Developing a Data Enclave for Sensitive Microdata
Norman Bradburn, Randy Horton, Julia Lane, Michael Tilkin
Norman Bradburn
Senior Fellow Education & Child Development Department
Randy Horton
Director of Development, Technology Services Information Technology Department
Julia Lane
Senior Vice President Economics, Labor, and Population Studies Department
Michael Tilkin
Senior Vice President and Chief Information Officer Information Technology Department
Email address of corresponding author: lane-julia@norc.uchicago.edu
US federal agencies and departments disseminate statistical data to external researchers for a number of reasons. These include allowing external researchers to conduct analyses in areas of interest to the agencies, to identify data quality issues, or to find new and innovative research and educational uses for existing datasets without further burdening respondents or increasing agency costs.
To disseminate public use data, federal agencies can use a number of established commercial and academic archives, such as the Inter-University Consortium for Political and Social Research at the University of Michigan. However, there is a more limited range of options available to federal entities seeking to disseminate sensitive microdata that have not been fully de-identified for public use. The largest federal statistical agencies (e.g. Census Bureau and Bureau of Labor Statistics) have sufficient economies of scale to develop advanced in-house solutions that serve the needs of external researchers. Smaller agencies often lack the resources to archive, curate, and disseminate the datasets that they have collected.
This paper describes an approach to developing a statistical data enclave. The core features of such an data enclave are the curation, indexing and archiving of microdata; the provision of researcher access to sensitive microdata; statistical protection; researcher training; dissemination of information to the research community; the simultaneous support of a wide range of agency-specific data requirements; and the ongoing interaction with the federal consortium.
