Evaluation of Improvements to the Imputation Routine for Health Insurance in the CPS ASEC

• Activity Number: 471
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Miche Boudreaux*+ and Joanna Turner and Michael Davern
• Companies: University of Minnesota and University of Minnesota and NORC at the The University of Chicago
• Keywords: CPS ASEC ; health insurance coverage ; imputation

Abstract

The Annual Social and Economic Supplement of the Current Population Survey (CPS ASEC) is a widely used source of data on health insurance coverage. About 10 percent of respondents do not answer any of the ASEC supplement questions and the entire supplement for these respondents is imputed. Davern et al. (2007) identified problems with the imputation of health insurance variables in the CPS ASEC. They found the full supplement imputations cases were less likely to have private coverage and more likely to be uninsured. This discrepancy may partially be caused by a misspecification of the hot-deck routine. Only members of a policy holder's nuclear family can be covered as dependents. This family restriction contrasts with the instrument itself. We collaborated with the U.S. Census Bureau and evaluated possible improvements to the imputation routine to fix the identified problems. The study team recommended switching the order of the hot-deck allocation matrices with public coverage first and removing the nuclear family restriction to the private coverage allocation routine. Using 2009 CPS ASEC data, making these modifications reduces the uninsured rate from 17.3% to 16.8%.