The Government of Chance is a magisterial compendium of the uses of sortition across the ages and across cultures and for varying purposes. It is a must read for anyone interested in the application of random selection for a whole range of social and political purposes. The examples are endlessly fascinating to this reader and, I suspect, to the book’s author Yves Sintomer. We hear the author’s views on the Athenian case, Rome, Florence, Venice, China, the United Kingdom, the United States, and on various practices all over Europe. We hear about Roman Saturnalia celebrations in which a young man was selected by lot to parade in royal clothing during the celebration and have his every desire satisfied for thirty days, after which he was “forced to slit his throat upon the divine altar” (62). In Venice, Sintomer details the intricate system for the selection of the Doge, in which many stages of random selection combined with election were part of a power-sharing arrangement among the aristocracy, one that persisted from the thirteenth century until 1797 (74–76). Sintomer also breaks down the many stages of decision making involved in governing the Florentine Republic: “As in Venice the objective was to appoint public officials in the most neutral way possible, thereby limiting conflicts and power struggles” (84). In the modern period, we hear a great deal about Anglo-American juries and their influence on Tocqueville (but curiously not on J. S. Mill).

Sintomer divides the modern era of deliberative minipublics—the random selection of some sort, combined with deliberation—into “three waves.” In the first wave, he includes deliberative polls, citizen juries, and consensus conferences (as well as planning cells in Germany). These processes are all said to employ random sampling and to have a merely advisory function. However, in this wave, he claims that “the political imagination of practitioners is limited and hence diffusion is consequently hindered” (204). They are social- science efforts not linked to social movements. Because the kind of opinion they produce “differs from the wider public sphere measured by public opinion polls,” Sintomer argues that “the relation between these minipublics and the macropublic is therefore problematic” (204).

I am not sure about Sintomer’s characterization of this “first wave.” First, setting my own work aside (as the originator of the deliberative poll), I would certainly not say that the late Ned Crosby (who invented the citizen jury) or the late Peter Dienel (who invented the planning cell) lacked “political imagination.” Indeed, that is what inspired them to invent new political institutions that continue to have influence to this day. Second, as with his later discussion of the second wave, he draws no distinctions among types of random sampling. A jury of 12 or 24 or a consensus conference of similar size—and usually consensus conferences are drawn from volunteers sorted by demographics, initially from newspaper ads—are all somehow random sampling, regardless of the response rate, or whether there is a sampling frame, or what data are collected at what point to establish its representativeness: they are all just “sortition.” Whether this is justified depends on the purpose of the institution, a point I return to later. Third, many of the “first-wave” projects were part of public decision processes. This was certainly true for the citizen juries, particularly on healthcare issues in the United Kingdom. The eight Texas deliberative polls conducted with the major utilities and the Public Utility Commission beginning in 1996 were part of an official decision process that led Texas from being last in the amount of wind power it generated to becoming first, by 2007, among the 50 US states. In many instances, these processes were more than the expression of counterfactual public opinion: they were an effective component of public decision processes.

I also question the neat divisions that Sintomer draws between these phases of “government by chance.” The “second wave,” we are told, increased the involvement of deliberative minipublics in party politics, and the “third wave” combined deliberative minipublics with direct democracy. Yet these initiatives are not easily separated chronologically. Sintomer notes that the Greek political party Pasok used a deliberative poll to select a candidate for parliament in 2006. Furthermore, the first national deliberative poll in Australia in 1989 was nationally televised to advise the public as it voted in a national referendum (on the “Republic”). In Sintomer’s schema, the link to party decisions is supposed to be in the second wave, and the connection to direct democracy is supposed to be in the third wave. But these projects were clearly in the period that he assigned to the first wave.

From my perspective, this marvelous book has other limitations, which perhaps are the inevitable result of its synoptic scope. Given that we live in an age of disinformation and the manipulation of public opinion, one of the key aspirations of deliberative minipublics should be to improve public will formation within these microcosms to find out (and perhaps make consequential) what the people would really think under reasonably good conditions. For this purpose, it is essential that the samples be of high quality and that both attitudinal and demographic data be collected about them from the start. If we don’t know where they start on the issue, why should we (or decision makers) listen to where they end up? Further, the design of these minipublics should have a track record of avoiding the key distortions that sometimes affect juries; for example, domination by the more advantaged and small-group polarization, as argued by Cass Sunstein in Going to Extremes (2009). Yet these defects of juries, now standard in critiques of deliberation, go unmentioned in Sintomer’s book. Nor does he offer criteria to distinguish credible random samples from convenience samples. Indeed, many of the models that are lauded in the book are obviously defective for representing the will of the community. For example, we are told that “the Berlin model, where the jury is partly drawn from the most active people in the community and partly selected by lot, is certainly one to be followed” (271). But what does such a small, hybrid collection of individuals represent? The activists are likely to dominate, and the overall group is some arbitrary mixture.

The problem may be that Sintomer gives pride of place to epistemic arguments for these deliberative minipublics and discusses representativeness almost entirely in terms of demographics. Perhaps for the “wisdom of the crowds,” the exact composition of a sample, so long as it is diverse, is not as crucial as it would be for a response to our debilitated processes of public will formation, in which disinformation and manipulation are influences everywhere as the public is trying to make up its mind. Hence, my perspective will not be shared by some readers, but I think that, in our current age, the manipulation and distortion of public opinion are prime challenges.

A last omission is the use of technology to perfect deliberation, make it more cost effective, and spread it to larger populations. All the projects Sintomer discusses were conducted face to face. But deliberation is moving online. Furthermore, online processes with AI-assisted video-based discussions are being used in many countries now to make deliberations more practical and to spread organized deliberation to the broader population. I believe the last section, “Sortition and Politics in the 21^st Century,” should have included a discussion of technology and how it might transform deliberation.

I have mostly criticized this book for what it does not include. But it is well worth reading for what it does include on sortition and the use of random sampling for democratic ends. I recommend it accordingly.