The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Evaluating Welfare Reform in an Era of Transition
coordination with states, both in developing the data and in making them more comparable across states. Neither of the relevant agencies within DHHS (ACF and ASPE) can devote the kind of resources needed and still fulfill their own missions. The need for a more comprehensive, long-term effort to develop administrative data sources is another factor that leads the panel to propose the establishment of an authority within DHHS that is responsible for social welfare program data collection (discussed in Chapter 6).
Any multistate monitoring or evaluation study must confront the issue of comparability of administrative reporting. This is a special issue for TANF program administrative data since the programs and administrative structures for operating the programs now vary widely across states. Some state TANF administrative data sets contain certain types of recipients while other states contain other types of recipients. For example, Wisconsin has moved administration of child-only TANF cases to its child welfare system so these cases would not be part of the TANF data sets, unlike other states. Definitions of a case and family unit also vary across states. For example, some states classify cases in which the adult has been sanctioned as child-only cases, while other states consider children receiving TANF benefits that do not live with a parent as child-only cases. States are providing a wide variety of services to beneficiaries and, as the population being served changes, these services are evolving. As a result, there are few standard definitions of services or even types of services. The services provided under job search assistance in one state may not be classified the same way in another state. All these differences need to be well understood and reconciled if data from multiple states are to be used for research purposes.
Conclusion 5.3 The lack of cross-state comparability is a major barrier to the use of state-level administrative data sets for cross-state monitoring and evaluation.
The panel concludes that more can and should be done to improve the cross-state comparability of administrative data sets if these data are to reach their full potential. These improvements should move toward a common set of definitions of services and service units, which will not be an easy task. However, as the programs become more stable, it should be easier to identify broad types of services that can be defined and to implement common types of service units. Child-only cases are a good example of those for which a standard definition and service unit could be created.
Improvements in common data formats must also be made, including updated systems for storing and managing administrative data. Systems vary significantly across states and agencies. For example, some records are still kept on paper and so are far from being ready for research use, while others are readily