Untangling altruism and parochialism in human intergroup conflict

The scale of violent intergroup conflict in humans is astonishingly large compared to other mammals. This capacity for war is closely linked to our exceptionally cooperative abilities. The parochial altruism model formally describes how within-group cooperation and between-group competition could be dynamically intertwined. However, whether this influential model correctly captures the fast-paced processes of preference adaptation in humans has not been systematically scrutinized yet. Here, we develop the psychometric toolkit required for this task and test key assumptions and predictions of the model in groups involved in real intergroup conflicts of varying intensities (total N = 1,121). Conceptually corroborating the model, we find that our new measures which cleanly separate interindividual altruism from intergroup parochialism characterize individuals’ preferences better than previous metrics and improve behavioral predictions of contributions to conflict. However, our results also show that parochialism varies for different outgroups, a finding that is not anticipated by the model. Thus, the five studies we report here provide new methods for studying individual- and group-level social preferences in the context of intergroup conflict and present new evidence that can inform substantive theoretical improvement.