Association for Public Policy Analysis & Management

The national workfare program in India guarantees employment to unskilled laborers in rural areas and hires them to work on public projects. This paper focuses on the state of Jharkhand where an audit was randomly-assigned over the course of a few years and carried out at gram panchayats, the smallest implementing government body of the workfare program. The paper studies how the audit and announcement of the audit affected the lags and leads of monthly performance of bureaucrats responsible for implementing the program and the productivity of the program’s commissioned public projects.

Results show that bureaucrat performance on the amount of employment generated sees a stark decrease around the time of audit and performance reverts to pre-treatment trends afterward. The results are similar when the audit is announced. Subsequent analyses attempting to disentangle mechanisms behind the bureaucrat's response suggest the decline in employment generated is due in part to a decline in fabricated employment. These results question the persistence of the effect of audits on bureaucrat behavior found in the positive results of audits in previous literature, and suggests that the design of the audit policy and maintaining a persistent threat to being caught is necessary for a successful monitoring program.