Predictors of Medicaid Reporting Instability and Implications for Examining the Medicaid Unwinding

Written by:
Working Paper Number: SEHSD-WP2025-08

Introduction

Medicaid coverage numbers have decreased since April 2023, when the federal government ended the continuous coverage requirement that prohibited states from disenrolling people during the COVID pandemic. Surveys are a potential data source for examining monthly coverage changes; however, limited information is available on the reliability of monthly Medicaid survey measures. Each year, the Annual Social and Economic Supplement of the Current Population Survey (CPS ASEC) sample includes individuals who were also interviewed in the previous year’s survey. This study examines individual-level instability in Medicaid reporting at the monthly level using matched records for 46,500 individuals in the 2023-2024 CPS ASEC. Individuals were categorized as having unstable Medicaid reporting at the monthly level if they (1) reported current Medicaid in their 2023 interview month but were not recorded as having Medicaid for that month in the 2024 interview or (2) did not report current Medicaid at their 2023 interview month but were recorded as having Medicaid for that month in the 2024 interview. Multivariable logistic regression was used to identify variables associated with instability in monthly reports of Medicaid. In the matched sample, 13.6% of individuals had unstable Medicaid reporting for the 2023 interview month. Unstable Medicaid reporting was associated with demographic, economic, and health-related characteristics, including serious difficulty with concentration or memory and poor health status (self- or proxy-reported). Unstable Medicaid reporting at the monthly level varied by individual characteristics, suggesting that these survey data may not be appropriate for examining monthly Medicaid coverage changes. 

Page Last Revised - June 17, 2025