Cass Sunstein has a new book out, #Republic. It is the very thoroughly revised and updated third edition of Republic.com, which was published in 2002. While the wider public began a serious debate on the impact on democracy of social media in particular and the internet in general only much more recently, Cass Sunstein had discussed some core issues already fifteen years ago. Among these issues are filter bubbles, self-segregation of citizens into distinct groups, and the problem of information-spreading cascades that could lead to huge numbers of individuals basing their political reasoning upon false beliefs.
Sunstein begins his argument by identifying some fundamental requirements for a properly working democracy. One of those requirements, he emphatically argues, is a public forum, a place where everybody is allowed to voice his opinion and where there is a likelihood of being confronted with unexpected or even unwanted opinions. This is important, because some still common misconceptions of democracy understand it merely as a mechanism for preference aggregation. Instead, Sunstein emphasizes the deliberative nature of democracy. Preference formation and alteration through debate, argument and exposure to competing points of view are the key elements.
Clearly, a well-functioning deliberative democracy does require high levels of effort and openness on behalf of its citizens. In juxtaposing consumer sovereignty and political sovereignty, Sunstein argues that the latter requires that political decisions meet certain standards of quality. If political sovereignty is to be established, he argues, then political decisions need to be justified by deliberation and sound arguments. To this end, he concurs with Louis Brandeis, who considered it a political duty of citizens to participate in public discussions and, to put it in economic language, invest some of their resourced into forming well-informed and reasonable political preferences.
Hoping that actual democracies can come close to an ideal discourse is one thing, preventing them from turning into something entirely different than an ideal discourse is another, more pragmatic approach. This is where electronic media and the internet enter the stage. Sunstein gives a broad, deep and analytic survey of what social scientists have found out about the impact of new media on the democratic process. There is the increasing polarization of US politics following the rise of MSNBC (for Democrats) and Fox News (for Republicans), there is the importance of group identity and its impact on group polarization, and there is a lot more. Some of the theoretical concepts discussed are applied to current developments, even to the recent election of Donald Trump. The reader learns a lot about the nature of the problems that may arise when political deliberation and formation of political preferences takes place through electronic media.
Some of the evidence presented is truly discouraging, such as the fact that news reports which attempt to correct false beliefs can even reinforce these beliefs if people, for partisan reasons, want to believe in them. On the other hand, there seem to be simple instruments that can increase trust and reciprocity among citizens, and ultimately reduce polarization. In this respect, Sunstein emphasizes the value of shared experiences, even in the form of symbolic events such as national holidays with a meaningful background; Martin Luther King Day in the United States is an example.
But still: exposing oneself to diverse sources and surprising pieces of information, and learning to discount pieces of information according to the quality of its sources, remains to a large degree a matter of personal discipline and willingness to avoid the pitfalls of today’s media. This is also the lever for the core argument of the book. Individuals as citizens might have good reasons to prefer institutions that restrict their actions as consumers. Political sovereignty dominates consumer sovereignty, and the question is how consumer sovereignty should be curbed in order to secure political sovereignty. Sunstein takes an (in my opinion wholly unnecessary) detour through Robert Frank’s theory of rat races in consumption in order to mildly discredit consumer sovereignty. This seems unnecessary because conflicts between consumer and policy choices that need to be managed through mechanisms of self-constraint can exist even if we take both of them seriously. There is no reason to discredit consumer choices.
When it comes to possible regulations designed to promote good democratic deliberation, Sunstein proposes measures such as requirements for educational programming on television, or for giving free air time to political candidates. On the more controversial side, he toys with the idea of giving out government subsidies to some, desired types of political deliberation, but not to others. Other interesting proposals include transparency requirements aimed at increasing airtime for public issues, or the establishment of independent monitors that rate journalistic quality, or the requirement of links on websites that lead to opposing opinions.
#Republic is an accessible book addressed at a broad audience. Nevertheless, its arguments in the first part of the book rest on a solid foundation of behavioral economics, social psychology and constitutional analysis. In the second part, when proposals for policy and institutional reform are discussed, the book becomes more speculative. To some degree, that is a necessity, because there simply is only very little broad, let alone quantitative empirical evidence on these issues so far.
But viewed from this side of the Atlantic, where we have some experience with some of the proposed measures, their effectiveness appears doubtful. For example, the existence of large, highly influential and taxpayer funded broadcast media, who have an explicit mission to educate the general public, has done little to prevent the rise of populist parties, and even of populism within established centrist parties. And Europeans of younger generations are probably not any less susceptible to living in segregated digital environments than their American peers.
Or consider direct democracy, which as such is judged by Sunstein from a very skeptical point of view. But what we do know from empirical research is that much depends on the specific design of the institutions of direct participation. A Californian style referendum, where citizens may be voting on dozens of items on a single trip to the voting booth, does little to encourage careful deliberation on every issue. A system where a referendum is a once-in-a-lifetime experience — Brexit, for example — gives voters little opportunity to learn how to use the political tool of direct democracy responsibly. But in a country like Switzerland, where referenda are commonplace and where the process leading up to the referendum is carefully structured, referenda normally are associated with long and careful public debates, and on average they produce reasonable and efficient outcomes. Of course, there are always some spectacular outliers, but you find those in purely representative systems, too.
The big, still open question is how to design incentives that encourage citizens to involve themselves in public debates, to vote responsibly and not on purely expressive or emotional grounds, and to react to dissenting opinions with respect, rather than attempts to ostracize those who voice them. #Republic makes a number of suggestions in that respect, but not all of them seem very convincing. This, however, is not the fault of Cass Sunstein. It merely reflects the fact that Behavioral Political Economy is still in its early stages, and we have little robust empirical knowledge so far on institutions that reliably strengthen political sovereignty, as defined by Sunstein.
A final point, which is admittedly speculative itself: Maybe the debate around new media and democracy underestimates the adaptability of individuals. When broadcast media surfaced in the early to mid 20th century, this had changed dramatically the way public discourse functioned, the way election campaigns were waged, and so on. But by and large, indivduals themselves learned how to use the broader and faster information flows, and democratic societies continued to work quite well — most of them, at least. We simply do not know yet whether people will develop similar competencies for the new media. As it happens, however, very recent research by Boxell, Gentzkow and Shapiro indicates that people who frequently use the internet are even less politically polarized than those who do not. Is the learning already taking place?
Certainly, more empirical research is needed before something about the robustness and generality of such results can be said. But chances are that even when it comes to the further development of democracy, our first knee-jerk reaction that tells us to regulate phenomena whose effects are not yet entirely clear, may actually be wrong.