Civilizing Social Media

Strategies for depolarizing political discourse


Excerpted from Breaking the Social Media Prism: How to Make Our Platforms Less Polarizing by Chris Bail. Copyright © 2021 Chris Bail. Published by Princeton University Press. All rights reserved. Reprinted with Permission.
One of the most ancient ideas in Western thought is that rational deliberation will produce better societies.  The notion that societies function more smoothly when people form their opinions based on a wide range of evidence has become part of the bed-rock of democracy. Like so many of the ideals we hold dear today, this idea gained momentum during the Enlightenment. Reason had helped scientists conquer so much of the natural world, philosophers such as Denis Diderot argued—so why not use it to build better societies as well?
    A key venue for this progressive idea was the salon. Invented in Italy in the seventeenth century and later popularized in France in the eighteenth, salons were small-group discussions about current events organized by influential elites. Equal parts wit and erudition, salons were a critical precursor to modern democracy, some historians believe, because they provided people with a forum to discuss shared challenges. But other historians, such as Dena Goodman of the University of Michigan, argue salons were simply an excuse to engage in licentious behavior.

“Discussing the War in a Paris Café,” an illustration by Fred Barnard (1846–1896) from the 17 September 1870 issue of the Illustrated London News. The scene is during a brief interim between the Battle of Sedan and Siege of Paris in the Franco-Prussian War (1870–1871).


    Even if salons were not the crucible of democracy, that is how many social scientists describe them. According to the German sociologist Jürgen Habermas, salons laid the groundwork for the systems of mass communication that emerged in the twentieth century. Newspapers, radio, and television allowed for the emergence of the public, Habermas theorized, because they provided societies with forums expansive enough for large groups of people to deliberate about the issues of their day. By spreading information more efficiently, Habermas argued, these new  technologies facilitated a better competition of ideas. Such beliefs—which echo those of other prominent scholars of the public sphere—provided the foundation for modern theories of public opinion. It is not surprising, then,  that early observers of the Internet celebrated the potential of social media to scale up salon culture, creating a massive and open marketplace of ideas. Social media sites were not only open to anyone, these observers argued, but they also lacked the conventional gatekeepers (television producers, newspaper editors, and so on) who tightly policed the boundaries of the public sphere. What is more, social media offered people seemingly endless access to information they could use to form their views. And social media could also allow users to discuss such information with a much more diverse group of people than they might encounter in offline settings.
    This heavily idealized vision about social media may now seem whimsical. But the same logic that propelled these prophecies—that connecting people more easily will lead to more effective democracy—continues to motivate many technology leaders. Mark Zuckerberg, Facebook’s CEO, reportedly believes that Facebook users can effectively deliberate about what should be labelled as fake news, even though the term itself is a political football. Similarly, Jack Dorsey, Twitter’s CEO, has considered tweaking his platform’s algorithm to expose people to more diverse views because he believes that would increase moderation.
    Unfortunately, social media are less like an eighteenth century salon and more like a sprawling football field on which our instincts are guided by the color of our uniforms instead of our prefrontal cortexes. In fact, the tendency for our political identities to guide our opinions—instead of the other way around—was identified years ago in a clever experiment by Stanford University psychologist Geoffrey Cohen. In 2001, then at Yale University, Cohen recruited liberal and conservative students to evaluate hypothetical welfare policies. One of the policies offered very generous benefits, in line with the typical preferences of Democratic voters. The second policy was much more stringent, consistent with typical Republican preferences for economic redistribution. But in the study’s treatment condition, Cohen attributed each hypothetical policy to the “wrong” political party. Democrats strongly endorsed the policy that they were told was supported by their party, regardless of whether the benefits described were generous or stringent. Similarly, Republicans preferred the policy that they were told was endorsed by their party, notwithstanding its content.

Surrounded by members of the French court in a Paris salon in the 1780s, Benjamin Franklin is being crowned with a laurel wreath. Lithograph is credited to Anton Hohenstein, published in the 1860s.


