Modernizing Cell Suppression Software at the U.S. Census Bureau

• Activity Number: 455
• Type: Topic Contributed
• Date/Time: Wednesday, August 3, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Paul B. Massell*+
• Companies: U.S. Census Bureau
• Keywords: Cell Suppression ; Sensitive Cells ; Protection at the Company Level

Abstract

An Economic Census of the United States is conducted every five years by the economic directorate of the U.S. Census Bureau. The main data products are additive magnitude data tables that typically involve NAICS categories as rows and geographic entities as columns (some tables have a 3rd dimension). The p% rule is used for determining which cells are sensitive, and how much protection each such cell requires. First, the sensitive cells are suppressed. Then a cell suppression program is run against a file with information about each cell including an identifier for each establishment that contributes to the cell value; its associated company, and the contributed value. This program calls an optimization routine for each sensitive cell in order to find the optimal set of additional cells that must be suppressed in order to find the set of cells with the minimum total value, that, when suppressed, lead to a protection of the sensitive cells at the company level. We discuss a number of complex aspects of the software and how each of these was modernized.