Estimating the effects of Head Start on parenting and child maltreatment

We examine the effects of Head Start participation on parenting and child maltreatment in a large and diverse sample of low-income families in large U.S. cities (N=2807), using rich data from the Fragile Families and Child Wellbeing Study (FFCWS). To address the issue of selection bias, we employ several analytic approaches, including logistic regressions with a rich set of pretreatment controls as well as propensity score matching models, comparing the effects of Head Start to any other arrangements as well as specific types of other arrangements. We find that compared to children who did not attend Head Start, children who did attend Head Start are less likely to have low access to learning materials and less likely to experience spanking by their parents at age five. Moreover, we find that the effects of Head Start vary depending on the specific types of other child care arrangements to which they are compared, with the most consistently beneficial protective effects seen when Head Start is compared to being home in exclusively parental care.