    This experiment and many subsequent studies indicate that Americans are actually much less polarized about many social policy issues than most people realize. Though rates of disagreement about social policies have remained relatively steady over the past few decades, our attitudes toward each other have become much more negative. Since 1960, the American National Election Studies have asked thousands of Americans whether they would be upset if their child decided to marry someone from the opposing political party. In 1960, only 5 percent of Republicans and 4 percent of Democrats reported that they would be displeased in this scenario. But in 2010, the shares were 27 percent of Republicans and 20 percent of Democrats. And in 2018, the shares were nearly 50 percent of both Republicans and Democrats. Social science has produced a litany of troubling indicators of how easily our political identities override our rational instincts, our ability to empathize with each other, and even our ability to connect with each other around nonpolitical issues. For example, one study showed that Democrats and Republicans will accept lower financial compensation in online labor markets to avoid working with each other. Other studies reveal that Democrats and Republicans award hypothetical jobs and college scholarships to members of their own party, even if those candidates are less qualified than others from the opposing party. Political identity has even become so central to our sense of self that it shapes to whom we are attracted. When a group of political scientists randomized people’s party affiliation and asked respondents to judge how physically attractive those people were, they discovered that respondents rated members of opposing parties less favorably—even when they were judged to be very attractive by other respondents when the people were labeled as belonging to the same party as the respondent. Because of the mounting evidence that our political identities shape the way we understand the world around us, social scientists have mostly abandoned the idea that people dispassionately deliberate about the merits of each other’s arguments.
    But there is another, even deeper, problem that social scientists have not yet solved. Our political identities are not simply a jersey that we put on each time we log onto social media. Instead, our identities evolve as we interact with information in our news feed and with other social media users. We humans are unusual creatures because we care so deeply about what other people think about us. Consciously or unconsciously, we expend unreasonable amounts of energy presenting different versions of ourselves in varied social settings to figure out which ones “work.” Our obsession with social status long predates social media, of course, and some people care much more about gaining status on social media than others do. But we care so much about our identities because they give us something that we all strive for: a sense of self-worth. We use social media platforms as if they were a giant mirror that can help us understand our place within society. But they are more like prisms that bend and refract our social environment—distorting our sense of ourselves, and each other.
    The most pernicious effect of this “social media prism” is that it exacerbates what social scientists call “false polarization.” This concept first emerged in the United States in the mid-1990s, amid the culture wars that pitted the family values of the Republican Party against the Democrats’ belief in redistributing wealth to the poor. The two parties traded increasingly negative attack ads about each other’s candidates, and elected officials engaged in a series of seemingly intractable debates about abortion, welfare, and crime, among other divisive issues. Newt Gingrich, a Republican who served as the Speaker of the House of Representatives, led the first campaign to impeach a U.S. president for lying about having a sexual affair. Cable television amplified partisan conflicts. These developments might seem tame by today’s standards, but at the turn of the twenty-first century, many people worried that the growing culture wars might never stop.
Our World in Data

