Collective Action Problem in Heterogeneous Groups with Punishment and Foresight

The collective action problem can easily undermine cooperation in groups. Recent work has shown that within-group heterogeneity can under some conditions promote voluntary provisioning of collective goods. Here we generalize this work for the case when individuals can not only contribute to the production of collective goods, but also punish free-riders. To do this, we extend the standard theory by allowing individuals to have limited foresight so they can anticipate actions of their group-mates. For humans, this is a realistic assumption because we possess a "theory of mind". We use agent-based simulations to study collective actions that aim to overcome challenges from nature or win competition with neighboring groups. We contrast the dynamics of collective action in egalitarian and hierarchical groups. We show that foresight allows groups to overcome both the first- and second-order free-rider problems. While foresight increases cooperation, it does not necessarily result in higher payoffs. We show that while between-group conflicts promotes within-group cooperation, the effects of cultural group selection on cooperation are relatively small. Our models predict the emergence of a division of labor in which more powerful individuals specialize in punishment while less powerful individuals mostly contribute to the production of collective goods.