An Analysis of the Representativeness of the Low-Income Population in the HRS and SIPP

Published: 2011
Project ID: UM11-16

Abstract

The HRS is a key data source used to analyze the health and economic status of the middle-aged and older population in the United States and many analyses are specifically interested in the low-income population that is eligible for various means-tested programs. However, the validity of studies that rely on the HRS depends on whether it accurately represents the size and composition of the low-income population. Results from a recent RAND study using HRS and SIPP survey data matched to SSA administrative records suggest the the HRS may underrepresent low-income individuals. This project will compare SSA administrative data matched to the HRS with population-wide SSA administrative records to more rigorously assess the representativeness of the low-income samples in the HRS. If biases exist, improved weights will be produced to allow researchers to generate more accurate estimates.