    But in reality, the culture wars were not growing. Though Americans were divided about contentious issues such as abortion, the sociologist Paul DiMaggio, then at Princeton University, and his colleagues discovered that the rates of disagreement about this and many other divisive issues did not increase between the 1970s and 1990s. Meanwhile, the social psychologist Robert Robinson, then at Harvard University’s Graduate School of Busi-ness Administration, was leading a team that was about to make a parallel discovery. Robinson’s team recruited college students who took liberal or conservative positions on abortion and racial conflict. In addition to measuring each student’s opinion on a range of questions related to each issue, the team also asked the students to estimate what people from the opposing party thought about each issue. The results would not have surprised DiMaggio and his colleagues: both liberals and conservatives drastically overestimated the difference between their views and those of the other side. They also underestimated the amount of difference in views within their own side.
    Robinson and his team revealed that false polarization was rampant among college students, but what about the rest of the country? In 2016, the political scientists Matthew Levendusky, of the University of Pennsylvania, and Neil Malhotra, of Stanford University, searched for evidence of false polarization among a nationally representative group of Americans. On average, the researchers found that perceived differences between the two parties were roughly two times larger than the actual differences. This discrepancy held across a range of topics that Robinson’s team had not studied—including immigration, tax reform, free trade, and campaign financing. Later research showed that the perception gap between Democrats and Republicans broadened at the same time as the culture wars were escalating in the mass media. Though false polarization was negligible in 1970, the political scientists Adam Enders, of the University of Louisville, and Miles Armaly, of the University of Mississippi, found that it grew by nearly 20 percent over the next forty years. What is more, they discovered that false polarization explained why Democrats and Republicans came to view each other so negatively over this time period. The more people misperceived the views of people in the other party, the more they disliked the members of that party. The latest research suggests that Democrats and Republicans also overestimate how much people from the other party dislike them. When we think that people in the other party dislike us more than they actually do, it makes us more likely to dislike them. A key strategy for reducing political polarization is to find ways to help members of opposing political parties correct the misperceptions they have about each other.
    If previous research is our guide, closing the perception gap between Democrats and Republicans should be a top priority for reducing polarization. Fortunately, many social scientists have studied how such gaps emerge. In my 2014 book, Terrified: How Anti-Muslim Fringe Organizations Became Mainstream (Princeton University Press), I analyzed how the mass media warped public opinion of Muslims after the September 11 attacks, for example, by amplifying the voices of emotional extremists at the expense of the more moderate majority. In The Outrage Industry: Political Opinion Media and the New Incivility (Oxford University Press, 2014), Tufts University professors Jeffrey Berry and Sarah Sobieraj show that such angry outbursts were almost completely absent from major US newspapers in the 1950s. By 2009, however, the average newspaper article included nearly six statements of outrage. Mass media can not only amplify extreme voices, they can also create misperceptions about the scope of polarization. Levendusky and Malhotra scanned dozens of newspapers, television show transcripts, and magazines to count the number of mentions of polarization, uncivil discourse, or lack of compromise. The researchers discovered a twofold increase in such mentions between 2000 and 2012. When they exposed people to such messaging in an experiment, they discovered that their subjects were more likely to overestimate the amount of political polarization in the United States than those who had not been shown such messages.
     If legacy media outlets fomented false polarization, then social media has sent it into hyperdrive. According to a recent study by the Pew Research Center, 73 percent of tweets about politics are created by just six percent of all Twitter users in the United States—and this small group of people have more extreme views. The link between social media and false polarization becomes even more vivid in cross-national perspective. In 2016, a group of fourteen scholars examined the gap between perceived and actual polarization in ten countries. They discovered online news consumption was the strongest predictor of false polarization in nearly every country. Social media also exacerbate the mass media’s contribution to false polarization. Journalists often use social media to monitor public opinion, and this makes false polarization even worse.
    How can we disrupt the feedback loop between social media and false polarization? It has become fashionable to argue that we must all delete our accounts in order to counter polarization online. But this clarion call is simply unpractical. Americans—and particularly young people—are using social media more than ever. We will not stop using social media any sooner than we will stop caring about our identities and social status. Instead, we need to think more about how the design of our platforms shapes the types of identities we create and the social status we seek. Only then will we be able to turn down the volume of extremism on our platforms and increase moderation.
    Asking how the design principles of social media shape social cohesion is revealing. What is the purpose of Facebook? The company tells us its mission is to “bring the world closer together.” But the platform began as a sophomoric tool that Harvard undergraduates used to rate each other’s physical attractiveness. What is the purpose of Twitter? Its motto is to “serve the public conversation,” but it was reportedly built to help groups of friends broadcast short messages to each other. What is the purpose of Instagram? We are told it is to “capture and share the world’s moments.” But the app was originally called “Burbn” (as in the drink) and was built to help people make plans to hang out with their friends. Hopefully, my point is already clear: Should we really expect platforms that were originally designed for such sophomoric or banal purposes to seamlessly transform themselves to serve the public good? Should we be surprised when they create the kind of leaderless demagoguery from which anyone can invent a kind of status, no matter how superficial or deleterious to democracy? Is it any wonder that people find themselves so rudderless on social media, when there is no common purpose for posting in the void?
    Imagine if we created a platform on which status was tied to a more noble purpose. Imagine a platform that gave people status not for clever takedowns of political opponents but for producing content with bipartisan appeal. Once we articulate the purpose of our platforms more clearly, we can build such principles into the architecture of the entire system. Instead of boosting content that is controversial or divisive, such a platform could improve the rank of messages that resonate with different audiences simultaneously. Instead of recommending that people follow others who already share their views, a platform could expose them to those whose messages cross social divides. A more noble platform could also fight false polarization, using the type of tools that we have developed in the Polarization Lab at Duke University. “Like” counters could be replaced by meters that show how people from across the ideological spectrum respond to people’s posts in blue, red, and purple. Artificial intelligence could be used to ask people who are about to post uncivil or ad hominem content to reflect upon their goals or to help people rephrase their messages using values that appeal to the other side.

