A Population-Based Analysis of Race and Poverty As Risk Factors for Maltreatment

Background. This study provides a population-level examination for racial disparities in child maltreatment reporting, substantiated victimization, and foster care entry. Using a unique dataset constructed by linking child welfare to birth records, over half a million children born in California in 2002 were prospectively followed through their fifth birthday. Sociodemographic risk factors gleaned from birth records were used to model the likelihood of contact with the child welfare system, from a first report to entry to foster care.

Method. The relative risk of a referral to CPS, substantiation, and entry to foster care before the age of five was computed using Generalized Linear Models (McCullagh & Nelder, 1989). Adjustments were made for sociodemographic and biomedical risk factors gleaned from the birth records (i.e., child sex, birthweight, birth abnormalities, prenatal care, maternal birthplace, maternal race/ethnicity, birth payment method, maternal age, maternal education, abortion history, paternity information, and birth order).

Results. Fourteen percent of children in the 2002 birth cohort (74,182) were reported for maltreatment, six percent (27,805) found to be victims, and under one percent (4,388) entered foster care. We found significant interactions between a number of the covariates and birth payment method, which led us to stratify multivariate models into an indicator that separated Medicaid recipients from others (who mostly had health insurance). This payment indicator can be viewed as a crude SES indicator, with the Medicaid group presumed to be of lower SES. Forty eight percent of the Black children were born to mothers in the Medicaid group, compared to nineteen percent of the White children. Our analyses demonstrated how important it is to look below summary indicators of disparity. While overall Black children were more than twice as likely to be reported (Risk Ratio 2.25), victimized (Risk Ratio 2.48), and/or enter foster care (Risk Ratio 2.55) as White children, stratification by birth payment clarified that disparities were virtually non-existent, or even reversed, for the Medicaid (lower SES) group.

Conclusions. Results indicate that while Black children were more than twice as likely as White children to be reported, victimized, and enter foster care in the crude models, disparities were either non-existent or reversed after adjusting for other risk factors and stratifying by the birth payment method. White children born into poverty were actually at greater risk of entering foster care than similarly situated Black children. This presentation will discuss the implications of these race findings in the context of a broader discussion of interactions between poverty and race.