The Determinants and Consequences of the Placement of Government Programs in Indonesia

Most estimates of the consequences of public programs rely onthe cross-sectional association between region-specific programsand program outcomes. Such estimates assume that the spatialdistribution of programs is random. This article reports estimatesof the effects of public programs on basic human capital indicatorsand the biases in conventional cross-sectional estimates ofprogram effects due to non-random program placement. The estimatesare obtained from pooled observations on human capital outcomes,socioeconomic variables, and program coverage at the kecamatan(subdistrict) level. The observations are based on successivesets of Indonesian cross-sectional household and administrativedata during 1976–86. The determinants of the spatial allocationof programs in Indonesia in 1976–86 are also estimated. The empirical results indicate that the presence of grade andmiddle schools in villages has a significant positive effecton the school attendance rates of teenagers. The presence ofhealth clinics in villages also positively affects the schoolingof females ages 10–18. However, no evidence is found ofany significant effects of the presence of family planning andhealth programs on either the survival rates of children oron cumulative fertility. The estimates also suggest that theuse of cross-sectional data results in substantial biases inthe estimates of program effects because of the evident nonrandomspatial allocation of public programs.