Many Medicaid expansion enrollees with incomes less than 138% of the federal poverty level (FPL) face
greater cost-sharing and higher out-of-pocket costs when they tum 65 years old and transition to Medicare. In
Medicare, there is a subsidy cliff at 100% FPL, above which beneficiaries are not eligible for comprehensive
subsidies that cover medical cost-sharing in Medicare Parts A (inpatient) and B (outpatient). Beneficiaries with
incomes between 100-150% FPL and limited assets can qualify for partial financial assistance through a
complex set of programs. Take-up of these subsidies is limited, and evidence suggests that Black and
Hispanic beneficiaries are more likely to be eligible for but less likely to be aware of subsidy programs vs.
White beneficiaries. Moreover, higher cost-sharing has been associated with reductions in necessary care and
worse outcomes, and in some cases, widening of racial/ethnic disparities in care. Sixteen states have
expanded income and asset eligibility for comprehensive cost-sharing subsidies above federal minimums, but
there is little evidence on whether such policies mitigate racial/ethnic disparities. In Aims 1 and 2, we will use
linked, individual-level Medicaid and Medicare enrollment, claims, and encounter data from 2016-2023 to
follow beneficiaries as they transition from Medicaid to Medicare at age 65 and examine whether the transition
between programs leads to disparities in the receipt of Medicare subsidies and contributes to disparities in
use, clinical event outcomes, and spending. For Aim 1, we will examine if there are racial/ethnic disparities in
Medicare coverage among Medicaid expansion beneficiaries that transition to Medicare-only and use machine
learning approaches to identify factors (e.g., policy, health care system, individual) that predict receipt of
subsidies among minority vs. White beneficiaries. For Aim 2, we will assess if increases in cost-sharing upon
entry to Medicare contribute to racial/ethnic disparities in service use (e.g., outpatient visits), clinical event
outcomes (e.g., emergency department visits, hospitalizations, mortality), and spending (out-of-pocket and
total costs). We will use a difference-in-difference approach to compare changes in outcomes for beneficiaries
just above vs. below income limits for comprehensive cost-sharing subsidies, after vs. before entering
Medicare. For Aim 3, we focus on a subset of states that expanded income or asset eligibility for Medicare
subsidies above federal minimums in 2008 or later to assess if these policy changes reduce racial/ethnic
disparities associated with Medicare cost-sharing. We will use 2006-2023 Medicare data and a difference-indifference
approach to examine changes in outcomes for beneficiaries who gain comprehensive cost-sharing
subsidies before vs. after state subsidy expansions. In all aims, we will examine whether effects differ for Black
and Hispanic vs. White beneficiaries. Findings from this study will inform coverage policies that could reduce
adverse outcomes and racial/ethnic disparities in care as beneficiaries transition across these two public
programs.
