- » Differential Response to Reports of Child Abuse and Neglect
- » Evaluation Findings
Differential Response to Reports of Child Abuse and Neglect
Series: Issue Briefs|
Child Welfare Information Gateway. |
|Year Published: 2008|
With any systems reform effort, evaluation is critical to understanding whether the program is being implemented as intended, assessing overall effectiveness, and identifying and sharing lessons learned. Several States—including Florida, Minnesota, Missouri, North Carolina, Texas, Virginia, and Washington—have conducted evaluations of their statewide or pilot differential response systems. In addition, the first large-scale, multistate study was published in 2005 (Shusterman et al.) based on an analysis of case data reported from six States (Kentucky, Minnesota, Missouri, New Jersey, Oklahoma, and Wyoming) to the National Child Abuse and Neglect Data System (NCANDS).
Overall, the evaluations of differential response systems have demonstrated positive outcomes, particularly in terms of sustained child safety, improved family engagement, increased community involvement, and enhanced worker satisfaction. Evaluations of pilot programs have generally led to decisions to expand implementation. Several evaluations, however, noted that continuing problems with the adequacy of resources such as staffing and services limited both implementation and the degree of positive change.
Specific findings from these evaluations are presented below as they relate to the following topics:
- Referral and substantiation
- Child and case characteristics
- Child safety
- Investigations and prosecution
- Services to families
- Family satisfaction and engagement
- Cost effectiveness
- CPS staff perspectives and issues
Referral and Substantiation
The research revealed:
- The proportion of reports diverted to an alternative response varied greatly across States. The multistate study found that during 2002, referrals ranged from a low of 20 percent to a high of 71 percent across the six States studied. An analysis of multiyear trends suggested that States were experiencing growth or steady use of the alternative approach over time (Shusterman et al., 2005).
- The proportion of investigations that were substantiated increased (Loman & Siegel, 2004a; Virginia Department of Social Services, 2004). This reflects the inclusion of a larger share of serious cases in the investigation track after less serious cases were diverted to receive services, which is in line with the stated intentions of differential response systems (Shusterman et al., 2005).
- Differential response resulted in a decrease in the numbers of both victims and nonvictims identified by States. The amount of the decrease varied by State, however, and in one State the number of nonvictims increased (Shusterman et al., 2005). Decreases are to be expected given that cases on most assessment tracks do not require a decision on substantiation.
Child and Case Characteristics
Research on child and case characteristics noted:
- An alternative response was more likely to be used for cases with less immediate safety concerns, and less likely to be used in sexual abuse cases (Shusterman et al., 2005; Virginia Department of Social Services, 2004; Loman and Siegel, 2004a; U.S. Department of Health and Human Services, 2003a). This finding is consistent with the stated intentions of differential response systems. While both the multistate study and single-State research found that the link between maltreatment type and referral track was strong, the relationships varied across States (Shusterman et al, 2005).
- Older children generally were more likely to receive an alternative response, while younger children were more likely to be assigned to investigation (Shusterman et al., 2005; Siegel & Loman, 2000; Chipley, Sheets, Baumann, Robinson, & Graham, 1999; English, Wingard, Marshall, Orme, & Orme, 2000). This suggests that track assignment may take into account the greater vulnerability of younger children.
- Children and families who were referred to an alternative response were similar in demographics (e.g., gender, race, ethnicity, family structure) to those who received traditional investigations (Shusterman et al., 2005; English et al., 2000; Siegel and Loman, 2000).
- Prior victimization was often related to a decreased likelihood of an alternative response, but not in all States (Shusterman et al., 2005).
- Referrals from social workers, medical personnel, and legal or criminal justice sources were less likely to receive an alternative response. Referrals from parents, relatives, friends, schools, or the children themselves were more likely to be referred to an alternative response (Shusterman et al., 2005; English et al., 2000).
The ability of differential response systems to protect child safety is a significant concern. Research findings suggest that:
- Child safety was not compromised under differential response systems. Single State studies revealed that children whose cases received an alternative response were less likely or as likely as children who received an investigation to be the subject of a subsequent report or investigation (Chipley et al., 1999; English et al., 2000; Loman & Siegel, 2004a; Loman & Siegel, 2004b; Virginia Department of Social Services, 2004; Center for Child and Family Policy, 2004). Likewise, the multistate study found that the rate of recurrence within 6 months was comparable for children whose cases received an alternative response versus investigation; in Oklahoma, the likelihood of receiving a subsequent CPS response within 6 months was lower for children receiving alternative response (Shusterman et al, 2005).
- Safety was maintained even when comparable families were randomly assigned to tracks. In an experimental study conducted in Minnesota, families randomly assigned to assessment were significantly less likely to be re-reported than families randomly assigned to investigations (27 percent versus 30 percent) (Loman & Siegel, 2004b).
- Increased services to families lowered recurrence. Analyses using Minnesota's experimental design support this expected outcome. In addition, the nonadversarial and participatory approach to families was linked to reduced recurrence whether or not services were delivered (Institute of Applied Research, 2005).
Investigations and Prosecution
Limited research has focused on investigations and prosecution. Findings thus far suggest:
- States differed widely in the extent to which the existence of an alternative response option resulted in fewer investigations (Shusterman et al, 2005).
