Paying £1 or nothing in Dictator Games: No differences

We conducted a Dictator Game experiment in Prolific in which 1,195 participants were endowed with £1 (63% female, average age 32.2). The purpose of the experiment was to examine whether paying participants real money would lead to different outcomes compared to not paying them anything (hypothetical condition) in the dictator game. Participants were randomly assigned to one of three groups: those who were guaranteed to be paid for real, those with a 1/10 probability of being paid (BRIS), and those who were not paid at all (self-reported or hypothetical condition). Our findings indicate that paying participants real money had no significant impact on the results. There were no noticeable differences in the mean amounts given by participants, the proportion of participants giving zero, or the frequency of equal splits. However, the fraction of hyper-altruistic drops significantly in the hypothetical setting (from 6% to 2% of the sample). Therefore, our experiment provides evidence that self-reported measures of altruism in the dictator game are indeed informative.