UC San Diego

This dissertation studies questions related to child maltreatment and the child protection system (CPS). In the first chapter, I provide new evidence of the effects of income on child maltreatment. I exploit a discontinuity in child-related tax benefits, which results in otherwise-similar families receiving considerably different refunds during the first year of a child's life. I use 20 years of linked administrative data from California to determine the effects of this additional income on CPS involvement. A one-time $1,000 transfer to low-income households decreases referrals to CPS in the first 3 years of life (3%), investigations (3%), substantiated referrals (4%) and days spent in foster care (8%). Effects persist through at least age 8. These findings suggest that providing low-income families with additional resources during the first year of a child’s life is a fruitful strategy for reducing child maltreatment.

In the second chapter, we ask whether providing decision-makers with a machine learning tool can reduce racial disparities. Our context is the implementation of the Allegheny Family Screening Tool (AFST), a predictive risk model that aims to help CPS workers decide which allegations of maltreatment to screen in for investigation. Using difference-in-differences and triple difference designs, we find that the AFST reduced disparities in screening decisions, as well as case opening and home removal rates for investigated referrals involving Black vs. white children.

In the third chapter, I study the effects of a major immigration reform on the incidence and reporting of child maltreatment in Hispanic households. Secure Communities ties federal immigration enforcement to local law enforcement, effectively increasing the likelihood of deportation for undocumented immigrants who are arrested for a crime. I exploit the staggered rollout of Secure Communities across counties to estimate an event study triple differences model. I find that Secure Communities decreased the rate of investigated child maltreatment cases involving Hispanic children. These results suggest that the policy had the unintended consequence of reducing the willingness of Hispanic individuals to report potential cases of child maltreatment.