Moral universalism: Incentivised evidence

• Uses incentivised dictator and trust games to measure moral universalism • Finds wide individual variation and moderate stability across contexts • Stability across identity dimensions is weaker than prior survey studies • Relationships with altruism, race and political views differ from prior research There has been a recent surge in interest among economists in moral universalism, the extent to which individuals treat in-groups and out-groups equally in altruism and trust. I provide novel incentivised evidence, adding to the literature which has relied on survey-based measurements of universalism. Using data from dictator and trust games involving 17 identity groups across five dimensions, I find broad consistency with existing studies: universalism varies widely across individuals and is moderately stable across contexts. Relative to previous research, however, I observe weaker stability in universalism across identity dimensions and some differences in its predictor variables.