The Internet, argues Cass Sunstein, has had a polarizing effect on democracies. Although it has the capacity to bring people together, too often the associations formed online comprise self-selecting groups with little diversity of opinion, explains the Frankfurter professor of law. This confounds the constitutional vision of the founding fathers through a perversion of the notion of free speech. Such environments reinforce preexisting viewpoints, undermining the constructive dialogue that promotes progress in democracies. 

Speaking on September 17—Constitution Day—Sunstein (who is bound for Washington; see Brevia, page 51) said the founders made only one “truly original contribution” to constitutional thought. Their predecessors, influenced by Montesquieu, thought that successful self-government required everyone to be alike. The founders, in contrast, believed heterogeneity and diversity constitute a creative force. “When Hamilton explained the system of checks and balances with what he called ‘the jarring of opinions’ in the legislative branch,” Sunstein noted, “he said that it promotes circumspection and deliberation, and serves to check the excesses of the majority.” This idea “turns traditional republican thought on its head.”

Protection of free speech is one element allowing Hamilton’s “jarring of opinions” to succeed. But Sunstein worries that the conception of free speech emerging in today’s communications market emphasizes “an architecture of control…by which each of us can select a [customized] free-speech package.” Google News asks, “[W]hy not set up your pages to show you the stories that best represent your interests?” The New York Times offers “Mytimes”; Amazon and Netflix employ collaborative filtering to ensure “a kind of personalization…by which your communications universe can be yours.” (MIT’s Nicholas Negroponte first identified this “daily me.”) The resulting self-segregation creates numerous small republics of like-minded individuals of the sort Montesquieu preferred, but the founders considered “destructive of self-government….” 

Sunstein buttresses his argument with data from three studies he has worked on in the last decade. In the first, he and colleagues assembled a group of liberal-minded citizens from Boulder and a separate group of conservatives from Colorado Springs to discuss climate change, same-sex civil unions, and affirmative action. “We were particularly interested,” he says, “in finding what would happen to the private, anonymous statements of views expressed” before and after the discussions. On each issue, the like-minded groups became more extreme and the internal diversity of views “evaporated,” Sunstein reports. Pre-deliberation, for example, some liberals wanted to know more about the costs, especially for the poor, of an agreement to reduce greenhouse gases, and some conservatives were open to same-sex civil unions. Post-deliberation, the diversity of views on all three issues dropped precipitously.

Sunstein found a similar effect within juries, and even among federal judges on courts of appeals panels. When comparing the voting records of judicial appointees, the split between Democratic- or Republican-appointed judges increased from 10 percent on mixed panels to 30 percent on panels consisting exclusively of single-party appointees.

These findings suggest, he says, that free speech is not enough to ensure a healthy democracy. Important as well are “unchosen serendipitous, sometimes disliked encounters with diverse ideas and topics,” as well as “shared communications experiences that unify people across differences.” Public spaces such as city parks and sidewalks provide the “architecture of serendipity” that fosters chance encounters with a “teeming diversity” of ideas. Newspapers, magazines, television, and radio—which Sunstein calls the “great general-interest intermediaries”—played a similar role in the twentieth century. “If you are reading a daily newspaper, not online, the real thing,” he says, “chances are your eyes will come across a photograph or a headline that will attract your interest, produce curiosity, make you read maybe a paragraph, and eventually an article and conceivably change your life”—the sort of thing your Google News feed filters out.

The shared “general-interest intermediaries” not only exposed readers to diverse topics and points of view, but created “a shared experience, a social glue,” Sunstein believes. In their absence, the current system of self-sorting—only 2 percent of Daily Kos readers, for example, are self-identified Republicans—diminishes the serendipity that alerts us to “the occasional, maybe infrequent legitimacy of the concerns of our fellow citizens.”

Yet the “new technologies here are more opportunity than threat,” Sunstein suggests, “and what is limiting the realization of the opportunity is the absence of relevant ideals in the minds of the people who are using and developing and innovating [these] technologies.” For a partial solution to the problem, he says, Americans must “recover our constitutional aspirations as citizens and as providers of information.” While not denying market pressures—“the information we receive is a product of what information we demand”—Sunstein advises seeing the notion of the “daily me” as “a kind of science-fiction story rather than as a utopian ideal.” And, he says, we should create twenty-first-century equivalents of the kinds of public spaces and institutions where diverse people will congregate.