Flip a Coin or Vote: An Experiment on Choosing Group Decision

Abstract

Before a group can take a decision, its members must agree on a mechanism to aggregate individual preferences. In this paper we present the results of an experiment on the influence of private payoff information and the role of the available alternatives on individuals' mechanism choices in such group choice situations. While efficient mechanisms are desirable, we experimentally show that participation constraints can prevent their implementation. We find strong indications that individual preferences for choice rules are sensitive to individual expected payoffs. Our results highlight the importance of considering participation constraints when designing choice institutions.JEL classification: C91,C92, D70, D82

Interpreting the Will of the People: A Positive Analysis of Ordinal Preference Aggregation

“…Are money-metric measures of well-being compelling when consequences are non-monetary? Another potential line of inquiry would investigate the relationships between preferences over rules revealed by choices over outcomes, preferences over rules revealed by choices over rules (as in Engelmann and Grüner, 2017;Hoffmann and Renes, 2017;Engelmann et al, 2020), and preferences over rules implied by approval of axioms (in the spirit of Nielsen and Rehbeck, 2020 and others). It would also be of interest to investigate how awareness of manipulability influences rule selection.…”

“…Other related work elicits subjects' preferences over (five or fewer) voting procedures, rather than over outcomes (Engelmann and Grüner, 2017;Hoffmann and Renes, 2017;Engelmann et al, 2020). 11 In these experiments, it is up to the subjects to imagine what each rule might imply for any particular preference profile.…”

Minority Protection in Voting Mechanisms – Experimental Evidence

“…across sessions, for the second part of an experimental session. 21 In each round of the second part, the chosen distribution was shown to the participants and the valuation of each participant was drawn according to the distribution. The participants were informed about their own realized valuation but not about the realized valuations of others.…”

“…4. Probabilities matter: Although distributions with the same outcomes but different probabilities of these (for example, distributions 1 and 6 in Table 1) have the same theoretically optimal rule, 21 To make sure that subjects could not lose money, we divided the distributions into three groups, according to the largest possible negative value (distributions 1, 2, 6, 7, 12, 20, 21 in group 1; distributions 3, 8, 9, 13, 14, 18, 19 in group 2; distributions 4, 5, 10, 11, 15, 16, 17 in group 3), and then randomly selected one distribution from each group. With this procedure, the probability that a given distribution is selected for the second part is the same as with unrestricted selection of triples of distributions.…”