- Issue: An estimated 40 percent of low-income Medicare beneficiaries spend 20 percent or more of their incomes on premiums and health care costs. Low-income beneficiaries with multiple chronic conditions or high need are at particular risk of financial hardship. High cost burdens reflect Medicare premiums and cost-sharing, gaps in benefits, and limited assistance. Existing policies to help people with low incomes are fragmented — meaning that beneficiaries apply separately, sometimes to different offices — and require Medicare beneficiaries to navigate complex applications.
- Goals: With the goal of enhancing access and affordability for people vulnerable due to low incomes and poor health, this issue brief proposes a policy that would reduce Medicare’s cost-sharing and premiums for beneficiaries with incomes below 150 percent of the federal poverty level.
- Methods: Profile current cost burdens by income groups and assess the potential impact of a policy to expand cost-sharing and premium assistance using the 2012 Medicare Current Beneficiary Survey projected
- Results and Conclusion: The policy described could help 8.1 million low-income beneficiaries, significantly lowering their risk of high cost burdens. It also could also simplify the administration of assistance provided to these enrollees.
Although Medicare provides a stable, trusted source of health insurance for elderly and disabled beneficiaries, the program’s benefit design can leave beneficiaries exposed to high out-of-pocket costs. These include high deductibles for hospital care and 20 percent cost-sharing for physician visits and other ambulatory care services. In addition, there is no limit on out-of-pocket costs for covered benefits. Core benefits exclude dental, vision, hearing, and long-term care services and support.
Approximately 18 million Medicare beneficiaries live on annual incomes below 150 percent of the federal poverty level — less than $18,000 a year for a single person (Appendix 1).1 For people with low incomes, Medicare’s Part B annual premium or hospital deductibles represent a high share of income, let alone the costs of physician services, medications, or uncovered services. Although very low-income people may be eligible for Medicaid to supplement Medicare, one-third of Medicare’s poor (i.e., poor is defined as income at 100% of poverty or less than $12,120 if single) are not on Medicaid, and there is only limited help for the near-poor unless they incur high expenses and exhaust their savings. Of those beneficiaries with incomes below 150 percent of poverty, less than half (43%) receive Medicaid assistance with premiums and cost-sharing. Nearly half at this income level spend 20 percent or more of their income on premiums and health care each year. Prior research has shown broad evidence of unmet care needs as well as financial hardship.2
With a goal of enhancing access and affordability, this issue brief describes an approach to reducing premiums and cost-sharing for Medicare’s low-income beneficiaries, specifically those with incomes below 150 percent of poverty. The policy has three core design elements:
- Medicare Cost-Sharing. Reduce Medicare cost-sharing for hospital, physician, and other services and implement a new limit on out-of-pocket costs for people with incomes up to 150 percent of poverty.
- Medicare Part B Premium. Provide full premium help for people with incomes up to 135 percent of poverty and partial subsidies to those with incomes up to 150 percent of poverty, using the Part D premium sliding-scale design.
- Simplify Administration. Administer cost-sharing and premium help through Medicare and eliminate the asset test used by states to determine Medicaid eligibility.
To provide a baseline for comparison, we profile current spending relative to income and describe the policies currently used to assist low-income Medicare beneficiaries. We then highlight an illustrative policy and estimate its impact on beneficiaries and Medicare costs if fully implemented in 2016. Spending estimates and the impact of policy provisions use the 2012 Medicare Current Beneficiary Survey projected to 2016 (see Methods).
Current Policies Leave Low-Income Medicare Beneficiaries at High Financial Risk
In 2016, 56 million people were covered by Medicare.3 One-third (18 million people) had incomes below 150 percent of the federal poverty level (less than $18,000 annually for an individual) with, at best, limited assets to last their lifetimes.4 For people with low incomes, paying Medicare’s Part B premium ($1,259 a year if on Medicare before 2016 or $1,463 if new to Medicare in 2016) or the hospital deductible ($1,288) or a 20 percent share of physician bills, puts them at risk of going without care or accumulating unaffordable bills. For people with incomes near poverty ($12,120 if single), the sum of the Part B premium, hospital, and medical deductibles amount to nearly 25 percent of income. (See Exhibit 2 and Appendix 1.)
