Do Real Stakes Always Matter? Experimental Evidence on Social and Anti-Social Preferences

In this paper, we investigate experimentally whether social and anti-social preferences are influenced by the nature of the stakes, i.e. whether the stakes are real or hypothetical. Our experiment consists of a control treatment in which incentives are real and a experimental treatment where they are hypothetical. Participants were randomly chosen among a broad and representative subject pool of adult professionals enrolled in continuing education, and played three games: i) a Modified Dictator Game, ii) a Giving Game, and iii) a MoneyBurning Game. We find significant quantitative differences across treatments in most individual decisions. However, our data also indicate that while the effect sizes differ by treatment, players facing real and hypothetical stakes generally respond in the same qualitative way to changes in either the initial payoff distribution or the cost and effectiveness of giving or money-burning decisions