Abstract

This paper evaluates a public childcare program for children ages 0–4 in poor urban areas in Nicaragua. Our identification strategy exploits the program's neighborhood-level randomization as exogenous variation to tackle imperfect compliance with the original treatment assignments. We find a positive impact of 0.38 standard deviations on socio-emotional skills and a 12-percentage-point increase on mothers’ work, which makes the program highly cost-effective. We do not find evidence of substantial heterogeneity of impacts across observed or unobserved household characteristics, and we present suggestive evidence of the importance of center quality for generating positive impacts.