On the Development of Cooperative and Antagonistic Relationships in Public Good Environments. A Model-Based Experimental Study

The importance of prosocial behavior is currently widely acknowledged. This not only holds for the social sciences, including economics, but also the life sciences where this kind of behavior is observed across the evolutionary ladder. Evolutionary continuity consequently suggests that caring for others may be due to both strategic motivations, based on deliberation and reasoning, and more impulsive (affective) non-strategic motivations, in line with the two major mental systems distinguished by Kahneman (2011). This study estimates and applies a dynamic affective-ties model, allowing for strategic behavior, on a novel experimental data set to investigate four underexplored issues regarding public good environments. First, do negative (antagonistic) relationships as well as positive (cooperative) relationships develop in a public good game environment once equal space is given for their development? Second, do people react differently to the positive versus negative behavior of others in such an environment? Third, does the affective-ties model outperform other relevant models in a proper out-of-sample prediction horse race regarding the same game? Fourth, is this model helpful in explaining behavior across different contexts? Our results provide a clear yes to each of these four questions. Negative relationships do develop, but seem less stable than positive relationships in the long run; people appear to react more strongly to the positive compared to the negative behavior of others; our estimated (two parameter) model outperforms other models; and our model helps explain why and how people switch behavioral rules (like tit-for-tat) as the parameters of a repeated prisoner’s dilemma game change.