The Problem Isn’t Twitter. It’s That You Care About Twitter.

Political leaders believe that the views they encounter online are representative of the “general public.” They’re not.

Ever since Donald Trump made Twitter his preferred medium for communicating with the country, the platform has taken an outsize hold on the American imagination. Once a forum on which users could discuss the day’s news, Twitter now just as often sets the day’s agenda.

Being active on Twitter has practically become part of the job description for some of the most influential people in the country. Any politician, journalist, or CEO who does not engage with social media gives up a precious chance to shape the conversation. And any public or semipublic figure who fails to monitor what is happening on the platform risks missing attacks or accusations that can quickly find their way into the headlines of national newspapers and the chyrons of cable-news shows.

Obligation breeds habit and habit addiction. The most active Twitter users I know check the platform as soon as they wake up to see what they missed. Throughout the day, they seize on the little interstices of time they have available to them—on the way to work, or in between meetings—to follow each new development in that day’s controversies. Even in the evening, when they are settling down to dinner, they cheer attacks against their enemies, or quietly fume over the mean tweet some anonymous user sent their way. Minutes before they finally drift off to sleep, they check their notifications one last time.

It is not the mental health of Twitter addicts that most concerns me, though; it is the well-being of the nation they collectively rule. To decision makers who spend most of their days ensconced in an elite bubble, Twitter can seem like a way out, a clear window into pure public opinion. In reality, it’s an extreme distortion.

Each week seems to throw up another example of organizations capitulating to outrage mobs on social media, whether they originate on the left or the right. In the past year, CNN fired Marc Lamont Hill for controversial remarks about Israel and Disney dismissed (and then rehired) James Gunn over offensive jokes he tweeted a decade ago—in both cases due in part to anger on Twitter.

But while the best-known cases of social media influencing large institutions involve famous celebrities losing their jobs, the sway is just as strong in shaping the implicit assumptions and priorities of the country’s political class. The vast gulf between the great importance pundits ascribed to the Mueller investigation and the apparent disinterest with which most Americans have greeted its findings is Exhibit A.

Judging by the conventional wisdom on Twitter, the publication of the Mueller report should have been the defining event of the Trump presidency. If Mueller found Donald Trump guilty of obstruction of justice, the president’s approval ratings would tank. Conversely, if Mueller exonerated Trump, there would be a broad backlash against Democrats; Trump would then be well on his way to reelection in 2020.

Instead, the most anticipated news event of the year has barely left a trace in public opinion. According to Nate Silver’s FiveThirtyEight, the government shutdown, which affected the lives of millions of Americans, had a clear and immediate impact on Trump’s popularity; the Mueller report did not. In fact, 42 percent of people approved of Trump at the beginning of March, before Mueller delivered his report to Attorney General William Barr, and 42 percent approved of Trump at the beginning of April, after Barr released a summary of the report that seemed to exonerate Trump. Now that much of the report is public, the number stands at, yes, 42 percent.

According to just about every study that has been conducted on the question, Twitter is not representative in the slightest. The Pew Research Center, for example, has found that less than a quarter of Americans log on to Twitter with any regularity. And as The Atlantic’s Alexis Madrigal points out, those regular users differ from the wider population: “In the United States, Twitter users are statistically younger, wealthier, and more politically liberal than the general population.”

Politics Twitter is a bubble in itself. Among the minority of Americans who regularly use Twitter, a majority never tweet about politics. According to a 2016 study, fewer than one in five active Twitter users—which is to say about one in 20 Americans—report posting about politics “some” or “a lot” of the time.

According to a recent analysis by The New York Times, left-leaning Twitter users who regularly post about politics are richer, better educated, and less diverse than the Democratic Party as a whole. In fact, Twitter Democrats are about 50 percent more likely to have a college degree than the average Democratic voter, but only about half as likely to be African American. Among the overall Democratic electorate, less than 50 percent consider themselves liberal, as opposed to moderate or conservative. Among Democrats on Twitter, more than 70 percent do.

One study surveying the evidence for who talks about politics on Twitter, by Pablo Barberá of New York University and Gonzalo Rivero of YouGov, found that “users participating in the political discussion were mostly men, living in urban areas, and with strong ideological preferences.” Another study, by researchers at the University of Pennsylvania and the National University of Singapore, came to an even starker conclusion: “Only self-reported extremists appear to devote much of their Twitter activity to politics.”

