One key finding of behavioral economics is that many people are motivated by social concerns. However, most of the robust mechanism design literature focuses on beliefs, and takes selfish preferences for granted. We study two classic challenges in mechanism design - bilateral trade à la Myerson and Satterthwaite (1983) and redistributive income taxation à la Mirrlees (1971) and Piketty (1993) - to show that some standard mechanism design solutions systematically fail with social preferences, while others are robust. We thus introduce the notion of a social-preference-robust mechanism which works not only for selfish but also for social preferences of different nature and intensity, and characterize the optimal mechanism in this class. We compare the performance of the optimal mechanisms for selfish agents and the optimal social-preference-robust mechanisms with the help of a series of laboratory experiments and find that behavior can indeed be better controlled with social-preference-robust mechanisms.

VIEW PAPER