Understanding the Decline in Cooperation: The Role of Confusion and Social Preferences in Public Goods Games, Investigated Through Manipulation of Other Players' Behavior and Decision-Making Model Comparisons

In repeated public goods games, a robustly observed phenomenon is the gradual decline in cooperation that initially appears. Two main explanations for this decline have been proposed: the Social Preferences Hypothesis (SPH), which posits that individuals adjust their behavior based on altruism or equality toward others, and the Confusion Learning Hypothesis (CLH), which suggests that participants are simply trying to maximize their own benefit and learn the optimal behavior over time. This study aims to investigate which hypothesis is more valid through two approaches: (1) deriving a unified mathematical decision-making model and fitting the data and explains both hypotheses, and (2) conducting experimental manipulations by adjusting the feedback participants receive about others' behavior. The results of Study 1 weakly supported CLH, while Study 2 provided strong evidence in favor of SPH. This study discusses the dual approach of combining decision-making models and experimental manipulation to explore both the decline and maintenance of cooperation in public goods games.