Low-income provisions to help pay for Medicare premiums or cost-sharing for Medicare covered core services or for Part D prescription drugs are limited, fragmented, and complex. Beneficiaries with incomes below poverty may qualify for full Medicaid to supplement Medicare but only if they have meager savings. In 25 states, they must have income levels lower than 75 percent of poverty.5
Beneficiaries with incomes below 135 percent of poverty who are not eligible for full Medicaid may qualify for help with Medicare premiums and cost-sharing through Medicare Savings Programs (MSPs) (Appendix 3). All states are required to provide MSPs through Medicaid, with the federal government setting minimum income and asset standards. Beneficiaries with incomes up to poverty receive help with paying for Medicare premiums and cost-sharing. Those with incomes between 100 percent and 135 percent of poverty receive help with premiums only. In 2016, the federal asset standard for MSPs was $7,280 if single and $10,930 for a couple, not including allowance for burial. Four states have higher income limits, and 11 have higher asset limits than the federal minimums.6
Similarly, Medicare Part D’s low-income Extra Help provisions assist in paying for prescription drugs and premiums for people with incomes up to 150 percent of poverty. Administered through the federal government, the program provides full subsidies below 135 percent of poverty and partial subsidies to 150 percent of poverty for those meeting asset tests. Beneficiaries apply separately to state Medicaid for MSPs and to Social Security for Medicare Part D Extra Help (Appendix 3).7
However, these existing provisions leave substantial gaps in protection for Medicare’s low-income beneficiaries. One-third with incomes below poverty and half of those with incomes below 150 percent of poverty did not have Medicaid in 2016 (Exhibit 1). Of those with any Medicaid coverage, one of six (12%) had only partial coverage for Medicare premiums.
Because there are no limits on out-of-pocket costs of core Medicare benefits, most higher-income beneficiaries seek supplemental coverage. However, as Exhibit 1 shows, low-income beneficiaries are the most at risk for having only Medicare and the least likely to be protected by employer-sponsored retiree coverage. Given the cost, few buy private Medigap supplements. Only about a third of those with incomes between 100 percent and 150 percent of poverty have any help from Medicaid.
We estimate that 41 percent of beneficiaries with incomes below 150 percent of poverty — 8.1 million people — spent 20 percent or more of income on premiums and health care in 2016 (Exhibit 2). This figure reflects the limits of current policies to help people with low incomes.
The high levels of health care spending are a result of both poor health and low income. We estimate that two-thirds of beneficiaries with incomes below 200 percent of poverty have three or more chronic conditions and/or serious physical or cognitive impairments (Appendix 4). Their out-of-pocket costs averaged more than $3,000 a year in 2016, rising to more than $6,000 for those with only Medicare coverage (Exhibit 3).
Policy Option to Enhance Financial Protection of Low-Income Medicare Beneficiaries
Using Medicare Part D as a guide, we outlined a policy that would reduce Medicare’s cost-sharing and premiums for beneficiaries with incomes up to 150 percent of poverty and eliminate asset tests for those not otherwise eligible for Medicaid or MSPs (Exhibit 4). Our policy would:
- Reduce Medicare Cost-Sharing. For beneficiaries with incomes up to 150 percent of poverty, Medicare would lower the hospital deductible to $100 per admission, reduce Part B cost-sharing to 10 percent from 20 percent, and set a new annual limit on out-of-pocket costs for Parts A and B services of $2,000. This approach would come close to the actuarial value and cost protection for people with incomes under 150 percent of poverty under the Affordable Care Act (ACA).8
- Extend Partial Premium Subsidies. MSPs currently provide a full Part B subsidy for beneficiaries with incomes up to 135 percent of poverty. This policy would extend premium help to 150 percent of poverty, using the Part D design that phases out premium help in three steps.
- Simplify Administration. Have Medicare administer the cost-sharing reduction and phased-premium reduction as it does for low-income assistance in Part D. Eliminating the asset test and using income alone to determine eligibility would also streamline administration. To identify those likely eligible for cost-sharing and premium help, Medicare could draw on data from Part D and sources used to determine premium surcharges for high-income beneficiaries.
