Normative and rational forces driving corrupt decisions among students within social networks: A vignette study

Beside the benefits of social capital for individuals and the society as a whole (such as weak ties for obtaining information or generalized trust) there is also a dark side of social capital, for instance, corruption. Corrupt actors try to veil their illegal actions which points to the fact that most corrupt actions do not occur between strangers, but rather between actors who are embedded in long-term social networks. Corruption can be considered as a result of a deliberation process including the evaluation of the trustworthiness of the partner in crime, benefits from the deal, potential cost, and the probability of being detected.

Corrupt relationships are usually considered within a principal-agent-client model. This model presumes an agent (here: a teaching assistant) to pursue his/her opportunistic interests and the interests of the client (a student) also, while both actors violate the principal’s (university) guidelines. Although the principal promotes legal/regulative principles (universalistic code of conduct), the agent takes advantage of his/her position and follows a particularistic norm which benefits the client.

A vignette was presented to 2.262 students describing a typical principal-agent-client corruption situation at university. We varied the situational conditions (trustworthiness of the client, benefits, costs, probability of detection). The students were asked whether or not they would commit corruption if they were the agent in that particular situation.

Our findings support both a decision perspective and the assumption of conflicting norms where norm divergence is a stronger explanation factor than deliberation. For the largest subgroup of the sample universalistic norms prescribe (non-corrupt) behavior ceteris paribus while in another subgroup universalistic norms are debased in favor to particularistic motives. Finally in a smaller group we find evidence for the dominance of hard instrumental incentives.

Our study implies that policy and organizational measures against corruption could be improved when norm-conflicts are regarded.