Policy Making by Randomly Selected Citizens: The Perspective of Elected Politicians

An institutional reform proposal which recently seems to have gained currency is to introduce at least in some areas policy making by citizens’ assemblies, that is, population samples whose members are selected from the public by lot. Analysts and commentators offer an array of sound arguments in support of this reform. However, short of enacting this institutional change by revolution, its adoption relies on elected politicians-legislators accepting it. But policy making by randomly selected assemblies of ordinary citizens decreases the domain of policy areas decided by politicians, and consequently it reduces their authority and prestige. Under what circumstances, if any, would politicians consent to such a change in policy making? This is the issue investigated here. The paper explores a model of institutional choice under uncertainty and examines the payoffs of politicians-legislators under different policy rules. Using the spatial decision framework, it identifies circumstances where an elected politician may be willing to grant policy making powers to a randomly drawn assembly of citizens. The choice is found to depend on the interplay of the following factors: the probability that he wins the election and so he implements his ideal policy compared to the probability that the assembly implements his ideal policy; the cost of fighting an election versus the (lower) cost of policy making by a randomly selected assembly; the policy difference with his rival politician; the deviation between assembly decision and his preferred policy; the size of office rents from winning an election; and the marginal utility of the policy compared to the marginal utility from holding office.