Although this program could be administered through Medicaid, we envision using Medicare administration to reduce complexity for beneficiaries and to coordinate with Part D Extra Help. This approach avoids the excess administrative costs incurred when Medicaid pays and then bills the federal government for Medicare-related costs. For providers, Medicare administration would assure continued payment at Medicare payment levels for covered services.9
Impact of the Policies
In the analysis, we assumed that all beneficiaries with incomes below 150 percent of poverty would be eligible to participate, except for those with employer–retiree supplements. This policy follows the ACA subsidy provisions. In practice, relatively few poor or near-poor beneficiaries have such coverage. Moreover, those with employer benefits would be unlikely to drop such coverage.
We assumed full participation of all eligible beneficiaries who were not already on Medicaid with help paying Medicare premiums and cost-sharing. This could include some people who might already be eligible for Medicaid or MSPs but had not applied.
With this assumption, we estimate that reducing Medicare cost-sharing and expanding premium assistance up to 150 percent of poverty would help an estimated 8.1 million additional Medicare beneficiaries who do not currently receive full Medicaid assistance (Exhibit 5). This estimate includes 1.6 million beneficiaries with no supplemental coverage and 1.6 million with Medicaid assistance for Medicare premiums but not cost-sharing. The rest include individuals with Medigap or Medicare Advantage plans (Appendix 5). As Exhibit 5 details, the new participants would equal 45 percent of the 18 million beneficiaries with incomes below 150 percent of poverty. Combined with existing Medicaid and MSP enrollees, nearly all beneficiaries with incomes below 150 percent of poverty would have financial protection. An unknown share of the new participants might already be eligible for help; others might be excluded from Medicaid or MSPs due to asset tests.
Assuming all income-eligible beneficiaries participate and those with Medigap drop their plans, the policy could reduce beneficiary expenses for premiums and care by an estimated $11.7 billion a year.
The policy could reduce the share of low-income beneficiaries with high-cost burdens as a share of their income from 40 percent to 30 percent (Exhibit 6). This means an estimated 2 million fewer people each year would be exposed to such high costs. Because beneficiaries would continue to face costs of uncovered services and continue to pay a share of costs for benefits, their burden would still remain relatively high.
Federal costs for expanded assistance would be an estimated $8.4 billion, assuming full participation of eligible people who are not already receiving help from Medicaid or in employer–retiree plans. We also assumed that premium subsidies would apply to Medicare Advantage as well as traditional Medicare beneficiaries (see Methods). The $8.4 billion federal cost includes $6.4 billion for premium assistance and $2 billion for reduced cost-sharing.
Medicare’s current benefit design and limited provisions for poorer adults puts low-income beneficiaries — particularly those with chronic diseases or debilitating cognitive and physical impairments — at risk for incurring high costs and going without needed care. Concerns around affordability do not solely affect those at or near the poverty level and require a response that addresses premiums and costs of care.
In recognition of affordability concerns, the Medicare Payment Advisory Commission (MedPAC) in 2008 recommended extending help paying for Medicare’s premium to people with incomes up to 150 percent of poverty. Congress did not act on the recommendation. The recommendation suggested that Social Security administer this enhanced support along with Part D.10 This incremental step would not address the open-ended cost-sharing Medicare beneficiaries face.
Reducing Medicare premiums and cost-sharing for those with incomes up to 150 percent of poverty would begin to ensure more adequate financial protection. Doing so would also ease transitions as low-income adults become eligible for Medicare to avoid the loss of ACA subsidies that currently are substantially higher for those under age 65 than Medicare low-income provisions. The policy could be crafted in a way that aligns with Part D provisions to streamline application and administration.
Such expansions would require new federal expenditures. Even if only half of the eligible population participated, costs would exceed $4 billion. The policy could be phased in to lower initial federal costs, starting either with premiums or cost-sharing.
There are a number of ways to finance enhanced affordability. The Congressional Budget Office lists several revenue options that could be earmarked to finance the expansion: cigarette taxes ($3.7 billion a year), alcohol taxes ($5 billion a year), and savings on prescription medications ($7 billion a year). A value-added tax or tax on luxury goods, like yachts and high-priced cars, could also raise substantial revenues.11 By simplifying administration, the policy could enable federal and state programs to redeploy resources currently devoted to paperwork to services.
Even with such a policy in place, low-income beneficiaries would remain at risk for services not covered by Medicare. This includes the population of disabled adults who are unable to afford long-term services and supports that could enable them to live independently in the community.12 But by limiting the financial pressure of paying for Medicare-related costs, the policy could help delay or postpone beneficiaries spending down into Medicaid. Helping low-income beneficiaries stretch their budgets without turning to Medicaid will become increasingly important as the next generation of workers age into Medicare, many with limited savings.13
When enacted, Medicare held out the promise that seniors would be financially protected with access to care based on need rather than income. Realizing this promise will require targeted efforts to enhance support of poor and near-poor beneficiaries. The policy option profiled in this brief offers a path forward.
