More Medicare Advantage Enrollment, More Duals? Enrollment in Medicare Advantage is Correlated with Increased Medicaid Take Up

Before 2018, Medicare’s risk adjustment model increased payments to Medicare Advantage (MA) plans for beneficiaries who were dually enrolled in Medicare and Medicaid (dual-enrollees) by 10 to 20 percent to compensate for those beneficiaries’ higher-than-average health costs. Conditional on health status, risk adjustment also compensated plans more for some beneficiaries than others; namely, risk adjustment over-compensated for dual-enrollees who were eligible for partial Medicaid benefits (partial duals) by 10 percent and undercompensated for dual-enrollees eligible for full benefits (full duals) by 10 percent. Additional risk adjustment payments may have given MA plans an incentive to help beneficiaries who were eligible for Medicaid enroll in that program. Given that as many as half of all Medicare beneficiaries who are eligible for Medicaid are not enrolled, MA plans may have significant scope to enroll additional beneficiaries in Medicaid (Medicaid Payment Advisory Commission, 2017.)

We test whether Medicare beneficiaries are more likely to take up Medicaid after enrolling in Medicare Advantage using detailed administrative data on a panel of 7.6 million Medicare beneficiaries for the years 2007-2016. Controlling for demographic and health characteristics, we find that beneficiaries who join MA are 33 percent more likely to become dually enrolled than beneficiaries who remain in FFS, despite being 13 to 33 percent less likely to become dual in the years prior to joining MA. This effect occurs quickly, with beneficiaries being 50-60 percent more likely to enroll in Medicaid in the first two years after joining MA. The likelihood that MA beneficiaries become dual falls thereafter. We also find that increased Medicaid enrollment among MA beneficiaries is primarily driven by partial duals rather than full duals.

We further test whether MA enrollees’ take up of Medicaid is plausibly linked to plan’s financial incentives by testing whether enrollees are more likely to take up full Medicaid in states where full duals may be healthier -- because income and asset eligibility thresholds for full Medicaid are higher. We find both that full duals in states with more generous eligibility standards for Medicaid have lower spending than full duals in other states, and that MA enrollees in those more generous states are twice as likely to take up full Medicaid benefits as MA enrollees in other states.

Taken together, our results suggest that MA enrollment is associated with an increased likelihood of subsequent dual-enrollment, and that dual enrollment appears to be related to beneficiaries’ profitability relative to their risk scores. Although a new risk adjustment system was implemented in 2018 in order to reduce the differential attractiveness of partial duals as compared to full duals, eligibility criteria for Medicaid benefits continue to differ across states. Whether risk adjustment updates will change plans’ incentives to enroll dual-eligible beneficiaries or help them take up Medicaid remains a matter for future research.