The Chronicle Review

Social Disorder

Cass Sunstein exaggerates the web’s polarizing effects

Errata Carmona for The Chronicle Review

April 23, 2017

F or more than 15 years, in between serving as a law professor and as the nation’s top regulatory officer, not to mention writing a small library of books and articles, Cass Sunstein has investigated the extent to which innovations in media and technology are transforming American politics.

His Republic.com (Princeton University Press, 2001) examined the problem during a time when our ability to filter the news was something of a novelty. The sequel, Republic .com 2.0 (Princeton, 2007), explored similar questions in the context of the "blogosphere." #Republic (Princeton, 2017), his new book, replaces the guarded optimism of his previous works (including 2006’s Infotopia) with a world-weary resignation that serious damage may have already been done.

REVIEW

#Republic: Divided Democracy in the Age of Social Media By Cass Sunstein

(Princeton University Press)

Across those books, Sunstein has repeatedly invoked Benjamin Franklin’s famous description of American government: "A republic, if you can keep it." Sunstein seems less certain than ever that we can. The republic is under siege, he argues, because of our ever-increasing ability to isolate ourselves in ideological and partisan enclaves. These enclaves are not physical but virtual, and they are of our own design. With each click and link we share, we are passing along information to technology companies to better curate a unique "Daily Me."

Coined by the tech evangelist and MIT Media Lab co-founder Nicholas Negroponte in 1995, the "Daily Me" connotes a news site with information pertaining only to us, our likes and dislikes. What we do not want to see, the Daily Me excludes; what we want, it brings to the fore.

Once a mere metaphor, the Daily Me has now become reality, mostly but not exclusively via our Facebook feeds. For Negroponte, the Daily Me was a promising vision; for Sunstein, it is closer to a nightmare. For one thing, it impedes our ability to encounter information outside of our comfort zone. If we consume only content that aligns with our beliefs, democracy — which depends on compromise and a willingness to agree to disagree — becomes harder to sustain.

For another, the Daily Me reduces the frequency and power of our shared experiences. We spend our days only with the like-minded and do not build the bridges that connect us across the chasms of ideology and party. As Sunstein writes, "Unplanned and unchosen encounters often turn out to do a great deal of good, for individuals and society at large." Those encounters are becoming rarer.

The Daily Me, and our reliance on it, can leave real harm in its wake. It might make us hate members of the opposing political party. (This is known as "partyism" in the academic literature.) Some might eventually turn to violence, attacking those with whom they disagree. And, insofar as it restricts our choices to those that follow from previous choices, our reliance on the Daily Me might damage our freedom. One learns only what one already wishes to learn, sees what one already wishes to see. What appears to be infinite choice quickly becomes a restricted set.

A lthough most of the research referred to in #Republic was conducted before the 2016 election, Sunstein’s concerns have become only more urgent since then. The widespread availability of conspiracy theories, the ubiqui­ty of clickbait "fake news," and the polarization of political media all very likely played some role in the electoral outcome.

Sunstein is hardly the first to consider these questions. He is not primarily an empirical researcher (by my count, only one of the empirical papers discussed in the book was written by Sunstein himself). He is instead a voracious reader and synthesizer, clarifying the theoretical stakes of empirical work that others are pioneering. But he’s a fine writer, adept at turning dry academic research into (mostly) readable prose. His ability to articulate how media and technology pose threats to American democracy is one of his strengths.

As with many of his books, #Republic uses descriptions of other academic studies to buttress his main argument. The book sometimes reads as if other social scientists have been, intentionally or otherwise, engaged in testing out — and largely affirming — Sunstein’s fears. Indeed, many scholars from disparate fields, including political science, economics, communications, and psychology, have been investigating the phenomena that he writes about.

Yet relying on the empirical research of others has a major risk: It is unlikely that all available evidence supports the points on offer. This is certainly the case here. While Sunstein dwells on research that paints people as unwilling to think outside their partisan boxes, a rich literature has emerged that challenges this consensus — literature that Sunstein doesn’t engage with. For instance, as the political scientist John Bullock and others have shown, the partisan discrepancy on key factual questions shrinks considerably with small monetary incentives — meaning that partisans’ interpretations of the empirical world are not hopelessly divided. In similar work, the political scientist Seth Hill and his co-authors have demonstrated that contrary to prevailing assumptions, people have the capacity to learn political facts, even when the facts are complicated. Meanwhile, the "backfire effect," which proposes that people become more committed to misinformation when the facts challenge their partisan commitments, has been seriously challenged by several papers, including one I’ve co-authored.

 This isn’t to say that people don’t self-segregate, or that partisanship doesn’t affect judgment. They do, and it does. But Sunstein’s anxieties are frustratingly imprecise given the state of available evidence. For example, he seems to believe that the lack of overlap between left- and right-wing news sites is a major problem, going so far as to suggest that The Weekly Standard and The Nation should agree to trade advertisements on each other’s websites to introduce people to counterpartisan news. Recent research by Andrew Guess, however, has shown that Americans have a surprisingly large appetite for centrist online news. Only a small (if influential) segment of the population consumes the kind of ideologically driven news that Sunstein is most concerned about. (And his idea of cross-partisan outreach hardly seems persuasive: One imagines a Nation reader laughing off a link to The Weekly Standard.)

Sunstein’s pessimism may be precisely what is preventing him from offering bolder solutions. While acknowledging the enormous role that Facebook plays in the trends he finds so worrisome, he expresses hope that the company will take action on its own. Perhaps, he suggests, Facebook should offer users ways to see articles that challenge their political beliefs, or should send them into entirely unfamiliar territory.

But there need not be a "perhaps" about it — the company should compel users to confront such information. And if the company does not want to do so, then perhaps — as with previous firms that have loomed like colossi over the media and communications landscapes — the government should take a more active role in regulating it. (Well, at least a hypothetical future government, one concerned with the public interest.)

Such an aggressive approach to regulation might not be palatable to our former regulatory czar. But if we wish to keep the republic, it might be necessary.

Ethan Porter is an assistant professor in the School of Media and Public Affairs at George Washington University.