Humans work together in groups to tackle shared problems and contribute to local club goods that benefit other group members. Whereas benefits from club goods remain group bound, groups are often nested in overarching collectives that face shared problems like pandemics or climate change.
Such challenges require individuals to cooperate across group boundaries, raising the question how cooperation can transcend beyond confined groups.

Agents belong to different groups with their own club good (CG)
and a group-independent, nonexcludable public good (PG)
(A). Costly contributions to the club good (“parochial cooperation”) create a benefit bCG for the group that the agent belongs to. Costly contributions to the public good (“universal cooperation”) create a benefit bPG for all agents regardless of their group membership. In a second stage, agents interact in a dyadic exchange
(B). In the role of the donor, an agent can pay a cost ch to create a benefit bh for another agent (their assigned receiver).
Depending on the fluidity of group boundaries, agents may meet other agents for dyadic interactions only from their own group (p = 1) or may meet in- and out-group members with equal likelihood (p = 0.5).
Here, we show how frequent intergroup interactions allow groups to transition from group-bound to universal cooperation. With frequent intergroup interactions, reciprocity of cooperative acts permeates group boundaries and enables the evolution of universal cooperation.
As soon as intergroup interactions take place frequently, people start to selectively reward cooperation aimed at benefitting everyone, irrespective of their group membership.
Simulations further show that it becomes more difficult to overcome group-bound cooperation when populations are fragmented into many small groups. Our findings reveal important prerequisites for the evolution of universal cooperation.
Previous research has shown how individuals can fail to solve shared collective action problems when they have the ability to tackle such problems on their own, even if such “self-reliance” is less efficient. Only when individual members become more interdependent and, concomitantly, less able to solve shared problems on their own will group cooperation emerge. Here, we extend these findings to the intergroup context.
With lower interdependence across groups (operationalized as a lack of intergroup interactions), group-bound cooperation is favored, while more efficient universal cooperation suffers. Similar to the transition from self-reliance to group cooperation, higher interdependence across groups (operationalized as more frequent interactions across group boundaries) promotes a transition from localized to more efficient global cooperation.
Especially with global challenges like climate change or pandemics that require cooperation across group boundaries, more fluid group boundaries and increased exchange across group borders can help to establish public goods from “bottom-up” without requiring a central authority or complex institutions to curtail free-riding. Globalization, characterized by increased trade relations and mobility of individuals and the integration of nation-states into larger unions, may therefore help to create global public goods, supported by reciprocal relationships between members of different groups.