The authors build on the work of Engelmann and Strobel and of Ackert, Martinez-Vazquez, and Rider to examine the potential role of social preferences in tax policy design. They randomly assign each participant in a session to a group with five members. The payoffs to participants are determined by majority vote of the participants in a given group. The authors randomly assign each participant a pretax income that remains the same throughout the experiment to control for the potential confounding effect of risk-pooling equilibria. On the basis of statistical analysis of the data generated by these experiments, they find support for the hypothesis that the minimax preferences of Rawls or the inequality aversion model of Fehr and Schmidt help explain individual choices over alternative tax structures.