Abstract
During human evolution, individuals interacted mostly within small groups that were connected by limited migration and sometimes by conflicts. Which preferences, if any, will prevail in such scenarios? Building on population biology models of spatially structured populations, and assuming individuals' preferences to be their private information, we characterize those preferences that, once established, cannot be displaced by alternative preferences. We represent such uninvadable preferences in terms of fitness and in terms of material payoffs. At the fitness level, individuals can be regarded to act as if driven by a mix of self-interest and a Kantian motive that evaluates own behavior in the light of the consequences for own fitness if others adopted this behavior. This Kantian motive is borne out from (genetic or cultural) kin selection. At the material-payoff level, individuals act as if driven in part by self-interest and a Kantian motive (in terms of material payoffs), but also in part by other-regarding preferences towards other group members. This latter motive is borne out of group resource constraints and the risk of conflict with other groups. We show how group size, the migration rate, the risk of group conflicts, and cultural loyalty shape the relative strengths of these motives.