Suchergebnisse

The public Indian health care system is plagued by high staff absence, low effort by providers, and limited use by potential beneficiaries who prefer private alternatives. This artice reports the results of an experiment carried out with a district administration and a nongovernmental organization (NGO). The presence of government nurses in government public health facilities (subcenters and aid-posts) was recorded by the NGO, and the government took steps to punish the worst delinquents. Initially, the monitoring system was extremely effective. This shows that nurses are responsive to financial incentives. But after a few months, the local health administration appears to have undermined the scheme from the inside by letting the nurses claim an increasing number of "exempt days." Eighteen months after its inception, the program had become completely ineffective.

This paper studies the problem of experiment design by an ambiguity-averse decision-maker who trades off subjective expected performance against robust performance guarantees. This framework accounts for real-world experimenters' preference for randomization. It also clarifies the circumstances in which randomization is optimal: when the available sample size is large and robustness is an important concern. We apply our model to shed light on the practice of rerandomization, used to improve balance across treatment and control groups. We show that rerandomization creates a trade-off between subjective performance and robust performance guarantees. However, robust performance guarantees diminish very slowly with the number of rerandomizations. This suggests that moderate levels of rerandomization usefully expand the set of acceptable compromises between subjective performance and robustness. Targeting a fixed quantile of balance is safer than targeting an absolute balance objective. (JEL C90, D81)