Online Program

Trend and BED Statistics Using Synthetic and Multiplicative Noise Models for Disclosure Limitation
Michael Buso, Bureau of Labor Statistics 
*Shail Butani, Bureau of Labor Statistics 
Ali Mushtaq, NORC 
Santanu Pramanik, NORC at the University of Chicago 

Keywords: disclosure limitation, confidentiality, multiplicative noise, synthetic models, small area estimation, skewed population, Business Employment Dynamics

The U.S. Bureau of Labor Statistics (BLS) and NORC at the University of Chicago have been jointly developing a new disclosure limitation methodology for the Quarterly Census of Employment and Wages (QCEW), building on the micro-level noise perturbation ideas of Evans, Zayatz, and Slanta (EZS). One aspect of the new methodology uses synthetic models based on small area estimation techniques. The data from QCEW program serve many purposes. One is creating longitudinal files for the purpose of measuring and analyzing trends over time. Related to this is the collection of Business Employment Dynamics (BED) statistics for purposes of studying employment fluctuations by establishment status (opening, expanding, contracting, or closing). In this paper, we discuss alternative perturbation methodologies considered for QCEW given the trend and BED uses of the data, for the highly skewed QCEW population. We offer some possible solutions, now under study.