On the fairness of random procedures

We consider procedures that use randomness to make a decision that involves several individuals. We asked subjects to compare the fairness of six pairs of seemingly equivalent procedures. We propose a classification of subjects into two categories: those who are “emotional” in the sense that they exhibit a systematic ranking of procedures on the basis of intuitive notions of fairness, and those who are “consequentialists” in the sense that they systematically view each pair of procedures as being equally fair. According to this classification, about 30% of the subjects are emotional while 31% are consequentialists.