Partial Information Releases for Confidential Contingency Table Entries: Present and Future Research Efforts

Abstract

Tabular data have been a staple product for disseminating information derived from the confidential microdata that fuel social science research and inform policy decisions. This paper outlines recent results on disclosure risk assessment associated with the release of high-dimensional contingency tables, and discusses some related research problems. The main focus is the partial information release strategy, through which agencies can release relevant marginal and conditional tables along with the sample size instead of a full contingency table. The most recent approaches in this area combine statistical and operations research methodologies with tools from computational algebra.