Abstract:
We study the effect of unemployment on child maltreatment in the United States for the period 2004-12, at the county level. We use a new administrative dataset containing every report of child abuse and neglect made to the Child Protective Services, and identify the effect using a Bartik instrument. A one percentage point increase in the unemployment rate causes a 20 percent increase in neglect. We provide evidence that parents' lack of social and private safety nets is an important mechanism behind the effect. Our results capture an effect on the actual incidence of neglect and not reporting behaviour.