Is Ignorance Bliss?
Is Ignorance Bliss?
复制标题
无知是福吗？
DOI：
10.1016/j.jebo.2019.02.025
复制
发表时间：
2019
影响因子：
2.2
通讯作者：
Yang Yang
中科院分区：
文献类型：
--
作者：
Yang Yang
本文提出了一个两回合的囚徒困境游戏，回合之间的重新匹配涉及两种参与者类型，“给予者”和“接受者”。前者比后者具有更多的合作偏好。两个信息条件进行了比较：在一个，球员类型的行为之前，选择。在另一种情况下，类型仍然是私有信息。在这两种情况下，第一轮的决定都将在第二轮向（新）合作伙伴披露。当给予者的比例足够高时，对该游戏的完美贝叶斯均衡（PBE）分析预测，如果类型信息未被揭示，给予者基于对手的先前行为进行区分，并且当类型信息被揭示时，基于对手的类型进行区分，而不管对手先前的行为。这意味着类型信息的披露减少了合作。我们使用实验室实验来研究这个游戏中的行为。我们观察到，给予者的歧视，在这两种治疗的理论预测。然而，与PBE预测相反，我们发现当类型被揭示时，玩家更经常选择合作。我们发现，基于图像评分的另一种理论方法可以解释60%-70%的数据。
This paper presents a two-round Prisoner’s Dilemma game with rematching between rounds involving two player types, ‘Giver’ and ‘Taker’. The former have more cooperative preferences than the latter. Two information conditions are compared: in one, player types are revealed before actions are chosen. In the other, types remain private information. In both cases first-round decisions are revealed to (new) partners in round 2. When the proportion of Givers is sufficiently high, a Perfect Bayesian Equilibrium (PBE) analysis for this game predicts that Givers discriminate based on opponents’ previous actions if type information is not revealed and, when type information is revealed, discriminate based on opponents’ type regardless of their previous actions. This implies the revelation of type information decreases cooperation. We study behavior in this game using a laboratory experiment. We observe that Givers discrimination as predicted by theory in both treatments. However, in contrast to the PBE prediction, we find players choose to cooperate more often when types are revealed. We show that an alternative theoretical approach based on image scoring can explain 60%–70% of our data.