Do civil servants respond to behavioral interventions? : a field experiment

Introducing financial incentives to increase productivity in the public sector tends to be politically and bureaucratically cumbersome, particularly in developing countries. Behavioral interventions could be a low-cost alternative, both politically and financially, although evidence of their effectiveness remains scarce. We evaluate the effect of redesigning the notice requiring civil servants in Buenos Aires to comply with citizens requests under Argentina's freedom of information act. The new notice, sent to the treatment group, attempts to exploit salience, deterrence, clarity, and social norms to increase adherence to deadlines. The results show an increase in the share of requests fulfilled by the second deadline, possibly because of a strong anchoring effect. These findings indicate that behavioral interventions can affect civil servants' actions. The fact that the intervention occurred at the same time as a civil service training program with sessions attended by members of both the control and treatment groups allows us to evaluate spillover effects. The evidence suggests that the time it takes a members of the treatment group to respond to a request increases with her interactions with members of the control group at the workshops. These findings have implications for policy design. First, they indicate that behavioral interventions could affect task compliance and productivity in the public sector. Second, they provide evidence that workshops may not always have the intended consequences, particularly when they increase interactions among employees with high and low incentives for task compliance.