Evaluating a Constrained Hotdeck to Perturb American Community Survey Data for the Census Transportation Planning Products

• Activity Number: 606
• Type: Topic Contributed
• Date/Time: Thursday, August 4, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Government Statistics
• Author(s): Tom Krenzke*+ and Jianzhu Li and David Judkins and Michael D. Larsen
• Companies: Westat Inc. and Westat Inc. and Westat Inc. and The George Washington University
• Keywords: Synthetic data ; semi-parametric imputation ; model-based imputation ; data utility ; disclosure limitation ; confidentiality

Abstract

The Census Transportation Planning Products (CTPP) contains tables for 150,000 travel analysis zones (TAZs), other geographies and journey to work flows, which results in millions of tables involving dozens of variables. The data source is moving from the Census Long Form to the smaller American Community Survey (ACS) five-year sample. Under cell suppression, data loss at the TAZ level would be substantial on ACS data using Census Bureau Disclosure Review Board (DRB) disclosure rules. For this reason, initial research has been conducted to perturb ACS values in a feasible way that retains the usability of the data, satisfies the data users' analytical needs, and pass the DRB rules. Also, even though the CTPP will be available in tabular form only, it was necessary to consider the confidentiality protection of the ACS public use microdata. We report on an evaluation of three perturbation