Statistical Disclosure Control of Sensitive Tabular Data by Complementary Cell Suppression- Myth & Reality

Currently, complementary cell suppression procedures are mostly used by statistical agencies to protect sensitive tabular data from disclosure. It is generally believed that the linear programming (LP) based complementary cell suppression procedures offer the best protection from wrongful disclosure of statistical information. In recent years LP-based automated audit procedures have been advocated and are being used to ensure the adequacy of protection offered by cell suppression patterns. LP-based lower and upper bounds for suppressed tabular cells are typically used to determine the adequacy of disclosure control measures. This paper identifies limitations of conclusions drawn using LP-based audit procedures. We utilize commonly used analytical procedures to demonstrate the relative ease with which statistical disclosure of sensitive tabular data could occur. We conclude by providing additional safeguard measures required to avoid such disclosures