Abstract

January 30, 2006

Title: Synthetic Data and Confidentiality Protection Systems for the Census Bureau Data
Presented by: John Abowd, Director, Cornell Institute for Social and Economic Research

This lecture is co-sponsored by IDRE and by the UCLA Human Complex Systems Program, http://www.hcs.ucla.edu/

PowerPoint Presentation

Video Podcast

Abstract

John Abowd is the lead Principal Investigator for a multi-million-dollar National Science Foundation Information Technologies Research Grant. The goal of this grant is to implement new confidentiality protection systems for Census Bureau public use data products. The new system is based on the concept of synthetic data, which is what this lecture will highlight. The cornerstone of this scientific model is a tight feedback relationship between the synthetic micro-data products, which must be developed using the Census Bureau's secure research facilities in Suitland, MD and on the Research Data Center network but may be released for public use after disclosure review, and the underlying "gold standard" confidential micro-data, which must always remain under the Census Bureau's direct control. Abowd will discuss how this feedback relationship is used to develop the data synthesizers, test their analytic validity, and improve their performance. All of the lecture’s examples will come from current Census Bureau synthetic data projects.

Biography

John M. Abowd is the Edmund Ezra Day Professor of Industrial and Labor Relations at Cornell University, Director of the Cornell Institute for Social and Economic Research (CISER), Distinguished Senior Research Fellow at the United States Census Bureau, Research Associate at the National Bureau of Economic Research (NBER, Cambridge, MA), Research Affiliate at the Centre de Recherche en Economie et Statistique (CREST, Paris, France), and Research Fellow at IZA (Institute for Labor Economics, Bonn, Germany). He has taught and done research at Cornell University since 1987, including seven years on the faculty of the Johnson Graduate School of Management. Professor Abowd’s current research focuses on the creation and use of linked, longitudinal data on employees and employers.

In his work at the Census Bureau he provides scientific leadership for the Longitudinal Employer-Household Dynamics Program, which is creating research data integrating demographic surveys, economic surveys, and administrative data. His original and ongoing research on this subject uses data created at the Institut National de la Statistique et des Etudes Economiques (INSEE), the French national statistical institute. Prof. Abowd’s other research interests include statistical methods for confidentiality protection of micro data; international comparisons of labor market outcomes; executive compensation, again, with a focus on international comparisons; bargaining and other wage-setting institutions; and the econometric tools of labor market analysis.

Prof. Abowd is currently the Principal Investigator or Co-Principal Investigator for multiyear grants and contracts from the National Science Foundation, the National Institute on Aging, the Alfred P. Sloan Foundation, and the U.S. Census Bureau. He has published articles in the American Economic Review, Econometrica, the Review of Economics and Statistics, the Quarterly Journal of Economics, the Journal of the American Statistical Association, and other major economics and statistics journals. Prof. Abowd was on the faculty at Princeton University, the University of Chicago, and the Massachusetts Institute of Technology before coming to Cornell.