What Do Mini‐Publics Prove?

Lottocracy is a novel political arrangement that replaces elected politicians with ordinary citizens selected at random. It is unlike any political system that currently exists, so it is natural to have anxiety about how it would work in practice. To mollify these concerns, many lottocrats point to the successful performance of mini‐publics as evidence that lottocracy would work well, if tried. This inference resembles a claim about external validity, I argue, and develop a framework for evaluating it by relying on work from the social sciences and philosophy of science about when such claims are warranted. With this framework in hand, I show that mini‐publics teach us little about how lottocracy would work in practice, although the strength of this conclusion depends on which specific version of lottocracy is under consideration. Mini‐publics, however, do provide evidence about the abilities of ordinary citizens more generally, undermining some of the strongest claims about their lack of political competence. The framework developed here applies to a broader class of arguments in political philosophy: arguments that use the performance of existing institutions to predict how similar but untried institutions would perform.