Estimating health insurance impacts under unobserved heterogeneity: the case of Vietnam's health care fund for the poor.

Vietnam's health care fund for the poor (HCFP) uses government revenues to finance health care for the poor, ethnic minorities living in selected mountainous provinces, and all households living in communes officially designated as highly disadvantaged. As of 2006, the program, which started in 2003, covered around 60% of those eligible. Those who were covered (about 20% of the population) were disproportionately poor, and around 80% of those covered were eligible. Estimates of the program's impact were obtained using a method that takes into account unobserved heterogeneity--including unobserved idiosyncratic returns--but requires minimal assumptions. The downside is that it provides an estimate only of the program's impact on those covered by it; it cannot therefore answer the question of how those currently uncovered will fare when they are eventually covered. The results suggest that HCFP has had no impact on use of services, but has substantially reduced out-of-pocket spending.