Estimating Heterogeneity in Intuitive Prosociality

This paper uses a randomized experiment on a representative sample from the Norwegian population (N = 1390) to test whether prosocial behavior is intuitive. We use time pressure to capture intuitive decision making and a dictator game to measure prosocial behavior. We also estimate and test for “population heterogeneity” in the effect size. First, we exogenously vary subject experience with economic games to test whether experience influences the treatment effect. Second, we leverage our data structure to conduct a "pseudo meta-analysis," assessing effect size heterogeneity across several potential sources of heterogeneity. This helps us assess whether prior findings in the literature stem from false positives or genuine heterogeneity in effect sizes. Our results show no evidence that experience with economic experiments moderates the time pressure effect. Further, the estimated heterogeneity in effect sizes in the pseudo meta-analysis is negligible, we cannot reject the null hypothesis of a uniform p-value distribution, and our rejection rate aligns with the expected 5% from chance alone. These findings support the view that the apparent link between prosocial behavior and intuitive processing is likely due to false positives rather than heterogeneous treatment effects.