The most effective bipartisan communicators on Twitter

Chris Ball

    Needless to say, not everyone would use a platform on which you gain status for bridging political divides. But that may be a good thing. The trolls and extremists who gain notoriety on other platforms for taking down political opponents could not be entirely banned from the type of platform I am envisioning, but it would be a much less rewarding place for them to play. Instead of gaining attention for taking people down, their posts would be lower in rank because they only appeal to one side. Regulating extreme content would also become easier. Once platforms have a purpose, such as bridging partisan divides, it will be much easier to make policies to define acceptable and unacceptable behavior. Instead of moderation policies that are so broad that they are difficult or even impossible to enforce, a platform built to depolarize users could moderate any posts that were personal affronts or uncivil. These terms would still be difficult to define, but I think that a platform guided by top-down principles that are transparent for all users to see when they sign up will always be more effective than a platform that relies upon norms to emerge from the bottom up—particularly when the bottom is composed of extremists whom moderates are unable or unwilling to police. Calling for a new form of social media may seem like a moon shot, given Facebook’s enormous market share. But taking the long view teaches us that platforms come and go. Friendster, the first broadly successful social media site, was once so popular that Google offered to buy it for $30 million. Only two years later, MySpace supplanted Friendster, eventually becoming the most visited site on the Internet in the United States by 2006. And Facebook once seemed unlikely to spread beyond the elite college campuses where it was born.
    In more recent years, Facebook has also shown signs of vulnerability. It took five years for Facebook to supplant MySpace at the top of the social media hierarchy. But beginning four years later, Instagram attracted fifty million users in its first two years of existence, prompting Facebook to acquire the new platform. But even Facebook and Instagram’s combined monopoly on the selfie market did not go unchallenged for long: by December 2013, Snap-chat users were sending more than 400 million messages per day, and by April 2015, the platform had more than 100 million monthly users. Similarly, TikTok arrived in the United States in 2018 and surpassed 100 million users just two years later.
    While the idea of building a decentralized platform to reduce polarization from the bottom up is romantic, I don’t think it is realistic. A full-fledged effort to create a platform to depolarize us would require a significant investment from a major funder. Governments would be ideal candidates, and proponents of publicly funded platforms, such as the internet activist Ethan Zuckerman, have argued that they could be funded by a digital advertising tax. However, if governments do not step in, there is money to be made by entrepreneurs willing to bet that a place where people can build reputations as effective bipartisan communicators will be more attractive to businesses, governments, and nonprofit organizations alike. As the United States confronts the most daunting challenge to its social fabric in a generation, developing a reputation for bridging partisan divides may become an increasingly valuable social asset. And, of course, it is not too late for existing social media platforms to be first to adopt many of the principles outlined above.--CB

Recent Stories

The way they live, the food they eat, and the effect on us

A true but unlikely tale

Story and Photographs by William Rowan

Increasing day length on the early Earth boosted oxygen released by photosynthetic cyanobacteria.

Genomic evidence shows that Denisovans and modern humans may have overlapped in Wallacea.