When there are two groups of officials in a public organization, we show that depending on the groups’ behavior – collusive or competitive – increasing the level of monitoring and punishment may have different impacts on corruption. If the two groups of public officials had been demonstrating collusive behavior, increased monitoring or punishment reduces both the level of corrupt activities and the corrupt officials’ bribe revenues. However, if the groups had not been colluding, increased monitoring reduces the level of corruption, but increases the corruption revenues collected. Only after reaching the optimum level of monitoring, is this result reversed.