Data-based organizational change: the use of administrative data to improve child welfare programs and policy.

Administrative databases hold the potential to have a significant impact on the development of effective child welfare programs and policies. This article discusses the strengths and weaknesses of administrative databases, issues with their implementation and data analysis, and effective presentation of their data at different levels in child welfare organizations. The development and use of administrative data in public and private child welfare agencies presents both challenges and opportunities. There is little doubt that the child welfare field is on a trajectory of increased development of administrative databases (ADBs). Most, but not all states have responded to the opportunities and requirements to develop administrative databases afforded by the 1986 Title IV E Social Security Act, and the 1993 Omnibus Budget Reconciliation Act (PL. 103-66). The Title IV-E Social Security Act mandated that states establish an Adoptions and Foster Care Analysis and Reporting System (AFCARS), and P L.103-66 provided funding incentives for Statewide Automated Child Welfare Information Systems (SACWIS). More recently, the SACWIS requirements have been translated into requirements for a National Child Abuse and Neglect Data System [NCANDS 1993]. In addition to federal efforts to promote a national child welfare database, national child welfare organizations and committees have been calling for the development of these data systems as part of the general movement toward outcomes and accountability measures (see Courtney and Collins [1994] and McDonald et al. [1989] for review). Both the American Humane Association and the Child Welfare League of America have promoted child welfare outcome-oriented initiatives to assist the field in developing and implementing program outcomes [Magura & Moses 1986; Gordon 1999; McDaniel 1999]. Administrative databases are a key component to the integration of outcomes into child welfare practice and policy development [Roos & Shapiro 1999]. Drake and Jonson-Reid [1999] outline many of the strengths and weaknesses inherent in administrative databases. Among the benefits of ADBs are an enhanced ability to meet expectations of accountability, the ability to examine policy-relevant questions on a longitudinal as well as cross-sectional basis, and opportunities for replication and linkage to other datasets to provide a more comprehensive analysis of public child welfare issues [Drake & Jonson-Reid 1999]. Identified limitations include predetermined variables that limit the scope of relevant policy and practice research questions, incomplete or inaccurate recording, and possible bias [Drake & Jonson-Reid 1999; Lurie 1990]. These and other authors argue, however, that administrative databases are a key resource for the current and future analysis of policy and program relevant questions [Drake & Jonson-Reid 1999; Worrall 1991; Raybould & Coombes 1992]. To realize the potential for these administrative data systems, their use needs to be integrated into policy development, practice innovations, and the development of program outcomes. Given the potential limitations and opportunities inherent in the development and use of administrative databases, two factors can determine whether an administrative database will prove useful in policy development or public agency practice: the extent to which the ADB is integrated into the decisionmaking process, and the nature and quality of the underlying data contained within the ADB. If the data in the ADB are unreliable, or if the data are reliable but no one values or uses the information, it is unlikely that the system will have long-term viability. This article discusses issues associated with the use of one administrative database, including the nature and the quality of the data contained within this management information system, and the integration of administrative data in decisionmaking in a public child welfare agency setting. …