Title:Liquid Democracy. Two Experiments on Delegation in Voting

Abstract:Proponents of participatory democracy praise Liquid Democracy: decisions are taken by referendum, but voters delegate their votes freely. When better informed voters are present, delegation can increase the probability of a correct decision. However, delegation must be used sparely because it reduces the information aggregated through voting. In two different experiments, we find that delegation underperforms both universal majority voting and the simpler option of abstention. In a tightly controlled lab experiment where the subjects' precision of information is conveyed in precise mathematical terms and very salient, the result is due to overdelegation. In a perceptual task run online where the precision of information is not known precisely, delegation remains very high and again underperforms both majority voting and abstention. In addition, subjects substantially overestimate the precision of the better informed voters, underlining that Liquid Democracy is fragile to multiple sources of noise. The paper makes an innovative methodological contribution by combining two very different experimental procedures: the study of voting rules would benefit from complementing controlled experiments with known precision of information with tests under ambiguity, a realistic assumption in many voting situations.