ABSTRACT
This article measures the impact of fostering on children's outcomes in Niger. We estimate a simultaneous equations model with three outcomes for children (school attendance, hours of market work and hours of domestic work) and a treatment variable (fostering). Nonlinearity provides identification by functional form of the causal effect of fostering. We further use information on the historical level of community integration of the child's ethnicity and the relative importance of the ethnicity in the neighborhood to build a plausibly exogenous exclusion restriction in order to correct for the potential endogeneity of fostering. Our results sustain the overall conclusion that fostering has a positive impact on children's school attendance and hours of domestic work. However, while foster boys and girls are both engaged in longer hours of domestic labor, only foster boys are advantaged in terms of schooling.
