Who cares for our children matters: early maternal employment, early childcare, and child development in Chile

Worldwide, non-maternal care during the first years of life has gradually become more prevalent. However, there is little evidence about the effect of non-maternal care— especially for under-three-year-olds—on child development. Hence, this thesis explores the association between both maternal employment and type of care at different stages during children’s first three years of life and child development in Chile. My results indicate that there is evidence that maternal employment during the child’s first year of life is detrimental to child development and that delaying maternal employment initiation decreases this detrimental effect on child development. On the other hand, the type of care that the child attends during this first year of life also matters. First, children who are looked after by their grandparent during their first year of life exhibit a positive association with child cognitive and socio-emotional development relative to exclusive maternal care. Second, there is a negative association between relative care and child cognitive and socioemotional development compared to exclusive maternal care. Third, there is a positive association between centre-based care and child cognitive development and a slightly positive association with child socio-emotional development. Finally, controlling for unobserved and fixed child characteristics, I analyse whether the positive association between centre-based care at 6 to 12 months old and child development is also observed on children who entered centre-based care between the ages of 24 and 36 months old. The association between centre-based care between centre-based care and child cognitive development is also positive and there is no significant association with child socioemotional development. In each of my empirical chapters, I test whether child vulnerability define as lowly educated mothers, single parent and low income families, moderates the association between early non-maternal care and child development. Overall, the previously described associations are slightly more detrimental for more vulnerable children. In my three empirical chapters, I use a novel Chilean longitudinal panel survey with waves in 2010 and 2012. To deal with selection bias, in two out three empirical chapters I control for an extensive set of child, mother and family characteristics using OLS regressions and propensity score matching techniques. In addition, in the last empirical chapter I control for (unobserved) individual fixed effects