Dictator giving and taking: Evidence from Africa

Replication lies at the heart of economic analysis and is of crucial importance for science progression. Previous studies find evidence that framing significantly matters in dictator games. These findings mostly rely on societies considered as being WEIRD. We explore the generalizability of these results by collecting data from 165 dictators in Botswana. Our findings show that average giving is significantly reduced when the action set includes taking, but by substantially less than what has been found previously in a WEIRD society. We calculate the post-study probability estimates showing how our priors change with the results reported here. Our study further highlights the importance of institutions and has significant implications for future research.