How Bad Do We Look? A Field Experiment on Reputational Incentives and Corrupt Transactions

• In Event: 3224 - The Meanings of Money, Property, and Relationships

Abstract

Can group-level reputational incentives--which affect how work colleagues perceive their personal behavior and each others' behavior--reduce transactional corruption during public service provision? To answer this question, I design and implement a field experiment in the maternity services of five public hospitals in Morocco. In this context, corrupt behavior denotes bribe payments made by patients to healthcare providers--doctors, midwives, and nurses--in exchange for improving on the quality of care or assisting with secondary services associated with the care experience. This is the first study to leverage social incentives in transactional corruption control and to accomplish this difficult task inside a public sector organization. The findings indicate that starting corruption levels determine how and if providers respond to reputational incentives. In particular, services with higher starting levels of corruption experience the greatest gains in corruption reduction. The experimental results, along with secondary analyses, are consistent with the explanation that providers have a threshold for allowable levels of corrupt behavior. These results depart from standard sociological and microeconomic expectations of how social influence shapes misbehavior. I discuss the theoretical and policy implications for the treatment and understanding of transactional corruption, as well as outline future research directions.

Author

• Diana Dakhlallah, Stanford University