A Report from Families USA
July, 1998

Shortchanged: Billions Withheld from Medicare Beneficiaries

INTRODUCTION

Congress created the Medicare buy-in program to protect low-income seniors and people with disabilities from the significant and growing costs required to receive Medicare coverage. Almost a decade after the enactment of the Medicare buy-in program, between 3.3 and 3.9 million of the people who are eligible for this protection are not receiving these critical benefits.

For these low-income senior citizens and disabled individuals, the Social Security Administration continues to deduct Medicare premiums of $43.80 each month for individuals, or $87.60 each month for couples, out of their Social Security checks, even though they are entitled to have those premiums paid by the government. Over the course of a year, this amounts to $525.60 per person, or $1,051.20 per couple, for a total of between 1.8 and 2 billion dollars in calendar year l998?funds that would not be deducted if they were receiving the benefits to which they are entitled by law.1 In addition, the federal government is withholding as much as $260 million in calendar year l998 from the Social Security checks of individuals who became eligible to apply for the Medicare buy-in as a result of the Balanced Budget Act of l997?but who are not receiving these benefits.

Today, senior citizens spend a higher proportion of their incomes for health care than they spent prior to the enactment of the Medicare program. Even with Medicare, out-of-pocket costs for senior citizens are higher today than they were in the early 1960s.2 Estimates show that Medicare beneficiaries spent $2,149, or 19 percent of their annual income, for out-of-pocket health costs in l997. For senior citizens at or below the poverty line, health spending consumed more than a third (35 percent) of their annual incomes.3

Congress enacted the Qualified Medicare Beneficiary Program in l988 to help defray the cost-sharing requirements of the Medicare program for low-income senior citizens and persons with disabilities.4 This program?commonly referred to as the Medicare buy-in?requires the Medicaid program to pay the Medicare premiums, deductibles, and copayments for senior citizens and persons with disabilities with incomes at or below 100 percent of the federal poverty guideline.5

Today, after a number of legislative changes, there are three major categories of buy-in programs6 for low-income senior citizens and persons with disabilities:

QMB: Those with incomes at or below the federal poverty guideline, known as Qualified Medicare Beneficiaries (QMBs), are eligible for financial assistance covering their Medicare premiums, deductibles, and copayments. The maximum incomes that a person and couple can have to be eligible for QMB protections are $8,292 and $11,100, respectively.

SLMB: Those with incomes between 100 and 120 percent of poverty ($9,900 and $13,260 in annual income, respectively, for individuals and couples), known as Specified Low-Income Medicare Beneficiaries (SLMBs), are eligible for assistance with their Medicare Part B premiums only.

Qualified Individuals (QI-1s): As of January 1, l998, senior citizens and persons with disabilities with incomes between 120 and 135 percent of poverty ($11,112 and $14,892 in annual income, respectively, for individuals and couples) are eligible through a block grant program to apply for payment of their Medicare Part B premiums. This block grant program serves potentially eligible persons on a first-come, first-served basis. Due to limited funds in the block grant, only a portion of those who are eligible to apply for the benefit will actually receive it.

For all three categories, the resource limit for eligibility is $4,000 for an individual and $6,000 for a couple.

KEY FINDINGS

This report presents the latest state-by-state estimates of the numbers of low-income senior citizens and disabled individuals who under federal law should have their Medicare premiums subsidized by the buy-in program?but, instead, are paying for those premiums through deductions in their Social Security checks.7 These are the low-income elderly and disabled individuals who qualify for the Medicare buy-in benefit but who are not receiving it in calendar year 1998.8

Nationally, between 3.3 and 3.9 million low-income senior citizens and disabled individuals are eligible for QMB and SLMB benefits but are not receiving them. Between 1.9 and 2.4 million low-income senior citizens and disabled individuals are eligible for, but not receiving, the QMB benefit and an estimated 1.4 million low-income senior citizens and disabled individuals are eligible for, but not receiving, the SLMB benefit.

Nationally, between 41.5 and 47.9 percent of the eight million low-income senior citizens and disabled individuals who are eligible to receive QMB and SLMB benefits are not receiving them.

