Using Publically Available Administrative Data to Improve Direct Estimates of Income and Poverty from the American Community Survey

• Activity Number: 418
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Richard Griffin*+
• Companies: US Census Bureau
• Keywords: Borrow Strength ; Multivariate Regression ; Measurement Error

Abstract

The American Community Survey (ACS) produces direct five-year estimates at the census tract level for income and poverty. Small area estimation using models that borrow strength from relationships between variables across geographic areas may improve the accuracy of these estimates. Typically, these approaches combine direct estimates with model estimates that make use of administrative data. A 2012 pilot study used simulated administrative data to demonstrate the potential gain in accuracy from using three model based estimation methods. The Longitudinal Employee-Household Dynamics (LEHD) Origin-Destination Employment Statistics files constitute a publically available source of administrative data potentially correlated with income or poverty. This paper uses the LEHD files and published ACS data to produce tract-level estimates and estimated mean squared errors using each of these three model-based estimation methods. Results are compared with the sampling variance of the published ACS estimates, which are assumed unbiased.