RESUMO
This paper evaluates the health effects of a large-scale subsidizing program of prescription drugs introduced in Brazil, the Aqui Tem Farmácia Popular program (ATFP). We exploit features of the program to identify its effects on mortality and hospitalization rates by diabetes for individuals aged 40 years or more. We find weak evidence for a decline in mortality, but a robust reduction in hospitalization rates. According to our preferred specification, an additional ATFP pharmacy per 100,000 inhabitants is associated with a decrease in hospitalization rates by diabetes of 8.2, which corresponds to 3.6% of its baseline rate. Effects are larger for Type II diabetes in comparison to Type I, and among patients with relatively lower socioeconomic status. Overall, the results are consistent with insulin-dependent patients being relatively less responsive to subsidies because of higher immediate life-threatening risks; and with lower-SES individuals being more responsive because of liquidity constraints. These results support the view that the optimal design of health systems and cost-sharing mechanisms should take into account equity concerns, heterogeneous impacts by health condition, and their potential offsetting effects on the utilization of downstream health services.