Voting by Veto

• Applicant: Professor Dr. Jan Sauermann
• Subject Area: Economic Policy, Applied Economics ; Political Science
• Term: from 2016 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 310624604

Voting by Veto [VbV] developed by Dennis C. Mueller is a voting mechanism suitable for decision making among small and medium-sized groups. In brief, VbV works as follows: In an n-member group, a VbV-decision consists of two stages. First, every group member makes a proposal. Those proposals, together with the current status quo (or another fallback option), form the set of available options so that there are n+1 alternatives to choose from. Then, a random mechanism determines the order of voting in the second stage. Here, group members sequentially eliminate one alternative each. After every group member has cast her veto, a single winning alternative remains. VbV has several theoretical properties that might make it a viable collective decision making mechanism under especially tough conditions such as high social and economic heterogeneity among group members. VbV is decisive, defining a unique winning proposal for any set of initial proposals and a given sequence of voting. It is neutral as no alternative (including the status quo) has an advantage in the voting stage. Most importantly, VbV is procedurally and distributionally fair because the winning proposal cannot be a proposal ranked last by any voter, and voters have an incentive to make proposals leading to an equal distribution of payoffs. Equal proposals rank relatively high in the preference functions of all voters. Hence, voters whose own proposals have already been eliminated during the decision making process, work for the victory of an equal distribution of payoffs. The probability that VbV produces an equal distribution of payoffs increases with group size. However, the fairness properties of VbV can be foiled by collusive behavior of subjects. Theoretically, the mechanism is prone to falling prey to the tyranny of the majority if some group members form coalitions in order to exploit other members of their group.The theoretical fairness properties of VbV are highly desirable in contexts where groups face intense conflicts. Thus, VbV might offer an innovative institutional resolution for moderating conflicts by fostering consensual decision making. However, there is hardly any empirical evidence on the workings of the decision making mechanism. To the best of my knowledge, it has never been applied in any real-world decision making. This project aims at filling this gap. When institutions such as VbV do not exist in the observational world, laboratory experiments are an ideal method for generating empirical evidence. Hence, this project will use experimental methods to seek answers to three questions:1) How does group size influence decision making under Voting by Veto?2) Is the consensual decision making under Voting by Veto robust against collusion among group members?3) How do social preferences such as reciprocity and inequality aversion influence decision making under Voting by Veto?

• DFG Programme: Research Grants