An Experimental Analysis of Third-Party Response to Corruption

It is a well-established fact that corruption is a widespread phenomenon. An important aspect of corruption is that two parties act jointly in order to further their own interests at the expense of a third party. The response of the third party has a significant impact on the persistence of corruption. This paper contributes to the experimental literature on corruption by analyzing third-party response to corruption. We report results from a laboratory experiment in which subjects participate in a one-shot, anonymous-pairing, three-person game, where one subject is a firm, the second is an official, and the third is a citizen. Contrary to the theoretical predictions of the model, we find that more than 50% of citizens choose to punish. The reasons given for choosing to punish are negative reciprocity, a sense of fairness, and a desire to correct wrong behavior. Increasing the amount of the maximum punishment that the citizens can impose on the firms and the officials allows us to observe that the number of firms who choose to offer a bribe decreases and the number of citizens who choose to punish increases. This is despite the fact that imposing a higher punishment amount imposes a higher cost on the citizen. Among the citizens, our results reveal that women in general are more willing to punish corrupt behavior and are less likely to accept the bribes offered to them. These findings are in agreement with the argument that since women are often the victims of corruption, they may be more critical of corrupt actions and may, therefore, be more willing to punish. The results also reveal that citizens in developing countries are more accepting of corrupt behavior. Given the high levels of corruption that exist in developing countries, this result provides experimental support for the theory that the frequency of corruption may be related to a society’s willingness to accept corrupt behavior, which, in turn, may help in the sustenance of corruption