Study Data and Methods
All estimates are based on analysis of the 2012 Medicare Current Beneficiary Survey (MCBS) with population and spending projected to 2016 based on the national health expenditure accounts. The 2012 MCBS includes 11,299 respondents with population weights representative of the entire Medicare population, including the disabled under-65 population and those primarily living in long-term care institutions. For 2016, the projected population was 56.1 million. The data brief displays results for the population-weighted data.
In the survey, beneficiaries report their health care spending, including on services not covered by Medicare, and premiums. In addition to beneficiary reports, the MCBS cost-and-use files include information about incurred liability for Medicare benefits and spending on Medicare premiums based on administrative data. The MCBS also includes information on Medicaid status regarding whether the beneficiary is eligible for full Medicaid, Medicaid only for Medicare cost-sharing and premiums, or Medicaid for Medicare premiums.
The database has a sufficiently robust sample to permit analysis of subgroups by income, coverage, and health. In the analysis, we grouped beneficiaries by income based on their reported annual income relative to the federal poverty level. We also grouped beneficiaries into one of five mutually exclusive insurance categories: Medicare-only, Medicaid, Medigap, employer-sponsored insurance (ESI), and Medicare Advantage. If beneficiaries had more than one source of supplemental coverage, we used the following hierarchy to assign them to a group: Medicaid, ESI if any ESI (and not Medicaid), Medicare Advantage (if not Medicaid or ESI), and Medigap.
Estimating the Impact of Extending Medicare’s Low-Income Subsidies to 150 Percent of Poverty
Using information from the MCBS “event” file, we estimated the change in patient liability of changing Medicare’s cost-sharing and implementing an out-of-pocket limit. For Medicare-covered services, such as hospital or physician visits, the MCBS indicates the total amount of the incurred expense, the amount paid by Medicare, and the patient liability for Medicare services (deductibles or co-insurance) for beneficiaries in traditional Medicare.
The event-level data enabled analysis of the shift to Medicare for changes in cost-sharing for Parts A and B services for each beneficiary. In the analysis, we calculated the federal cost of a shift to Medicare if Medicare were to cover a greater share of the bill. The event file also enables estimates of the impact on other payers — for example, the reduction in Medicaid if Medicare covers more of the costs.
We assumed utilization of care would stay constant. We referred to a study that indicated relatively little expenditure impact among the elderly for hospital or physicians as long as there continued to be a deductible and cost-sharing (i.e., if the person did not have first-dollar coverage). Based on this information, we decided not to use dynamic modeling with utilization changing as cost-sharing changed at the margin.14
In the analysis, we assumed that all beneficiaries with incomes below 150 percent of poverty would be eligible to participate except for those with employer–retiree supplements. This policy follows the ACA subsidy provisions. In practice, relatively few very low-income beneficiaries have such coverage. Moreover, those with generous employer benefits would be unlikely to drop such coverage.
We assumed individuals with incomes below 150 percent of poverty currently buying Medigap would drop their policies. Participation would lower their premium costs (Medicare Part B and Medigap) but expose them to somewhat higher cost-sharing than in Medigap plans. Although the impact would likely be small, use of Medicare services under the option might be somewhat lower for beneficiaries who previously had Medigap. Use might be somewhat higher for Medicare-only beneficiaries if changes affect use at the margin. For simplicity, we kept utilization constant.
We assumed that all beneficiaries with incomes below 150 percent of poverty who did not already have full Medicaid or partial Medicaid that paid for Medicare premiums and cost-sharing would participate. We assumed states would maintain their current Medicaid levels and would continue required Medicare Savings Programs meeting federal minimum standards. The handful of states that have MSP standards above federal minimums might lower their levels back to meet the federal levels. However, this would have only a small impact on the estimates as the federal government already matches states spending and only three states (Connecticut, Indiana, and Maine) and the District of Columbia significantly exceed the minimum income standards. The additional beneficiaries with assistance under the policy option include 8.1 million people who fall into the following categories: Medicare-only, Medicaid with only Medicare premium help, Medigap, and Medicare Advantage beneficiaries with incomes below 150 percent of poverty not already on Medicaid.
