ABSTRACT
The current policy emphasis on data-driven decision-making is creating the right incentives for government agencies around the world that have not traditionally disseminated their administrative data to do so. The literature on statistical disclosure control focuses on the technical aspects of a variety of methods designed to protect data confidentiality. There is, however, a void in the literature in regard to what other elements are necessary to create and sustain a successful initiative. This paper examines six case studies of individual-level datasets. It reviews current practice in several domains and summarizes recommendations from expert practitioners including challenges for future initiatives.
Acknowledgments
We are grateful to all the individuals who graciously shared their time and expertise and who provided most of the data and information reported here. Mark Boward led the initial stages of this investigation.
Notes
1 Available at http://www.fcsm.gov/committees/cdac/cdac.html.
2 Generalized Micro-Agglomeration, Substitution, Subsampling, and Calibration.
3 NHES includes geographic indicators to the Census’s level four region. Some Census products present information at the Public Use Microdata Area (PUMA) defined after a minimum threshold population of 100,000.
4 A similar result was found by Felso et al. (Citation2001) in their review.