Previous research on social preferences has found that reciprocal behavior is an automatic response, which requires less cognitive control than self-interested behavior. However, research on unethicality has demonstrated that cognitive control is required to resist the temptation to benefit from engaging in an unethical act, thus suggesting that self-interested behaviors are automatic. By manipulating information asymmetry (i.e., advantage) among trust game receivers, we examined whether the automatic tendency to positively reciprocate a kind gesture is driven by a genuine motivation to be fair, or whether it reflects a motivation to appear fair. In an equal-information condition, we replicated previous findings by showing that limited cognitive control, due to ego-depletion, promotes reciprocal behavior (Experiments 1 & 2). However, this effect did not replicate under time-constraint, an additional manipulation to limit cognitive control (Experiment 3). Importantly, in an advantaged-information condition, under ego-depletion (Experiments 1 & 2) or time-constraint (Experiments 3 & 4) manipulation, participants did not exhibit enhanced reciprocity. Rather, in three out of these four experiments, and as confirmed by a meta-analysis that additionally included a pilot experiment, participants exploited their information advantage and positively reciprocated to a lesser extent than participants whose cognitive control was intact. These results suggest that the automatic preference for reciprocity might be more driven by a self-serving motivation to appear (rather than to be) fair than is typically credited. We further discuss various other findings supporting the notion that what may appear as automatic pro-sociality may in fact reflect an automatic self-serving motivation of self-presentation.

先前对社会偏好的研究发现，互惠行为是一种自动反应，与自利行为相比，它需要更少的认知控制。然而，对不道德行为的研究表明，需要认知控制来抵制从从事不道德行为中获益的诱惑，从而表明自利行为是自动的。通过操纵信任博弈接收者之间的信息不对称（即优势），我们检查了积极回报善意姿态的自动倾向是否是由真正的公平动机驱动的，或者它是否反映了出现的动机公平的。在信息均等的条件下，我们通过显示有限的认知控制，由于自我耗竭，促进了互惠行为（实验 1 和 2），重复了之前的发现。然而，这种效果在时间限制下不会复制，这是一种限制认知控制的额外操作（实验 3）。重要的是，在信息优势条件下，在自我消耗（实验 1 和 2）或时间限制（实验 3 和 4）操纵下，参与者没有表现出增强的互惠性。相反，在这四项实验中的三项中，并且正如一项元分析所证实的那样，该荟萃分析还包括一项试点实验，参与者利用他们的信息优势，积极回报的程度低于认知控制完好的参与者。这些结果表明，对互惠的自动偏好可能更多地是由自私的动机驱动的，这种动机比通常认为的要公平（而不是公平）。我们进一步讨论了支持以下观点的各种其他发现，即看似自动的亲社会性实际上可能反映了自我展示的自动自利动机。