Corruption is a phenomenon that involves moral components. Despite the general agreement about the involvement of a moral transgression in the characterization of corruption, it has been difficult to find clear evidence of the implication of the participant’s morality in their choices in bribery games. To signal the presence of norms, researchers have relied on negative externalities, punishment and frames, but they have yielded mixed or null results. In a bribery game framed in a neutral way Abbink, Irlenbusch and Renner (2002) showed that the participants were sensitive to the possibility of receiving sanctions, but not to produce negative externalities on the other participants in the session. The insensitivity to increasing degrees of externalities was also the result obtained by Barr and Serra (2009) in their neutral version of a single round bribery game. According to these results, games with abstract or neutral frames do not seem to have the aforementioned moral dimension, which questions their validity as instruments to study corruption. This study searched for evidence of a moral component in a bribery game. The experiment relied on two independent groups of participants (N=106) who played the game either framed in neutral terms or with options labeled with relevant semantic content (for example, offer a corrupt deal). In addition, selfish monetary incentives in the bribery game were increased across within-subject conditions, whereas the payoffs for making no transaction or for playing the Trust Game remained fixed. Finally, to assess the moral connotation of decision-making scenarios, participants were asked to complete (post-decision) a Machiavellian scale, which has been previously associated with different aspects of moral decision-making. The experiment managed to identify a robust frame effect for both Players A and B. Results showed that both potential bribers and potential bribees were significantly less prone to choose the corrupt transaction in the loaded frame than in the abstract frame. In addition, the tendency for corruption increased as the selfish monetary incentives involved in the corrupt transaction increased across conditions. Moreover, responses to the game showed individual differences in the level of Machiavellianism, as high Machiavellian scores were directly associated with corrupt offers, though only in the loaded frame. The data also suggests that the present game managed to capture the expression of pro-social preferences (the idea of social preferences is that individuals evaluate an outcome not only on the basis of its consequences for themselves, but also based on consequences to others). This was evident in the fact that, in general, participants avoided the option that generated negative externalities when there was an option without externalities that was equivalent in terms of personal gain. In addition, possible ways of improving the sensitivity of the bribery game to variables expected to be associated with corruptibility were presented. These suggestions could be followed in future studies. The discussion draws attention to potential practical implications of present findings, which suggest that targeting socially undesired actions using labels with morally negative nuances could be an effective mean to discourage people from choosing them.
Read full abstract