- The use of a differential response system improved investigations and increased criminal arrests in one State. A recent study (Loman, 2005) examined the arrests in cases of child sexual abuse and severe and moderate physical abuse in Missouri's differential response system, which diverts the majority of other reports to a nonadversarial home visit. The study found that limiting the number of investigations led to more intense investigations and a greater likelihood of criminal prosecution of perpetrators of the more serious offenses.
Services to Families
Research on services provided to families on the assessment track showed:
- Services were provided more often to children and families on the assessment track (Shusterman et al, 2005; Loman & Siegel, 2004a; Virginia Department of Social Services, 2004; Hernandez & Barrett, 1996). In Minnesota, for example, 54 percent of families on the assessment track received specific services (other than case management), compared to 36 percent of families on the investigation track (Loman & Siegel, 2004a).
- The number of services received by families on the assessment track was greater than on the investigation track. Linkages of families to funded and unfunded community providers increased in both Minnesota and Missouri (Loman & Siegel, 2004a; Loman & Siegel, 2004b). In addition, the types of services delivered to families shifted in both States toward family support services related to basic financial needs (Institute of Applied Research, 2005).
- Services may be provided to families earlier on the assessment track. Missouri found that services occurred in a more timely manner under differential response (service activity was initiated on average within 17 days in the pilot areas, versus 35 days for comparison families) (Siegel & Loman, 2000).
- Greater use of community resources was reported in pilot areas of at least three States (Florida, Minnesota, and Missouri) (Siegel and Loman, 2000; Loman & Siegel, 2004b; Hernandez & Barrett, 1996). One evaluation report, however, indicated that community agencies were not always able to make contact with families or make contact within time frames anticipated by CPS (Washington State DSHS, 2005).
- Children were more likely to be placed in foster care if they received investigations, in several but not all States (Shusterman et al., 2005; Virginia Department of Social Services, 1999; Chipley et al., 1999; Loman & Siegel, 2004b). Recurrence of maltreatment resulting in removal declined for families receiving alternative response in the Minnesota study, but no corresponding evidence could be found in the Missouri study (Institute of Applied Research, 2005).
Family Satisfaction and Engagement
A few States conducted surveys to explore family and worker perspectives on family satisfaction and engagement. They found:
- Families reported satisfaction with the differential response system in Minnesota, Missouri, North Carolina, and Virginia (Loman & Siegel, 2004a; Siegel & Loman 2000; Center for Child and Family Policy, 2004; Virginia Department of Social Services, 1999). Further, the Minnesota families receiving alternative response reported that they were treated in a friendly and fair manner, were listened to by workers, were connected to other community resources, and benefited from the CPS intervention more often than families receiving a traditional response. These same families more often reported being hopeful and encouraged (Loman & Siegel, 2005).
- The family's sense of participation in decision-making increased in several States (Loman & Siegel, 2004a; Loman & Siegel, 2004b). In Minnesota, 68 percent of assessment families said they were involved a great deal in decisions that were made about their families and children, compared to 45 percent of control families (Loman & Siegel, 2005). In Virginia, families were included in planning for services in 95 percent of assessments and 67 percent of investigations (Virginia Department of Social Services, 1999).
- Workers reported families were more cooperative and willing to accept services. In Minnesota, for example, workers rated the primary caregiver as uncooperative in less than 2 percent of assessment families, compared to 44 percent of control families (Loman & Siegel, 2005). In Missouri, it was hypothesized that cooperation between families and the child welfare agency was linked to the more positive and supportive orientation and earlier service contacts (Siegel & Loman, 2000).
A cost analysis showed promising results:
- Differential response appears to be cost effective over the long term. Minnesota's cost-effectiveness study suggested that costs of alternative response in the early stages of a case, including worker time during case opening, were greater than in traditional CPS interventions. However, costs for case management and other services following the closing of the initial case through the end of the follow-up period were lower. Savings achieved later more than offset investment costs early on and, as such, total costs were less for the alternative response cases than the control cases (Loman & Siegel, 2005).
CPS Staff Perspectives and Issues
Surveys and interviews with CPS staff underscore:
- CPS staff like the differential response approach. In Missouri, workers and community representatives preferred the family assessment approach (Loman & Siegel, 2004a). North Carolina social workers and supervisors overwhelmingly agreed that the differential response system was a more respectful way to serve families and allowed them to consider all family circumstances (Center for Child and Family Policy, 2004). Virginia CPS staff also expressed favorable views of their multiple response system (Virginia Department of Social Services, 1999).
- Large caseloads and limited resources are obstacles to differential response effectiveness. Missouri's evaluation concluded that the impact of the demonstration was mitigated by large caseloads and limited resources. Although the results of the evaluation favored the family assessment approach over the traditional CPS approach, the effects were relatively modest. To achieve greater impact, the evaluators recommended reducing worker caseloads, as well as increasing and accelerating community development activities and resources (Loman & Siegel, 2004a). In North Carolina, staff members experienced increased challenges in managing cases while working with new reports, leading to increased stress levels. Evaluators recommended limiting caseload sizes to six to eight families per worker or implementing team models (Center for Child and Family Policy, 2004).
- Training is needed. The Virginia evaluation led to some specific recommendations for program expansion, including providing training for both frontline staff and administrators to communicate changes to other agencies and the community (Virginia Department of Social Services, 1999). Similarly, North Carolina evaluators called for additional and better training for line staff, supervisors, and management.
This material may be freely reproduced and distributed. However, when doing so, please credit Child Welfare Information Gateway.