Over the course of the year, the federal government is withholding between 1.8 and 2 billion dollars from the Social Security checks of these low-income senior citizens and disabled individuals?funds that would not be deducted if these beneficiaries were receiving the benefits to which they are entitled by law.

Nationally, there are an estimated 1.6 million senior citizens and disabled individuals who qualify to apply for the QI-1 buy-in benefit. As of June l998, only 4,723 individuals out of the 499,000 individuals who could be served by the program with the available funds for calendar year l998?less than 1 percent?were receiving buy-in benefits. Unless participation improves, $260 million will be deducted from the Social Security checks of low-income Medicare beneficiaries in 1998 who should be receiving the new benefits enacted through the Balanced Budget Act of l997.

The following 12 states have the highest numbers of poor seniors and disabled individuals eligible for QMB and SLMB benefits but not receiving them and, hence, are experiencing inappropriate reductions in their monthly Social Security checks: Texas (370,000 to 404,000); Florida (252,000 to 275,000); Ohio (233,000 to 264,000); Pennsylvania (202,000 to 231,000); Illinois (198,000 to 226,000); New York (148,000 to 192,000); Virginia ( 122,000 to 131,000); Michigan (103,000 to 118,000); Massachusetts (94,000 to 113,000); Washington (99,000 to 107,000); Alabama (92,000 to 100,000); and Georgia (88,000 to 103,000). (See Table 1.)

In the 12 states with the highest number of persons not receiving their buy-in entitlements, the following amounts are being deducted from such persons' Social Security benefits: Texas ($194 to $212 million); Florida ($132 to $145 million); Ohio ($122 to $139 million); Pennsylvania ($106 to $121million); Illinois ($104 to $119 million); New York ($78 to $101 million); Virginia ($64 to $69 million); Michigan ($54 to$ 62 million); Massachusetts ($49 to $59 million); Washington ($52 to $56 million); Alabama ($48 to $53 million ); and Georgia ($46 to $54 million). (See Table 2.)

In the following 12 states, the percentages of poor seniors and disabled individuals eligible for QMB and SLMB benefits but not receiving them are the highest: North Dakota (75 to 80.1 percent); New Hampshire (68.6 to 75.6 percent); Rhode Island (64.6 to 72.4 percent); Illinois (61 to 69.5 percent); Nevada (63 to 65.8 percent); Nebraska (60.1 to 68.6 percent); Delaware (61.5 to 66.8 percent); Ohio (59.4 to 67.1 percent); Arizona (59.6 to 63.3 percent); West Virginia (59.5 to 63.4 Percent); Pennsylvania (56.6 to 64.8 percent); and Maryland (56.8 to 64.2 percent). (See Table 1.)

Table 1
Low-Income Medicare Beneficiaries Eligible for, But Not Receiving, Buy-In

1 This column presents a high and low range rounded to the nearest 1,000 of QMB and SLMB eligibles not receiving the buy-in who, as a result, are experiencing deductions in their Social Security checks.
2 This column presents a high and low range percentage of QMB and SLMB eligibles not receiving the buy-in. The percentages given in this column are calculated from data which have not been rounded. As a result, they may not match percentages calculated from previous columns due to rounding error.
* We do not report for Alaska due to insufficient sample sizes.
** Less than 1,000.

Table 2
Annual Funds Lost By Low-Income Medicare Beneficiaries

1 Range of dollars lost by low-income Medicare beneficiaries who are experiencing Social Security deductions because they are not receiving their buy-in benefits.
* We do not report Alaska due to insufficient sample sizes.

** See last footnote on Table 1.

MEDICARE COST-SHARING

In l965, Congress enacted the Medicare program to help senior citizens pay their health care bills. Over the last three decades, Medicare has provided millions of senior citizens and people with disabilities with access to health care that they would not otherwise have had.

