Implementation of Differential Response in Child Welfare: A Comparative Case Analysis

Background and Purpose: Differential response refers to a set of policies that establishes at least two distinct pathways for families that are reported for child maltreatment, including: (1) a traditional investigation pathway for high risk cases; and (2) the differential response pathway for low to moderate risk families. Over half the states in the U.S. are now using this approach in child welfare.  Several prior evaluations have indicated that differential response increases family engagement, improves family and worker satisfaction and may improve child safety by reducing recidivism of child maltreatment (QIC-DR, 2011).  This study examined the implementation of Minnesota’s differential response approach, called Family Assessment (FA) response to better understand which components of FA are linked to improved outcomes. Minnesota has a county-administered child welfare system, and this study explored which aspects of FA implementation could help account for differences in outcomes by counties.

Methods:  In this comparative case study, counties were selected based on racial equity and child safety outcomes from analysis of county-level child welfare data from 2003 to 2010 (e.g., outcomes such as racial disparities at key decision making points and rates of re-reporting for children in FA).  The sample included 9 counties (3 counties from each of these categories) that demonstrated: (1) overall positive outcomes; (2) overall negative outcomes; and (3) mixed outcomes.  Data was collected from semi-structured focus groups with child welfare workers and interviews with supervisors.  The interview guides were informed by key concepts in implementation research and organizational theory focusing on these domains: organizational capacity; worker practices; contextual factors and organizational climate.  The use of comparisons in the data analysis helped to move from the focus on a single case to cross-case understanding; and allowed the researcher to check biases and assumptions through disconfirming evidence and divergent patterns (Corbin and Strauss, 2008; Miles & Huberman, 1994). 

Results: Themes emerged suggesting counties that had effectively integrated approaches that were both child-safety focused and family-engaging, had improved implementation and improved outcomes. Integrated implementation strategies included: more comprehensive assessment protocols; increased capacity of staff and supervisors through integrated training efforts; having “champions” that promoted team cohesion and support; and combined resources for group case consultation and on-going coaching of effective practices in child welfare. Successful outcomes were also linked to availability of resources, including “natural” or enduring supports for parents, supports to help families meet basic needs and culturally responsive services for families.

Implications: Many prior studies on FA were conducted as initial pilots; so the findings of this current study build unique knowledge about sustaining implementation strategies as contextual factors change over time – as a state moves from pilot phase to on-going, institutionalized practice. This is particularly timely, as differential response is a growing approach in child welfare. The findings from this study also suggest that research and evaluation methodologies are needed that more effectively measure and monitor the impact of multiple, comprehensive and integrated approaches to practice, as many child welfare jurisdictions have practitioners implementing several new strategies simultaneously.