Hélène Landemore Princeton University Press, Princeton and Oxford, 2012, xxi+279pp., $39.50 / £27.95, ISBN: 978-0691155654.

Hélène Landemore’s Democratic Reason concludes that an inclusive representative democracy coupled with a well-functioning technocracy embedded within a vibrant liberal pluralist society is likely to outperform rule by councils of experts or benevolent dictators. The fact that she can plausibly present this as radical says something about the state of advanced democracies, beset by financial crisis, wars, environmental threats and the rise of rival regimes. It also highlights a persistent skepticism in democratic thought about the capacities of the people for competent self-rule. Landemore, by contrast, makes an optimistic case for collective intelligence in political life. In Chapters 2 and 3 she neatly surveys theories of the ‘incompetent multitude’, and the main historical expressions of the idea of collective wisdom in the work of Aristotle, Spinoza, Machiavelli, Mill and Dewey, among others. However, her aim in this excellent book is to use current social science to specify the idea of collective intelligence in the context of democratic institutions today.

Landemore argues that rule by the many is better than rule by the few or the one because it can harness ‘democratic reason’. Democratic Reason refers to the collective political intelligence of the many, which in turn is a function of individual epistemic competence and the cognitive diversity of the group. Cognitive diversity is central to her argument. Drawing on the formal theorems of Hong and Page (2001) and Page (2007), she claims that a group containing diverse perspectives, interpretations, heuristics and predictive models, will be better at making decisions than a less diverse group of people, even if they are individually smarter. As the key feature of democracy is that it is an inclusive decision procedure, and inclusion tends to increase cognitive diversity, democracies are likely to outperform rival regime types.

At the heart of the book – Chapters 4 through 7 – is an extended discussion of two democratic institutional mechanisms or ‘cognitive artifacts’: deliberation and decision by majority rule. Deliberation, she argues in Chapter 4, is a mechanism suited to problem solving. Take the real-life problem of preventing muggings on a bridge in New Haven, CT. She picks this case because it has an ‘obvious and intuitive’ (p. 98) procedure-independent standard of success. She tells how a group including police, regular citizens, engineers and accountants started with an obvious but poor solution of stationing a police car near the bridge, but that through a process of collective inquiry eventually identified the ‘less obvious and more compelling solution’ of solar powered lamps on the bridge (p. 101). The group deliberation was the occasion for bringing participants to recognize what in retrospect seemed obvious, even though no participant could have found it on their own. In this, as in other examples, such as the film 12 Angry Men, the value of deliberation within a cognitively diverse group is that it prevented the group from reaching a premature consensus on a sub-optimal solution.

Majority rule, discussed in Chapter 6, has its epistemic value as a prediction device. By ‘prediction’ Landemore does not simply mean foretelling a future event. Rather, a predictive judgment is characterized by its aim, which is ‘to figure out the right answer to a given question, where the right answer is defined as the factual or moral “truth” of the matter: Will the candidate be a competent president? Is the defendant guilty or innocent?’ (pp. 145–146) Landemore argues that where there is a better option according to a common measure, majority rule makes it likely that the group will decide for that option. In conjunction with a pre-voting deliberative phase, majority rule can ‘turn imperfect individual predictions into accurate collective ones’ (p. 146). She discusses three epistemic arguments for majority rule. The first is Condorcet’s jury theorem, which states that for a binary question with a single true answer, majoritarian outcomes track that answer as long as the voters have a better than even chance of being right, they vote independently and they vote sincerely. The second is the ‘miracle of aggregation’, the mechanism by which the weight of a cow, in Francis Galton’s famous example, can be successfully approximated by taking the average of the independent estimates of a large group of equally placed observers. She gives a subtle and illuminating account of these two arguments, but leans most heavily on a third, drawn from Hong and Page’s cognitive diversity model, which claims that a prediction (of future sales figures, for instance) is enhanced by a diversity of predictive approaches. In contrast with the first two approaches, the diversity model requires not independence, but negative correlation between the predictive models of the participants, such that ‘mistakes cancel each other out not randomly but systematically’ (p. 160).

In these central chapters, Landemore focuses on the collective effects of deliberation and majority rule. This yields many useful insights. In Chapter 5, she discusses group polarization and the confirmation bias in the light of Dan Sperber’s argumentation theory (p. 130). The tendency of people to seek arguments that support positions they already believe may be a bad individual strategy for finding the right answer, but it can contribute to collectively better answers because it supports ‘a more exhaustive exposure to the arguments supporting the different sides of an issue’ (p. 142). In chapter 7, she criticizes Caplan’s (2008) The Myth of the Rational Voter not only for its elevation of a ‘cultural and possibly ideological consensus’ (p. 202) among economists as the standard against which to assess citizen error, but also for its focus on the individual rather than the collective as the cognitive unit of interest. For this reason, she later concludes, much research on the uninformed voter is misguided (p. 241).

Landemore’s central argument has radical implications. It ‘blurs the line between so-called technical and political questions’, and suggests the epistemic value of increasing citizen input on technical questions like those related to GMOs, nuclear energy, pension reform, electoral systems and so on (p. 14). This raises the possibility that many decisions made by what Ober (2008, p. 1) calls ‘cloistered experts’ would be better made by randomly selected panels of citizens with similar information but greater cognitive diversity. However, she does not fully develop this connection to work on institutional innovations in the realms of policy and governance. Rather, she later declares, her central concern is not with the relation between democracy and the technocratic branch of government, but with the comparison ‘between democracy and oligarchy when both are equipped with a competent technocracy of that kind’ (p. 204).

Instead, Landemore focuses her attention on the claim of democracy as a way of finding the correct answers to questions of not only means but also ends. She recognizes that this is a jarring claim, and in Chapter 8 she makes a careful defense of ‘political cognitivism’, the idea that some political questions have a ‘better’ or ‘worse’ answer according to an independent standard of correctness (p. 210). Here she generalizes the argument from those situations where there is a shared problem and an ‘obvious and intuitive’ (p. 98) independent standard, as in her recurring example of the jury, to those situations where the nature of the problem is at least in part the thing to be democratically decided. Are political decisions really like jury deliberations, or guessing the number of beans in a jar, or predicting next year’s sales figures? Is the election of a president really a kind of collective prediction (p. 162; my emphasis)? Are ‘white supremacists’ really best described as ‘incompetent people’ (p. 115; my emphasis)? Whether or not readers find her discussion of these questions convincing, the expansion of the argument means that she effectively ends up focusing on democratic inclusion where we already have it – in the selection of a president, for instance – and paying less attention to inclusion in technical policy domains.

Democratic Reason is an important contribution to democratic theory. It shows the benefits of thinking of democracies as systems that respond and adapt to their environment by harnessing the information and perspectives of differently situated knowers. Landemore’s analytic method of isolating the purest forms of the mechanisms of deliberation and majority rule in order to identify and explain their distinct epistemic properties is highly effective. The arguments developed here will be especially useful to deliberative democrats as the field moves beyond the mistaken opposition of ‘deliberative’ and ‘aggregative’ models and toward a more differentiated ‘systemic’ account of democracy (Parkinson and Mansbridge, 2012). However, Landemore leaves open a number of important questions about how democratic mechanisms interact, and how they might operate together in a mixed regime. It also leaves open the question when and where to best make use of democratic reason, which remains partly a matter of ‘experience’, and ‘partly a gamble’ (p. 156). Democratic Reason ultimately appears as a tool that may serve rather than displace political judgment.