Email questions to

PUBLICATIONS > Reports > An Inventory of Research Uses of Administrative Data in Social Services Programs in the United States 1998

An Inventory of Research Uses of Administrative Data in Social Services Programs in the United States 1998

The use of administrative data for policy research is substantial and growing around the country. The authors surveyed over 100 administrative data linking projects in 26 states. Linkages were most common within the public assistance programs (AFDC/TANF, Food Stamps, and Medicaid), but a majority of states had projects linking public assistance data to JOBS, UI earnings, or Child Support data. Around one-third of the states had projects linking public assistance data to child care, foster care, or Child Protective Services. Four-fifths of the states used outside researchers to do these projects, and about half of all the projects we identified were done outside state agencies. The vast majority of projects were one-time, but there is a small, and growing, trend toward ongoing efforts that link a number of programs. The “cross-program query-able system” is a goal of some states, but it remains to be seen how successful these projects will be. Although the majority of the linked research data sets described in the report have already been developed, many still remain under construction. This use of administrative data is certainly widespread, but it remains in its infancy.

Several factors contribute to the success of data linking projects, including system uniformity, system upgrades, administrative leadership, and experience of groups or organizations at working together. A centralized data linking effort facilitates this process. Many researchers attributed success in extracting and linking data to their prior experience with state agencies. States with a historic interest in evaluation and those with experience from the welfare demonstration projects tended to be more adept with data linking. States that upgraded or replaced their older systems are in a much better positions to link data. Finally, states that have invested in one-stop program delivery approaches have tended to invest in databases that allow for cross-program data linkages as these can facilitate smoother program operation by providing information for decision support at the caseworker level.
Further progress will require the implementation of better common identifiers, longer periods of data retention, and more consistent and flexible policies on confidentiality and sharing of data. The majority of administrative data systems in every program area except Unemployment Insurance generate their own identifiers rather than Social Security numbers as their primary method for identifying persons. Consequently, probabilistic matching techniques using name, age, sex, and other information are often necessary. Different definitions of case units across programs add substantial complexities to the matching task. Even when linking is possible, data storage and retention practices can hinder the development of a retrospective longitudinal database. Most states keep data active for less than five years, and archiving for more than ten years is unusual. These practices hinder the development of retrospective longitudinal databases; however, they would not affect prospective longitudinal database creation. Confidentiality barriers may be the toughest of all these difficulties because they may require legislative remedies and they involve legitimate concerns about protecting people’s privacy. Despite these difficulties, increasing technological capabilities and the desire of program managers to know what is happening in and across their programs strongly suggest that the use of administrative data will continue to expand.

Publication file


related publications

''Pencils Down!'' The Negative Effect of Welfare Reform on Opportunities for Higher Education
Health Status, Health Insurance, and Worker Mobility: a Study of Job Lock in California
The Poverty Research Programs of UC DATA
Recommendations for Research Priorities and Data Sources for Monitoring and Evaluating Welfare Reform
California Work Pays Demonstration Project, A Process Evaluation Phase IV: 1996
California's Immigrant Households and Public-Assistance Participation in the 1990s
Seasonal Employment Dynamics and Welfare Use in Agricultural and Rural California Counties
Latina/Latino Public Opinion Data Book
Animated Maps for USA: A Century of Difference
Supplementary Materials: A Century of Difference
Comparing Costs of Delivering Medical Benefits: Group Health and Workers’s Compensation
Unchurched Believers:Fewer Americans Have a Religion But Religious Beliefs Haven’t Changed Much
Data From the Census Bureau
Survey Documentation and Analysis (SDA)
The General Social Survey, 1972-2010
Census 2010: Redistricting Data and Beyond
Census 2010: Data Tools and Resources
Census 2010: Census Geographies
Census 2010: Issues in Data Across Time and Space
Census Data for Researchers
Foreclosure Data
Data and Statistics from International Governmental Organizations
National Longitudinal Study of Adolescent Health
GIS/Geoportal Services
Access to Data Collected by the Census Bureau
APLIC: Tools We Use: Survey Documentation & Analysis (SDA)
2008/12/18 15:41:29