Researchers from the University of Minnesota State Health Access Data Assistance Center (SHADAC) investigated why state and national household surveys often produce much lower estimates of Medicaid enrollment than Medicaid enrollment data collected by states. Determining why this discrepancy, known as the “Medicaid undercount,” exists will improve the data systems used to analyze insurance coverage and focus proposals to change health policy where they are needed.
To study the factors contributing to the Medicaid undercount, the researchers linked administrative records on Medicaid enrollment from 2000 to 2002 compiled by the Centers for Medicare & Medicaid Services (CMS) with survey data from the U.S. Census Bureau's Current Population Survey, the nation's main source for estimates of uninsurance rates.
Researchers found that discrepancies between the populations included in the Medicaid administrative records and the Current Population Survey (CPS) account in part for the Medicaid undercount. Adjusting the data to account for these discrepancies enabled researchers to reduce the Medicaid undercount in 2000 from 42 percent to 32 percent. They also found incorrect responses to Current Population Survey questions about Medicaid coverage that appear to be an even larger cause of the Medicaid undercount.
The findings suggest that Current Population Survey-based estimates of the uninsured and the population eligible but not enrolled are overstated. Therefore, policy-makers should consider whether some other data sources might better meet their need for insurance coverage estimates.