Soft Information and the Cost of Job Rotation: Evidence from Loan Officer Rotation

Job rotation, where a principal routinely rotates agents among tasks, is argued to be a powerful antidote for agency problems inside an organization. However, when soft information dominates transactions inside a firm, verifying the information set that led to a particular decision becomes difficult. This difficulty imposes costs on job rotation since an incoming agent cannot verify the information set that the outgoing agent utilized when arriving at a key decision. This lack of verifiability distorts incentives for effort when a decision straddles two agents since neither agent receives the entire marginal cost/benefit of her effort. In this study, we highlight this cost of rotation policies. We use unique data on over 50,000 loans sanctioned by 51 loan officers of a large public sector bank in India, which follows a fixed-tenure-based policy of loan officer rotation. We find default probabilities to be about 8% higher for loans that are likely to straddle an incoming and an outgoing loan officer when compared to other loans. This difference does not stem from any differences in hard information captured in borrowers' credit histories or from loan officer rotation destroying lending relationships. Finally, we find evidence of credit rationing as well as the new loan officer rejects 7% more loans for borrowers who have borrowed during the straddling period.