Differential Response and Children Re-Reported to Child Protective Services: County Data from the National Child Abuse and Neglect Data System (NCANDS)

Background and Purpose:

Child Protection systems that implement Differential Response (DR) systems screen to route referrals to an investigation response (IR) or alternative response (AR). AR responses emphasize family engagement, assessment of family needs, and service linkage. Usually, AR state level policy does not require child welfare staff to make a maltreatment determination. Families served in the AR track have been found to be more cooperative, realistic, motivated, and have higher self-esteem compared with families in the IR track. Families served in the AR track show higher levels of positive emotional response to the intervention and higher levels of satisfaction, associated with increased family engagement and increased use of services (Loman & Siegel, 2004 & 2013; Loman, Filonow, & Siegel, 2010; Merkel-Holguin et al., 2014; Hollinshead et al., 2015; Hollinshead et al., 2017). Jurisdictions implement DR systems differently, leading to variations in the proportion of AR cases, risk levels of cases served, and the ways families use services. In this study, we explore how differences in county AR utilization rates, after controlling for various county characteristics, are associated with county-level rates of maltreatment re-reports. The rate of re-reporting is examined for all cases, and separately for children whose initial response was IR or AR. 

Methods:

County data from the National Child Abuse and Neglect Data System were analyzed for six states from 2004-2013 that had implemented DR prior to 2004. Child-level NCANDS data were aggregated at the county level for each year to produce counts and rates of child response events. These data were drawn from a Child File maintained by the Children’s Bureau which encrypted the county identifiers and added American Community Survey data (U.S. Census Bureau, 2014 & 2014b) prior to analysis. Multilevel random-coefficients models were examined for each of the outcome variables: (a) county-year overall re-report rates (combined AR and IR re-report rates), (b) county-year AR re-report rates, and (c) county-year IR re-report rates. Variation in county-level AR rates were associated with county level re-report rates using regression models with risk adjustments for socio-economic and other county characteristics.

Results:

From the risk adjusted analysis, counties had three percent fewer re-reports overall for each percent increase in AR use; higher levels of AR use are related to lower levels of re-reporting. When county AR and IR cases were analyzed separately, increasing rates of AR were associated with lower re-report rates for IR cases, but higher re-report rates for AR cases.

Conclusions and Implications:

The findings of this study are applicable only to the six states; the implementation experiences of other states could be different. Findings for the AR and IR subgroup must be interpreted with caution since technical factors may be driving these results. Results from this study suggest that child safety, as measured by re-report rates, was improved by increased AR utilization in DR systems. This finding is consistent with the theory of change for engagement enabling policy; policies can be put in place that reinforce engagement practices which in turn enhance service uptake.