Working Paper: CEPR ID: DP14183
Authors: Benny Moldovanu; Andreas Kleiner
Abstract: We study how parliaments and other committees vote to select one out of several alternatives in situations where not all available optionscan be ordered along a \left-right" axis. Practically all democratic parliaments routinely use Sequential Binary Voting Procedures in or-der to select one of several alternatives. Which agendas are used in practice, and how should they be designed ? We assume that pref-erences are single-peaked on an arbitrary tree and we study convex agendas where, at each stage in the sequential, binary voting process,the tree of remaining alternatives is divided in two subtrees that are subjected to a binary Yes-No vote. In this wide class of situations weshow that dynamic, strategic voting is congruent with sincere, unsophisticated voting even if agents are privately informed, and no matterwhat their beliefs about other voters are. We conclude the paper by illustrating the empirical implications of our results for two large casestudies from Germany and from the UK.
Keywords: Voting; Agenda; Revealed Preference
JEL Codes: D02; D72; D82
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| sincere myopic voting (D72) | election of the Condorcet winner (D79) |
| structure of the voting agenda (D72) | outcome of the voting process (D72) |
| preferences being single-peaked on an arbitrary tree (C69) | identification of the Condorcet winner (D79) |
| voting behavior observed in empirical cases from Germany and the UK (D72) | alignment with theoretical predictions (C52) |
| designed voting agendas (D72) | final election outcomes (K16) |