2025 SOA Health Meeting: Session 3A: State-Specific Enrollment Modeling - Understanding Disenrollment Trends and Population Dynamics -06.23.25- Recording

This presentation introduces a novel approach to analyzing state-specific Medicaid enrollment patterns using data from the Centers for Medicare & Medicaid Services (CMS). We model Medicaid populations as a fluid system, where enrollees are treated as 'liquid' flowing in and out of a tank, with some individuals exiting and later re-entering the program. By examining the relationship between gross disenrollment and net enrollment changes, we build simple models that incorporate parameters such as lag effects and recirculation rates. Our early findings show strong linear trends in some states, which we visualize using Tableau to explore data fits across states.
This approach provides a deeper understanding of Medicaid enrollment dynamics, improving forecasts and supporting policy decisions. The presentation will also highlight how this model can be expanded for more sophisticated analyses. Learning objectives include exploring a new method for modeling Medicaid enrollment dynamics, understanding the use of lag and recirculation parameters in forecasting, learning how Tableau can enhance data analysis and visualization, and identifying new rates of churn and other patterns in the post-unwinding era of Medicaid.