Research has consistently shown that health insurance coverage surveys underestimate the number of people enrolled in Medicaid and the extent of this “Medicaid undercount” varies greatly across surveys and states. Among these surveys that measure health insurance coverage, the American Community Survey (ACS) has been shown to have a smaller Medicaid undercount, which is one of the reasons that SHADAC typically relies on this survey to track state-level health insurance coverage.

The “Medicaid undercount” refers to the discrepancies that exist between survey estimates of enrollment in Medicaid and the number of enrollees that are actually reported in state and national administrative data.

With 2021 ACS data being given its normal, official release by the Census Bureau following COVID-19 pandemic-related disruptions, SHADAC is again using the ACS as our primary data source for monitoring state-level health insurance coverage. However, as we and other data users have begun analyzing the 2021 ACS data, it has become apparent that the extent of the ACS’ Medicaid undercount was larger in 2021 than in prior years.

Given the evidence of a greater Medicaid undercount in the 2021 ACS, it is important for users to understand the undercount in the ACS when interpreting coverage estimates. In this brief, we review the research regarding the Medicaid undercount in the ACS, provide estimates of how the undercount varies across states, discuss the impact of assigning single coverage for those with multiple sources on the Medicaid undercount in the ACS, and provide guidance on how to understand health insurance coverage data in the 2021 ACS.

Related SHADAC Resources

Medicaid Undercount Doubles, Likely Tied to Enrollee Misreporting of Coverage (Issue Brief)
Understanding the Undercount of Medicaid Enrollees in the 2020 CPS (SHADAC Blog)
Medicaid Expansion and the Medicaid Undercount in the ACS (Research Article)
Medicaid Undercount Project (Collaborative Study)