Geographic Variation in Medicaid Participation among Low-Income Medicare Beneficiaries

Even though Medicaid provides financial assistance for Medicare premiums and cost-sharing, many low-income Medicare beneficiaries who are eligible for these benefits do not enroll. Potential barriers include beneficiaries’ lack of awareness about Medicaid benefits and burdensome application procedures. To increase Medicaid participation, the Social Security Administration (SSA) sends letters annually to low-income Medicare beneficiaries who are potentially eligible but not enrolled for Medicaid. Based on aggregate data about the number of letters sent in each zip code, we examined how Medicaid participation varied across the country in 2016. For each state and county, we estimated take-up of Medicaid benefits among low-income beneficiaries by comparing the number of potentially eligible Medicaid beneficiaries to the number of dual-eligible beneficiaries. These measures were based on dual-eligible participants who qualified for Medicaid due to income. SSA sends different letters to potentially Medicaid-eligible beneficiaries depending upon whether they already participate in the Part D Low Income Subsidy program (Part D LIS), which allowed us to estimate participation in Part D LIS among beneficiaries who might be eligible for Medicaid.

Nationally, SSA notified 2.3 million Medicare beneficiaries that they were potentially eligible for Medicaid benefits. Relative to the number of beneficiaries already enrolled in Medicaid, this figure suggests that 79% of eligible low-income beneficiaries participate in the Medicaid program. By state, this percentage ranged from a low of 46% in North Dakota to a high of 90% in the District of Columbia. Eight of the ten states with the lowest Medicaid participation rates were located in the Midwest or West.

About one quarter of beneficiaries notified about their potential Medicaid eligibility already participated in the Part D LIS program. This measure varied across counties with the lowest versus the highest Medicaid participation rates. In counties in the lowest decile for Medicaid participation (less than 66% take-up) 19% of letter recipients already had Part D LIS benefits. In contrast, 30% of letter recipients already had Part D LIS benefits in counties with the highest participation rates (more than 87% take-up). According to Census data, counties with the lowest Medicaid participation rates were more likely to be rural, had lower poverty rates, and had a higher share of residents who were white or completed high school.

These estimates are imprecise due to estimating Medicaid eligibility based solely on income, not assets. Furthermore, this approach might overestimate Medicaid participation in states that have expanded Medicaid eligibility beyond the federal minimum guidelines for Medicare beneficiaries. Nevertheless, these results provide detailed geographic information about Medicaid participation and suggest a significant number of low-income Medicare beneficiaries are not accessing Medicaid benefits. Initiatives that target areas with low participation rates might provide better health care access and financial benefits to low-income Medicare beneficiaries.