Accuracy of Medicaid reporting in the ACS: Preliminary results from linked data
The “Medicaid undercount” refers to the discrepancy between administrative counts of Medicaid enrollment and estimates from survey data. This working paper is a collaboration between SHADAC and the U.S. Census Bureau to expand previous research on the Medicaid undercount to the American Community Survey (ACS).
In this working paper we use a unique version of the ACS which has been merged to enrollment data from the Medicaid Statistical Information System (MSIS) to quantify the Medicaid undercount in the ACS. Preliminary results suggest that 22.9% of enrollees known to have Medicaid or expansion Children's Health Insurance Program (CHIP) from the MSIS are not coded as such in the ACS. We explore how the discrepancy varies across age, poverty level, and state of residence and discuss how the discrepancy impacts estimates of the uninsured. While differences in survey and question design make comparisons with other surveys challenging, the Medicaid undercount in the ACS is on par with previous results studying the undercount in other federal surveys.