Using W-2 Administrative Records to Compare Performance Across the SIPP and SIPP-EHC: An Analysis of Income and Poverty

Ashley Edwards, U.S. Census Bureau

In an effort to reduce costs and improve data quality in the Survey of Income and Program Participation (SIPP), the Census Bureau has re-engineered the survey by implementing a longer retrospective reference period (one year instead of four months) and has adopted an event history calendar (EHC) to facilitate respondent recall over the longer reference period. Of focus in this paper is the impact of the SIPP redesign on the capture of jobs and employment earnings over a calendar year. This analysis will compare the performance of the 2010 and 2011 SIPP-EHC field test to the 2008 SIPP Panel, using administrative W-2 records as controls for reference years 2009 and 2010. Specifically, this paper will evaluate differences across survey designs when looking at the number of jobs captured, reported annual employment earnings, and the impact of substituting W-2 records for survey data when calculating annual poverty rates across surveys.