道德许可是指当人们想起自己曾经做过比较道德的行为，或者是符合社会规范的行为后，更可能放任自己做出不良行为。同样，也存在观察者道德许可效应，即当观察者发现目标人物之前有过良好的道德行为纪录是，会更容忍其后的不道德行为。 本文从观察者视角考察道德许可对腐败容忍度的影响。研究假设，当腐败者之前有过良好政绩时，人们对其后的腐败行为会更加宽容。研究1和研究2采用实验法分别考察观察者对个体官员和单个政府组织腐败行为的道德许可效应。结果发现，提示腐败者之前的良好政绩后，被试对其腐败行为容忍度更高，这表明存在对腐败行为的道德许可效应。另外，对腐败者的道德评价和腐败严重度感知发挥了中介作用。研究3、研究4、研究5考察对总体政府层面腐败现象的道德许可效应。研究3采用相关法，发现对政府的政绩评价与腐败容忍度正相关。同时研究3还发现官方媒体使用频率可以正向预测腐败容忍度，其中政府功绩评价和腐败严重度感知发挥中介作用。这意味着官方媒体宣传政府政绩可能会引发人们对腐败的道德许可。 研究4和研究5考察群体身份对腐败道德许可的调节作用，将党和政府中的腐败现象作为研究问题，采用党员和非党员两类被试。因为党员被试与执政党中的腐败者都属于党员群体，故而将党员被试视为内群体成员，将非党员被试视为外群体成员。内群体偏好理论预期，党员被试对腐败的道德许可效应更强。研究4发现，群体身份调节历史功绩评价与腐败容忍度之间的关系，但调节模式却与研究假设相反：对非党员被试，功绩评价可以正向预测腐败容忍度，表明存在道德许可效应；而对于党员被试，历史功绩评价并不能预测腐败容忍度，表明不存在道德许可效应。研究5采用实验法进一步验证了研究4的结果，提示党的历史功绩后，非党员被试对腐败容忍度显著提升（相比于无提示条件），再次出现道德许可效应；但对党员被试在两种情况下无显著差异，即依然没有出现道德许可效应。 考虑到研究4和研究5中党员被试不在党政机关任职，对党政机关中的腐败行为涉入度较低，因而将他们视为腐败低涉入度内群体成员。为进一步检验腐败高涉入度内群体成员对腐败的道德许可模式，研究6采用与研究5类似的方法，以某高校学生会的腐败故事为情境，考察高涉入度的内群体成员（该校的校学生成员）、低涉入度的内群体成员（该校各学院学生会中的成员）、外群体成员（该校的非学生会成员）各自的道德许可模式。结果发现，提示学生会的良好行为纪录后，三类被试的反应存在显著差别，其中外群体成员与高涉入度内群体成员都对腐败行为表现出道德许可。但低涉入度的内群体成员的腐败容忍度却没有提升，即不存在道德许可效应，这种结果可以用害群之马效应来解释。此外，对于内群体成员，群体认同也会调节道德许可的程度，对于腐败高涉入度的内群体成员，群体认同越强，道德许可程度越高；而对于腐败低涉入度的内群体成员，群体认同越低，道德许可程度越高。 本研究结果表明，腐败者之前的政绩会提升人们对其腐败行为的容忍度，即存在对腐败的道德许可效应，其中道德评价和腐败严重度感知发挥中介作用；对个体腐败层面和群体腐败层面都存在道德许可效应；外群体成员、腐败高涉入度内群体成员，都对内群体腐败授予道德许可，但腐败低涉入度的内群体成员却不会。最后，群体认同高且腐败涉入度低的内群体成员，还会产生反道德许可效应，这可以运用害群之马效应来解释。

Moral licensing effect refers to the phenomenon that when thinking of the moral behavior performed before, individuals are inclined to indulge themselves into morally questionable conducts. Similarly, when seeing someone performed well before, observers would be inclined to endure his/her immoral behavior. This paper focused on whether observers would tolerate corruption more when they got known of the corruptor’s good record in his official career. Study 1 and Study 2 respectively manipulated the corruptors’ records on individual or institutional levels, and found that participants tolerate corruption more when corruptors once kept good record compared with no good record. That’s to say, observers licensed corruptions of individuals who performed well before. A mediatoring role of moral evaluation and perceptions of corruption’s severity was found in moral licensing effect. Study 3, Study 4 and Study 5 inspected the moral licensing effect on corruption of the whole government and the governing party . It was found that the evaluations on government’s contributions were positively correlated with corruption tolerance, and the frequency of official media contact could positively predict corruption tolerance via moral licensing. Study 4 replicated the research of Study 5 with group membership as a potential moderating variable. In Consistent with the result of Study 4, as for non-party members evaluations on the party’s contributions could positively predict corruption tolerance. However, no such an effect of found as for party members. In order to further check the moderating role of party membership, the contributions of party and government were manipulated. Findings showed when primed with high contributions of party and government, non-party members abhorred corruption less. Again, no such an effect was found as for party members. Study 6 carried on with the issue of Study 5, with group membership classified more specifically as high corruption-involved in-group members, low corruption-involved in-group members, and out-group members. Again, out-group members showed moral licensing effect on corruption. And it was the same with high corruption-involved in-group members. But low corruption-involved in-group members display no moral license on corruption, and this result could be interpreted in the framework of black sheep effect. As a conclusion, the good record of corruptors could improve the tolerance towards corruption. And moral evaluations and severity of corruption played mediators in the process of moral licensing effect on corruption. Group membership moderated moral licensing effect: out-group members and in-group members high involved in corruption showed moral licensing effect on corruption, in-group members low involved in corruption displayed no moral licensing effect. Furthermore, in-group members who were low involved in corruption and high identified with in-group showed the inverse of moral licensing of corruption, which could be interpreted as black sheep effect.