Accountability in public child welfare: Linking program theory, program specification and program evaluation

In public child welfare services, an increasing emphasis over the past decade on outcome accountability has not changed the prevailing perception of stakeholders that public child welfare agencies are not accountable. A host of demonstration projects have been required to measure and account for results. The evaluations of these projects have rarely provided evidence of program effectiveness or led to program enhancement. A major contributing factor has been the lack of fit between the model of causation that is the basis for most outcome-oriented evaluative research and the reality of how most large, complex public social welfare programs are developed and implemented. A realistic evaluation approach pioneered by British sociologists is more closely articulated to the reality in large, public child welfare agencies. The potential for transforming a large, public child welfare agency into a more accountable organization is illustrated with the developmental and realistic approach to program evaluation that has been initiated in Los Angeles County's Department of Children and Family Services (DCFS). This article describes the development of program theory, how program theory has been linked to program specification, how program specification has been linked to planning for program evaluation and how a realist approach solves some on-going problems in the evaluation of public child welfare programs.