James Fishkin has spent three decades endeavoring to make public policy in democracies around the world more responsive to the considered opinions of their citizens. He developed Deliberative Opinion Polls (DOPs) as the primary tool in service of this goal. DOPs bring random samples of regular citizens together under favorable conditions in order to deliberate in depth about pending policy issues. He argues that the results reveal the informed “will of the people.”
Democracy When the People Are Thinking is Fishkin’s fifth book-length discussion of deliberative democracy, and in many ways it is his best. Much of the territory will be familiar to readers of his previous work. The themes and arguments, however, have been refined and integrated into a more complete account of the ways in which DOPs can promote democratic legitimacy and effective public policy.
The book is divided into four sections. Part I is a brief introduction. Part II lays out Fishkin’s most up-to-date theoretical account of deliberative democracy and contrasts it with what he sees as its three main rivals: party competition, elite deliberation, and participatory democracy. Part III discusses four in-depth empirical case studies of DOPs, and Part IV presents a miscellany of 15 short topics and cases.
In Part II, Fishkin posits four core democratic principles: (a) political equality, (b) mass participation, (c) deliberation, and (d) nontyranny (roughly, rights that constrain democratic choice). In one of his most consequential, but (here) unargued, moves, he claims that we can effectively pursue only two of the four principles at a time, concluding that there are only four viable forms of democracy: competitive elitism (a+d), elite deliberation (c+d), participatory democracy (a+b), and deliberative democracy (a+c): “Variations that aspire to more . . . are either unworkable or merely utopian” (p. 24). Unsurprisingly, Fishkin argues against the relative desirability of the first three, settling on deliberative democracy as the most attractive option available. He stays true to his self-imposed constraints, conceding that deliberative democracy “is usually agnostic about” mass participation and nontyranny (p. 24).
Herein lies our first question about Fishkin’s argument. Whatever the merits of claiming that we cannot pursue more than two principles at a time (a topic we return to later), it seems descriptively inaccurate to say that deliberative democrats are “usually agnostic” about rights and mass participation. Certainly this would not describe Jürgen Habermas, John Rawls, Joshua Cohen, John Dryzek, Jane Mansbridge, or Amy Gutmann and Dennis Thompson, among other major deliberative theorists. Indeed, it is Fishkin’s conception of deliberative democracy that is, despite his enormous stature in the field, rather idiosyncratic in this respect. Idiosyncratic, of course, does not mean wrong. As mentioned, Fishkin claims that rival theories that aspire to more are unworkable (e.g., Rawls, p. 154) or merely utopian (e.g., Habermas, p. 157), which allows him to stay laser-focused on the potential of DOPs. But for deliberative theorists who do not stay agnostic about mass participation, rights, elite deliberation, and party competition, the worry is that Fishkin has developed a theory of DOPs more than a full theory of deliberative democracy.
A successful theory of DOPs, however, is still tremendously important, since DOPs can lessen the tension between democratic self-government and good policy outcomes. In Part III, Fishkin shifts to a more concrete discussion of four sustained case studies of DOPs in action: California, Mongolia, Uganda, and the European Union. These are deeply impactful additions to the deliberative democracy literature, to say nothing of their contributions to democratic practice. Each extends DOPs into new theoretical and practical terrain where they had not proven their mettle heretofore. The California DOP focused on deliberative agenda setting for more authoritative ballot initiatives. The Mongolian DOP extended random sampling into participatory budgeting techniques. The Ugandan initiative tested whether DOPs could work in contexts exhibiting extremely wide variation in wealth and educational attainment. And, finally, the Europolis DOP attempted to overcome the Tower of Babel that might attend deliberation across diverse linguistic and cultural divides. All four projects were broadly successful on the criteria that Fishkin uses to assess his DOPs (inclusion, choice, deliberation, and impact).
