Abstract
Using a unique longitudinal survey from Ethiopia, we investigate whether resource constrained parents reinforce or attenuate differences in early abilities between their children. We propose a simple model that allows for sibling interactions.
To overcome the endogeneity associated with measures of endowment, we construct a measure of human capital at birth that is plausibly net of prenatal investment. We estimate a sibling fixed-effect model to account for bias due to unobserved family-specific heterogeneity.
We find that parents reinforce educational inequality: inherently healthy children are more likely to attend preschool, be enrolled in elementary school, and have more expenses incurred towards their education. Health inputs are allocated in a compensatory manner.