Abstract
The present paper provides new causal evidence that private practice of public health professionals may encourage moonlighting with implications for the provision of public health services. Our identification strategy relies on the introduction of the 1997 Ministry of Health Regulation 916 in Indonesia that requires health professionals to apply for license for private practice after three years of graduation. Arguing that the timing of obtaining the license for private practice is random after the 1997 regulation, we use four rounds of puskesmas-level Indonesian Family Life Survey data over a period of 1993 to 2007 from their heads to assess the impact of private practice on dual practitioner's public labour supply indices. We employ a difference-in-difference regression discontinuity design with a view to test the validity of moonlighting hypotheses. Results are robust and suggest that private practice is associated with significantly higher number of public patients seen, but lower hours worked per week by the dual practitioner heads. We also show that the result is driven by those for the urban region where we expect the results to be stronger, thus allaying concerns for omitted factors and confounding events around 1997. Finally we show that the number of patients seen by the dual practitioner head is correlated with the referral of public patients to private clinics. The results can be attributed to the weak incentive mechanism for the heads of the puskesmas who are unlikely to be fired even if caught shirking, thus limiting the effectiveness of the regulation to redress the problems of moonlighting