The Challenges of Preparing Sensitive Data for Public Release

Survey piactitioners arechallenged to meet he ever rising demand for microdata fdes while protecting the confidentiality of the individual data provider. Data pmcessing tools are becoming quite sophisticated which is an aid to survey practitioners, but also a potential tool to data snoops. Additionally, there is a percqtioo that participation in surveys is declining world wide; partially due to a rising coucern about cout%lent.ia.lity. Thus, survey practitioners must thoroughly protect the identity of the individual respondents. However, it is also important to retain the usefulness of the original dataandforinferencesmadefranmaskeddatatobeno stronger than those made from the original data. This paper details the preparation of the public release data file for the Federal Reserve Board’s Survey of Consumer Finances (SCF), a triennial household survey that includes data on finances, employment, and demogtaphics. We detail our experiences from the 1989 and 1992 surveys. Including this inttoduction, there are six sections. A brief summary of the literature on disclosure methodology is provided in the second section. Next, we describe the SCF. the sample design, data collected, and disclosure issues. ’ In tbe fourth section, we detail the disclosure suategy currently used in the SCF. The effects of the disclosure adjustments on selected estimates are presented next. We summarize our experience and discuss future plans in the last section.