Polarization and Cooperation: A Behavioral Experiment (with Ignacio Jurado and Albert Falcó)

This paper explores the effects of affective polarization on cooperative behavior through a behavioral experiment conducted in Brazil and Spain. Participants were asked to perform a simple, one-shot asynchronous task: converting to capital letters either a neutral salad recipe or a politically charged text written by another individual. The experiment varied also both the political affiliation of the original author—presented as a supporter of the participant’s most liked or disliked political party—and the type of reward offered for task completion. The results show that cooperation significantly declines when the task carries political meaning compared to a neutral one. Furthermore, cooperation significantly decreases when the partner is identified as an out-partisan. While participants are more inclined to cooperate—and perform more effectively—when paired with a co-partisan, this positive effect is weaker than the negative impact of interacting with an out-partisan. Both private and collective rewards can enhance cooperation, but these incentives are insufficient to counteract the divisive effects of partisanship. Participants are even willing to forgo personal rewards if doing so prevents an out-partisan from benefiting. Moreover, although political tasks tend to suppress overall cooperation, they can enhance collaboration among co-partisans, especially for individuals that are more affective polarized, which in our experiment reinforces more in-group favoritism. These findings suggest that partisan animosity can undermine even the most basic forms of cooperation, including tasks that are simple and apolitical in nature. The implications are significant: when everyday collaborative interactions are disrupted by political divisions, the potential to foster broader, cross-cutting cooperation in politically diverse societies is severely compromised.