Social scientists have known for decades that the most politically active citizens are highly unrepresentative of the population as a whole. On average, citizens are more politically engaged the more affluent, educated, and powerful they are. “The heavenly chorus” of those who write letters to their local newspaper, attend PTA meetings, or ring up their senator, the political scientist Elmer Eric Schattschneider once wrote, “sings with a strong upper-class accent.”

For that reason, key social and political institutions are always in danger of being captured by their loudest and most resourceful constituents. Small but highly ideological factions have repeatedly taken over political parties. On college campuses, radical students have had a greater influence than their more moderate classmates.

Even so, the outsize influence that small, unrepresentative groups now exert via Twitter is a particular source of concern. Until the advent of social media, decision makers were only confronted with their angriest detractors on specific occasions—at town halls, say, or on the Letters page in newspapers—for which they could mentally brace themselves. And when they faced the local teachers’ union or chamber of commerce, they were well aware that it represented a particular slice of their electorate.

Twitter is different in two key ways. Because it allows anybody to speak up, leaders of political and cultural institutions seem prone to believe that the views they encounter there are representative of the “general public.” And because so many influential people check their Twitter notifications dozens of times a day, the opinions they see there become the constant soundtrack of their life. When deciding what to think or how to act, leaders may find it harder to tune out angry professions of outrage on Twitter than the “heavenly chorus” of yore.

“When you’re on Twitter, every controversy feels like it’s at the same level of importance,” one influential Democratic strategist told me. Over time, he found it more and more difficult to tune Twitter out: “People whose perception of reality is shaped by Twitter live in a different world and a different country than those off Twitter.” (I granted the strategist anonymity in exchange for candor.)

If elected representatives treat Twitter as representative of public opinion, they will fail to be responsive to the actual views of their constituents; political journalists will obsess over scandals and debates that don’t interest most of their readers; and political campaigns may lose eminently winnable elections.

President Trump is a case in point. He has rightly intuited that a significant portion of the American population is anxious about the influx of immigrants in the country illegally. But egged on by his die-hard fans—who ensure that his tweets about immigration are especially popular, with many of them attracting more than 100,000 likes—he has embraced policies, such as separating children from their parents, that are rejected by a vast majority of Americans.

Trump’s likely Democratic opponents fare no better. They are right about the fact that most Americans would like a strong public option for their health insurance. But, encouraged by activists on Twitter, some have brushed away concerns about what Medicare for All would mean for existing insurance plans—even though polls suggest that a clear majority of Americans do not want to lose what they already have.

To win the White House in 2020, presidential candidates will need to both win over swing voters whose views diverge from those of their party’s base and mobilize like-minded supporters who rarely think (much less opine) about politics. Paying less attention to Twitter may be the key to both.

America’s political class now lives in a bubble that has been made more, rather than less, impenetrable by the technological changes of the past years. Instead of connecting America’s elites to ordinary people, Twitter has amplified the beliefs of a small band of hyper-political partisans.

The solution to this problem is a lot more straightforward and achievable than much of the hand-wringing commentary about social media would suggest: It is for political leaders—and everyone else—to keep Twitter in perspective. What’s dangerous to democracy is not the existence of a forum in which extremists can talk to, and shout at, one another—it’s the possibility that decision makers will confuse the forum for the real world, and in so doing allow extremists to shape real-world culture.

A few months ago, I started to notice just how bad an influence Twitter was having on my grasp on reality, my productivity, and my serenity. For a brief period, I considered quitting Twitter. But that didn’t seem like the right solution. For one, I like sharing my work and my views with my followers. For another, I doubted my resolve: Over the past years, I’ve seen too many writers make grandiose announcements about quitting Twitter—only to rejoin the platform a few weeks, or days, after their departure.

Instead, I opted for a more modest solution: While I still access Twitter from my desktop from time to time, I have deleted the app from my phone, and stopped it from sending any notifications to my email account. It’s faintly ridiculous just how much my quality of life has improved as a result. I now know much less about the latest controversy—but have much more time to read that book I’ve been meaning to turn to for ages. I miss out on a few good jokes or interesting links—but have started to detox from the feverish anger that reigns supreme on the hyper-political corners of the Twitterverse.

The Democratic strategist who described Twitter as a “different world” recently quit the service. And he feels the same way as I do: “Instead of retweeting pundits, I’ve started watching candidates doing speeches and town halls. It’s made me a better political strategist—not to mention a better friend, a better boyfriend, and a better human being.”

If the most influential people in the country would follow his lead, we might just wind up with a better country.