Federal costs would increase for expanded premium subsidies for Part B, including costs for those eligible for full premium subsidies but not now participating. The estimates include the costs of expanding premium subsides for beneficiaries in Medicare Advantage with incomes below 150 percent of poverty. Appendix 5 details current sources of coverage at poverty levels for those newly participating.
As noted above, the net impact on beneficiaries includes the premium savings that would occur from dropping Medigap.
1. Analysis of 2012 Medicare Current Beneficiary Survey projected to 2016. See Appendix 1 for population counts by poverty level.
2. Cathy Schoen, Karen Davis, and Amber Willink, Medicare Beneficiaries’ High Out-of-Pocket Costs: Cost Burdens by Income and Health Status (Commonwealth Fund, May 2017); and Amber Willink, Karen Davis, and Cathy Schoen, Risks for Nursing Home Placement and Medicaid Entry Among Older Medicare Beneficiaries with Physical or Cognitive Impairment (Commonwealth Fund, Oct. 2016).
3. Centers for Medicare and Medicaid Services, National Health Expenditure Projections, Tables 1 and 17 (CMS, updated June 2015).
4. Analysis of the Federal Reserve Board’s ongoing survey (every three years) finds the bottom half of retirees are less likely to have any pension income than earlier decades or than high-income retirees. And as a group the bottom half of the income distribution have median private financial assets of under $7,000. Sebastian Devlin-Foltz, Alice M. Henriques, and John E. Sabelhaus, “The Role of Social Security in Overall Retirement Resources: A Distributional Perspective,” FEDS Notes (Board of Governors of the Federal Reserve System, July 29, 2016). Only 25 percent of the bottom quintile income group receive any income from assets. Half of this low-income group received less than $150. Ke Bin Wu, Sources of Income for Older Americans, 2012 (AARP Public Policy Institute, Dec. 2013).
5. Molly O’Malley Watts, Elizabeth Cornachione, and MaryBeth Musumeci, Medicaid Eligibility for Seniors and People with Disabilities in 2015 (Henry J. Kaiser Family Foundation, Mar. 2016).
6. National Council on Aging, Medicare Savings Programs: A Profile of State Options (NCOA, May 2016); and updated table (March 2018).
8. Gary Claxton and Nirmita Panchal, Cost-Sharing Subsidies in Federal Marketplace Plans (Henry J. Kaiser Family Foundation, Feb. 2015). The ACA specifies an actuarial value of 94 percent for incomes from 100 percent to 135 percent poverty and 87 percent up to 150 percent poverty. Out-of-pocket limits can be no higher than $2,250 a year. This Claxton brief finds limits are much lower, averaging $881 for the near-poor and $1,700 for those with incomes between 150 percent and 200 percent of poverty. We used $2,000 for both groups for simplicity.
9. In contrast, when Medicaid pays for Medicare cost-sharing, it may pay at rates below Medicare allowable levels. Providers are required to accept such payments, and thus get paid less for a Medicare patient who is low-income with Medicaid than they would be paid for a similar patient if at somewhat higher income levels. This practice penalizes physicians in low-income communities and allows for inequities in Medicare payment. See Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (MedPAC, Mar. 2008).
10. MedPAC, Report to the Congress, 2008.
12. For a discussion of potential ways to begin to expand support for those at risk because of complex illness and functional impairment, see Amber Willink, Karen Davis, and Cathy Schoen, Improving Benefits and Integrating Care for Older Medicare Beneficiaries with Physical or Cognitive Impairment (Commonwealth Fund, Oct. 2016).
13. Monique Morrissey, The State of American Retirement: How 401(k)s Have Failed Most American Workers (Economic Policy Institute, Mar. 2016).
14. Christopher Hogan, Exploring the Effects of Secondary Coverage on Medicare Spending for the Elderly (MedPAC, Aug. 2014; submitted 2012; appendix added 2014). This study found that the group with low cost-sharing (not first-dollar) and Medicare only had similar total expenditures although the mix differed with low cost-sharing using more preventive care, office visits, and eye exams. Compared to those with first-dollar coverage, both groups (Medicare only and modest cost-sharing) had lower total expenditures, adjusting for health status and sociodemographic factors.