In his Introduction to A Citizen Legislature by Ernest Callenbach and Michael Phillips, Peter Stone commends the authors for doing “an excellent job of presenting the idea of a representative House — a House that will truly be “of the people” — as an inspiring piece of democratic reform.” On the other hand, such inspiring presentation does not meet the philosopher’s standard of a good argument: “Their [C&P’s] efforts to defend their proposal, however, have a number of shortcomings.”
Stone rejects the implied argument that descriptive representation is a desirable end. Descriptive representation is a means, not an end, Stone argues: “descriptive representation is desirable because — and only to the extent that — it contributes to the goal of good lawmaking.” And while some may reject this point of view, and argue that the symbolism of microcosm is indeed an end in itself, I think that would be an evasion of the main point. The essential function of government is to generate good policy – i.e., policy that promotes the interests of the population – and, as Stone asserts, sortition is a useful tool if and only if it can be expected to produce a government that generates good policy. (By “interests” I mean to encompass whatever a person, or a group of people, perceive, upon informed and considered reflection, as important or desirable.)
While C&P imply the “sortition as an end” argument, it is quite clear that, in addition, they believe that descriptive representation promotes representation of interests. However, as Stone points out, they fail to make a detailed argument of why this would be so. It may be that they considered the argument to be too obvious to require explication: “a parliament that mirrors the population will act as the population would” may seem at first glance like all that can be said about the matter. But, as often happens, this turns out not to be the case. Just like the intuitive argument for elections (“elections represent the people’s interests because they produce the government that people think will do the best job”) turns out to require further elaboration (and eventually falls apart under examination – see section 3 of the presentation here for a brief discussion), the mirroring argument, in its naive form presented above, is far from being convincing.
A fundamental point of weakness of the mirroring argument is that it not clear what the mirror is of. “Act as the population would”, implies a situation in which the “population would act” in one way or another. Presumably this envisions a hypothetical situation in which the entire population engages in a deliberation which has the all-to-all characteristics that are associated with a small-group setting. This is a very remote hypothetical. Specifically it involves the members of the population being able to listen to, understand the ideas of, and weigh and discuss the proposals of, what could run into millions of people. Thus the hypothesis here involves, among other conditions, a relaxation of the cognitive constraints of humans.
Once such a hypothetical mirroring target has been assumed, the remaining difficulty is showing that the statistical sample would be likely to make decisions similar to those of the target. But with the mirroring target being such a strange object, it seems that it would be rather difficult to make any claims about its properties, including in particular claims that it would produce decisions that are similar to those yielded by a specific, real world procedure. And, of course, even ignoring the remoteness of the target, asserting “similarity” between decisions is another thorny issue.
Beyond those difficulties with the mirroring argument, there are the well-known difficulties associated with aggregating preferences, which are often taken to imply that representative policy is simply impossible. Social choice theory claims to show that any way to convert preferences of individuals in a group into group preferences must violate one of several properties, all of which seem essential to group preference in order to be representative. Sortition, like any other procedure for setting group policy, will have, then, to violate one of those properties and thus any argument for its representativity – not just the mirroring argument – must fail.
In the face of such difficulties, Stone and his collaborator Oliver Dowlen propose (1, 2) to abandon the representativity justification for sortition. Instead they offer a different justification, namely that sortition is a tool to prevent domination and corruption in government. Stone and Dowlen do not explicitly state the difference between “defence of an open, fair, inclusive, rule governed polity” and representativity, but the former is apparently taken to be a more modest, and thus more realistic, task. Prevention of domination and corruption is compatible, it seems, with continued existence of political elites holding significant power as long as the “gap […] between a professional political caste and the people at large” is diminished. Indeed Stone approvingly quotes Dowlen as saying that
[t]he most significant and fundamental reason that lot is used in the selection of public officers is to inhibit the power that any individual or group of individuals might seek to exercise over that process of selection,
which seems like an exceedingly narrow function.
The narrow justification that Stone and Dowlen are apparently making is in fact rather similar in content and in language to the standard academic justifications for electoralism. These justifications are essentially adornments of the Schumpeterian elitist “theory of democracy” (in fact, theory of electoralism) with vague terms such as accountability, responsiveness, rule-of-law and non-domination. These apologia for the existing political system are often coupled with various touch-up proposals that would supposedly alleviate various deficiencies in the system. In the hands of Stone and Dowlen sortition becomes not much more than one of those proposed touch-ups. Any hope of moving beyond the fog of apologia for electoralism and toward a clearer view of what democracy is and how to attain it is given up upon at the outset together with the notion of representativity.