People Intentionally Avoid End-Game Information to Facilitate Cooperation in Repeated Social Dilemma
Speaker: Associate Research Fellow Zou Wenbo
Time: 2:30 p.m., June 5th, 2025 (Thursday)
Place: Conference Room on the 8th Floor, Office Building, School of Economics
Speaker Information:
Zou Wenbo is an Associate Research Fellow at the School of Economics, Nankai University. Her research interests lie in behavioral and experimental economics, as well as development economics. She has led several research projects, including the National Natural Science Fund Youth Project and a research grant from the Asian Studies Center at Nankai University. Her work has been published in leading domestic and international journals such as Experimental Economics, Journal of Economic Behavior and Organization, and World Development.
More Information:
All long-term relationships ultimately face the possibility of dissolution. When information about the game horizon can be avoided, individuals who understand the role of dynamic incentives in supporting cooperation may choose to forgo such information, aiming to prevent or mitigate the unraveling of cooperation predicted by backward induction. This paper examines endogenous horizon information choices in repeated Prisoner's Dilemma (PD) games, both theoretically and experimentally. We identify subgame-perfect Nash equilibria (SPNE) in which either one or both players choose not to acquire horizon information. In both cases, cooperation can emerge if the continuation probability of the repeated game is sufficiently large. Under unilateral information avoidance, the informed player can exploit the information asymmetry to earn a higher payoff. In a laboratory experiment, we found that only a small percentage of participants chose to avoid horizon information. However, for one of the two horizon sequences that we used, the treatment in which horizon information was endogenously chosen at no cost yielded significantly higher levels of cooperation than the treatment with exogenously provided information. We also found modest evidence supporting the payoff advantage of the informed player in the unilateral information avoidance scenario.