The Persistence of Trust: Further Tests of the Cooperators'Advantage

在许多共享资源的问题中，通常被称为“公共产品”，或者一个人的决策结果高度依赖于其他人的行为的问题中，出现了关于信任和合作或缺陷行为的本质的重要研究问题。在这种情况下，合作对群体有利，但个人私下里不合作会更好。来自游戏的实验证据反映了高度相互依赖的结果和选择性参与的条件，表明：(1)意图合作的人比叛逃者更愿意进入这种情况（即“信任”）；(2)在实验博弈中，合作者比叛逃者更容易获得奖励，这就是所谓的“合作者优势”；(3)有意合作者和有意叛逃者都不具备识别陌生人意图的特殊能力；这意味着合作者的优势完全来自于与其他合作者不成比例的参与概率。这表明玩家决定通过启发式方式参与其中，他们将自己的意图投射给其他人。本研究将进一步检验这些结果的稳健性，通过调查经验对启发式的影响，从而产生合作者的优势。相互竞争的假设预测：(1)从人口池中反复“抽样”合作伙伴将导致期望趋同，从而消除合作伙伴的优势；或者(2)他人行为的现有样本是有偏见的，并为有意的合作者和有意的叛逃者提供了证实他们最初期望的证据，因此，很少刺激经验调整。这些假设将通过向受试者提供关于他人选择的各种反馈来验证。这些发现将会影响到合作发展的可能性，而不依赖于文献中讨论的以牙还牙或其他基于激励的机制。

In many problems of shared resources, often termed "public goods," or problems where the decision outcome of one person depends highly on the behavior of others, important research questions arise about the nature of trust and cooperative or defective behavior. In such situations, cooperation works to the advantage of the group but individuals are privately better off not cooperating. Experimental evidence from games that reflect the conditions of highly interdependent outcomes and optional participation, shows that (1) people who intend cooperation are more willing than defectors to enter into such situations (i.e. to "trust"); (2) cooperators are relatively more successful regarding rewards from the experimental games than the defectors, hence the term, "cooperators' advantage;" and (3) neither intending cooperators nor intending defectors have special capacity to recognize the intentions of strangers, meaning that the cooperators' advantage comes exclusively from the disproportionate probability of participation with other cooperators. This suggests that players decide to participate by a heuristic where they project their own intentions to others. This research will further test the robustness of these results by investigating the effects of experience on the heuristic that gives rise to the cooperators' advantage. Competing hypotheses predict: (1) that repeated "sampling" of partners from the population pool will lead expectations to converge such that the cooperators' advantage is eliminated; or (2) that available samples of others' behavior are biased and provide both intending cooperators and intending defectors with confirming evidence of their original expectations and thus, stimulate little adjustment from experience. These hypotheses will be tested by providing subjects with various types of feedback about the choices of others. The findings will bear on the possibility of cooperation evolving independently of tit-for-tat or other incentive-based mechanisms discussed in the literature.