This article examines why the Current Population Survey (CPS) estimates of Medicaid enrollment are significantly lower than actual Medicaid enrollment counts. Inaccurate Medicaid enrollment estimates pose a problem since data from the CPS is frequently used in health policy and research.
The authors linked data from the 2000–2002 Medicaid Statistical Information System and the 2001–2002 Current Population Survey. They analyzed the frequency of incorrect answers to the CPS question on Medicaid enrollment and frequency of imperfect concept alignment.
The CPS survey results are a poor measure of Medicaid enrollment, largely due to Medicaid enrollees not responding accurately to questions about their enrollment status.