This study aims to prove how kindness conducted by the group, the definition of violations and the value of doing good, can affect moral credential in the form of the leniency of punishment for corruptors. Sample 319 polices and doctors divided into two segments. This study uses a population-based survey experiment method. Participants divided into moral credential with the narrative noble duty, and non-moral credential with neutral narratives. Researchers require each participant to provide criminal punishment and social reactions to the vignette of violation cases due to receiving gratification without (study 1) and with group support (study 2). Statistical correlation test with Kendall's Tau-b finds that defining the behaviour of gratification influences punishment, the correlation shown in invisible victim is rτ = .455, p = .000, and visible victim, rτ = .191, p = .003, but there is no significant relationship between the values of selflessness action and defining the behaviour of gratification as a violation (rτ = -.006, p = .922). Through the Mann Whitney U test it was found that the moral credential and non-moral credential conditions do not affect the provision of informal social reactions to violators, both the perpetrators of gratification with invisible victims (Mdn = 2), U = 2757, p = .523, and visible victims (Mdn = 2), U = 2778.5, p = .575, and also in a situation of gratification with group support, the perpetrators of gratification with invisible victims (Mdn = 2), U = 3419, p = .937, and visible victims (Mdn = 2), U = 2978,5, p = .119.