The Society for Social Work and Research

2013 Annual Conference

January 16-20, 2013 I Sheraton San Diego Hotel and Marina I San Diego, CA

Safety and Well-Being of Child Maltreatment Victims in a Low-Removal State Compared to the Balance of the U.S

Schedule:
Saturday, January 19, 2013: 10:30 AM
Executive Center 1 (Sheraton San Diego Hotel & Marina)
* noted as presenting author
Mark F. Testa, PhD, Spears-Turner Distinguished Professor, University of North Carolina at Chapel Hill, Chapel Hill, NC
Ijeoma J. Nwabuzor, MSW, Postdoctoral Research Associate, University of Southern California, Los Angeles, CA
Qi Wu, MSW, PhD student, University of North Carolina at Chapel Hill, Chapel Hill, NC
Kevin White, MSW, PhD Student, University of North Carolina at Chapel Hill, Chapel Hill, NC
Background and Purpose: Advocacy by leading child welfare organizations to reduce the national foster child population to half its size in the coming decade raises the question of what effect a decrease of this magnitude might have on the safety and well-being of children substantiated for child maltreatment. This study compares short-term longitudinal outcomes for child maltreatment victims in the second round of the National Survey of Child and Adolescent Well-Being (NSCAW II) to a supplementary NSCAW II sample of child victims in the state of Illinois. The substantial reduction in the size of the Illinois foster care population in the late 1990s to per-capita removal rates that are now among the lowest in the nation provides a unique opportunity to approximate the counterfactual of what impact similar low-removal rates might have on the safety and well-being of child maltreatment victims in the balance of the nation.  

Methods:The proposed presentation used propensity score matching and weighting to balance the population characteristics of 815 Illinois child victims in NSCAW II with those of 2,798 child victims in the balance of the U.S. Both procedures successfully reduced differences in child race and ethnicity, age at investigation, maltreatment type, and other characteristics to statistical insignificance. After balancing the two samples, the relative odds of removal from birth families were still 2.1 times greater in the balance of the U.S. compared to Illinois. Furthermore, the conditional odds of placement in unrelated foster care were 1.9 times greater than in Illinois. These differences in removal odds and restrictiveness of placement are consistent with policies that Illinois implemented favoring family preservation and kinship foster care in the mid-1990s.

Results: Results suggest that low-removal odds in Illinois did not adversely affect child safety, health, and other well-being outcomes 18-months after investigation compared to similar children who experienced higher removal rates and more restrictive placements in the balance of the nation. For example, 9.2% of children from Illinois were re-reported for abuse, compared to 9.1% of children from the U.S. sample (p = 0.9603). In addition, 50.3 % of Illinois caregivers reported that their children had excellent health, compared to 47.5 % of U.S. caregivers (p = 0.1682). Even though the difference in out-of-home placement rates persisted a year and one-half after investigation, the higher U.S. rates of reunification from unrelated foster care reduced the difference between Illinois and the rest of the U.S. to statistical insignificance.

Implications:  Findings suggest that more aggressive family preservation policies to reduce out-of-home care and a greater reliance on kinship foster placements to lower the restrictiveness of foster care may not jeopardize short-term child safety and well-being and may possibly improve longer-term outcomes. The higher rate of restoration to parental custody following removal in the rest of the U.S. compared to Illinois suggests that greater efforts could be made to prevent removal without jeopardizing child safety and well-being. The paper also discusses the appropriateness of weighting estimates after propensity score matching to account for complex sampling design.