The influence of group size on distributional fairness under voting by veto

Voting by Veto [VBV], developed by Dennis C. Mueller, is a two-stage voting mechanism for committee decision making. In the first stage, every member of the committee makes a proposal. The proposals are joined with the current status quo (or another fallback option). For the second stage, a random mechanism determines the order of voting and the committee members sequentially eliminate one alternative each. After every committee member has exercised her veto, a single winning alternative remains. Game theory predicts that the equality of the resulting distribution under VBV varies with group size. Hence, decisions in two-member groups should result in highly unequal distributions of benefits. With increasing group size, however, VBV should generate more equally distributed outcomes. We examine the effect of group size under VBV in a series of laboratory experiments where groups with a varying number of members play a ‘divide the dollar game’. Our results confirm that larger groups choose more equal distributions than smaller groups. However, we also find deviations from the model's predictions. Some committees in our experiment do not select Pareto-optimal alternatives. An exploratory analysis of possible explanations suggests that individual behavior is most likely motivated by a combination of material self-interest and social preferences such as reciprocity and fairness.