Synthetic Data and Confidentiality Protection:Technical paper

摘要

The creation of demographic public use micro-data files fuelled a scientific and policy revolution. Restricted access to business micro-data in the early 1990's sparked a similar sea-change in the quality of analytical and policy work. In both cases a wide range of social scientists developed new empirical insights into social behavior, policy makers were able to base decisions on high quality statistical analysis, and entire generations of students grew up with the statistical and analytical tools that only access to micro-data can bestow. Even more compellingly, this massive social benefit was largely achieved simply by disclosure protecting already existing datasets--with little additional respondent or taxpayer burden. It is now apparent, however, that new challenges threaten the ability of national statistical institutes (NSIs) to release high quality public use data files.