(joint work with Charles Bellemare)

We propose an experimental design to measure the impact of stakes on the distribution of guilt sensitivity in a population. The design exploits exogenous variations in beliefs to identify guilt sensitivity at the individual level. The design rules out reciprocity motives and can separate behavior consistent with various levels of guilt sensitivity from behavior consistent with distributional concerns. We estimate a structural model to recover the distribution of guilt sensitivities for stake-sensitive and stake-insensitive players. Our estimates suggest that slightly more than half of subjects are stake-insensitive and have a relatively small heterogeneity of guilt sensitivities. We find significantly more heterogeneous guilt sensitivities at all stake levels for stake-sensitive players. For the latter group, we find that the average level of guilt sensitivity significantly decreases with the stakes. Finally, we find a positive and significant correlation between our economic measure of guilt sensitivity and a measure of guilt sensitivity often used by psychologists.