In our view, these and other DOP successes are among the most salutary innovations in democratic practice over the last few decades. Our concerns, then, are not with DOPs per se, but rather with how they get interpreted. In particular, we worry that DOPs do not warrant the very strong interpretation of revealing “the will of the people,” a phrase used repeatedly throughout the book. This term is historically encumbered, as well as theoretically and operationally ambiguous in myriad ways. Yet it carries tremendous rhetorical punch: One presumes that it is very bad to subvert “the will of the people,” and so we might reasonably infer that DOPs should carry nearly dispositive normative force. It is crucial, then, to get clear about what DOPs do and do not establish. Our best reconstruction of what DOPs can hope to recover is roughly: an estimate of the marginal distribution of mass opinion if the issue under consideration were subject to a thoughtful plebiscite. Even if we were to grant that DOPs do reliably recover such a distribution, however, there are a number of reasons to be wary of equating that with the will of the people.
First, Fishkin measures opinion privately, through anonymous surveys. Participants are not required to connect their responses to public deliberation. He offers some good reasons for this measurement strategy, but the result is something in between aggregated private opinion and public opinion in the full sense developed by Habermas and others. Second, and relatedly, Fishkin’s focus is on a simple count of informed individual opinions, rather than any systematic analysis of the content and quality of the public justifications themselves. Finally, the theory explicitly “remains agnostic” about mass opinion (other than a brief reprise of his previous book, Deliberation Day (2004)). But for most deliberative democrats, a decision is justified through reasoned justification to all who have to live under the decision, not a random sample of those who have to live under it. Just as standard polls are not sufficient to stand in for elections, DOPs are not sufficient to represent “the will of the people.”
Apart from our theoretical concerns, we also see empirical obstacles to Fishkin’s claim that require deeper inquiry. The inferences from his deliberative events rely on a set of strong assumptions, at least three of which need more justification. The first is the exclusion restriction, or whether assignment to the deliberative condition itself matters for outcomes. If this assumption is not met, then the deliberative process itself may not be the only causal factor at work, calling into question both the normative value of the events and their efficiency vis-à-vis less expensive ways to elicit opinion. The second assumption is noncompliance, or the fact that all experiments, especially elaborate field experiments, face selection, noncompliance, and nonresponse issues that confound straightforward inferences. Finally, there is the Stable Unit Treatment Value Assumption (SUTVA), which refers to whether assignment of others to one’s discussion group aggravates the intrinsic path dependence of deliberation. Failing SUTVA would call into question whether it is even possible to recover a unique will of the people. Because many of Fishkin’s inferences depend on these three assumptions, it makes sense going forward to focus on building more concrete, empirical justifications for them.
Part IV is full of fascinating facts and insights—too many to summarize and engage in detail. Toward the end of the book, however, Fishkin briefly addresses the so-called systemic approach to deliberative democracy. He does not embrace that approach, though he does offer some suggestions as to how DOPs might augment systemic thinking in practice. The suggestions are well taken, and point toward potential institutional innovations. However, they mostly stay focused on how variations on the DOP might help secure equality and deliberation at each site of the system. Doing so, again, strongly privileges the principles of equality and deliberation at the expense of mass participation and nontyranny (along with other desiderata that the systemic approach was developed to incorporate).
By largely eschewing the systemic approach, Fishkin occludes the ways that we might be able to aspire to an enlarged democratic theory—one that does not have to remain agnostic about any core principles. Showing that tensions and trade-offs arise when we try to promote more than two principles does not prove that they relate in zero sum terms. Indeed, there is reason to think that the principles reinforce each other as often as they stand in tension: For example, certain rights (nontyranny) are essential for achieving equality, and likely contribute to people’s willingness to participate on a mass scale. Working out the complicated ways in which democratic principles interact to promote sound and legitimate governance should be, in James Madison’s words, “the great object to which our inquiries are directed.” Despite our differences and reservations, however, there is no doubt that Democracy When the People Are Thinking contributes powerfully to Madison’s ongoing project.
