The Effect of Individual and Group Punishment on Individual and Group-Based Dishonesty

Economic dishonesty is a widespread behavior that has substantial implications for organizations and societies. Recent studies suggest that decision making in groups or commitment to other individuals can further increase such dishonesty in contrast to individual decision making. Various interventions have been suggested to curb dishonesty, with classical economic approaches emphasizing the use of punishment by focusing among others on the risk (“how likely am I to get caught?”). However, the effects of punishments on dishonesty are mixed and it is currently unclear if punishment can be effective in collaborative settings. We conducted an experimental online study (N = 717) collecting UK-based Prolific participants playing a tax evasion game for five rounds. We manipulated (1) the type of payoff (whether dishonesty earned money for the individual or for the group of three), (2) the risk of punishment (no risk or 30% audit chance), and (3) the type of punishment (whether the individual was punished or the whole group in case of getting caught). Against our prediction, we did not observe higher dishonesty for group-based payoffs compared to individual-based payoffs for treatments without punishment but higher compliance. As predicted, we found that punishment risk decreased dishonesty and increased compliance by 13.6 percentage points (Cohen’s d = 0.31). Further, in contrast to the predictions based on utility-based models, we found that group punishment was more effective in decreasing dishonesty and increased compliance by 5.9 percentage points (d = 0.11) compared to individual punishment. This effect was similar for individual and group-based payoffs. Overall, this study provides evidence for the effectiveness of punishment in curbing dishonesty across different decision-making structures.