However, senior citizens today are spending a higher proportion of their incomes for health care than they were prior to the enactment of the Medicare program. Even with Medicare, out-of-pocket costs for senior citizens are higher today than they were in the early 1960s.9

It is estimated that Medicare beneficiaries spent $2,149, or 19 percent of their annual incomes, for out-of-pocket health costs in l997. For senior citizens at or below the poverty line, it is estimated that health spending consumed more than a third (35 percent) of their annual incomes.10

In l998, Medicare cost-sharing and payments for Part B premiums are substantial. These costs for an individual are:

the premium for physician and outpatient coverage (Part B of Medicare), currently $43.80 per month (or $525.60 per year);

the Part B deductible, $100 per year;

copayments of 20 percent for Medicare-approved physician charges above the Part B deductible;

a hospital (Part A) deductible of $764 for each hospitalization;

substantial copayments for any hospitalization in excess of 60 days; and

substantial copayments for skilled nursing care stays longer than 20 days.

Table 3
Increasing Burden of Medicare Cost-Sharing, 1980 and 1998

Source: 1998 Annual Report of the Board of Trustees of the Federal Hospital Insurance and Supplementary Medical Insurance Trust Funds.

a Average spent per beneficiary.

Medicare-related beneficiary costs have risen quite dramatically over the last three decades. Table 3 compares these costs for l980 and l998. In l998, the annual Part B premium had grown to $525.60?an increase of 403 percent from l980, when it was $104.40. The average per capita expenditure for Part B copayments was up 500 percent over this time period?from $85.52 in l980 to $513.38 in l998. These percentage increases in health care costs were considerably higher than the 98 percent increase in the cost-of-living between 1980 and l998.

Medicare cost-sharing will continue to increase over time. The changes in the Balanced Budget Act of l997 are projected to increase Medicare premiums to $105.70 per month in the year 2008. Although Medicare premiums were expected to increase over time, this is $46 more than the increase projected prior to passage of the Balanced Budget Act.11 This increase in the Medicare premium will place an even greater burden on low-income beneficiaries in the future.

In addition to these substantial and growing costs, Medicare beneficiaries are likely to have additional medical expenses because the program does not cover the costs of such services as prescription drugs; vision, hearing, and dental care; and long-term care. Moreover, the low-income elderly have greater health care needs and higher health care costs than those with higher incomes.12

LEGISLATIVE BACKGROUND

Congress enacted the buy-in protection in recent years to prevent the extreme financial hardship that Medicare cost-sharing requirements create for low-income beneficiaries. In the Medicare Catastrophic Coverage Act of l988 (MCCA), Congress required the Medicaid program, starting in l989, to pay the Medicare premiums, deductibles, and copayments for low-income senior citizens and persons with disabilities below 85 percent of the federal poverty guideline?and the eligibility level was to increase in five percent annual increments until it reached 100 percent of poverty. The federal government pays for most (approximately 55 percent) of the costs of the benefit, with the states providing the remainder based on the federal-state matching formula under Medicaid. Those eligible for the protection were called Qualified Medicare Beneficiaries (QMBs). While the MCCA eventually was repealed, the provisions relating to the buy-in were left in place.

Table 4
Coverage Available to Low-Income Medicare Beneficiaries

	Part B Premiums	Copays	Deductibles
Qualified Medicare Beneficiaries Living at or below the poverty line
Specified Low-Income Medicare Beneficiaries (between 100 and 120 Percent of Poverty)
Qualified Individual Beneficiaries (between 120 and 135 Percent of Poverty)

In the fall of l990, Congress increased Medicare cost-sharing amounts as part of its deficit-reduction package. It also added buy-in protections for Medicare beneficiaries with incomes up to 110 percent of the poverty guideline in l993 and 120 percent of the poverty guideline in l995. Persons eligible for this benefit are called Specified Low-Income Medicare Beneficiaries (SLMBs). Individuals with incomes between 100 and 120 percent of the poverty guideline are eligible for Medicaid payment of Medicare premiums, but not for Medicaid payment of other Medicare cost-sharing. The federal-state matching formula for subsidizing SLMB costs are identical with those for QMB beneficiaries.

The Balanced Budget Act of l997 added another category of persons who are eligible to apply for buy-in protection. As of January 1, l998, senior citizens and persons with disabilities with incomes between 120 and 135 percent of poverty are eligible through a block grant program to apply for subsidization of Medicare Part B premiums.13

This block grant program differs from the existing SLMB program in several ways. The program is a "capped entitlement"?meaning that not all eligible beneficiaries are guaranteed protection. The federal government pays 100 percent of the costs up to a cap for each of the next five years. The total funding for this block grant is limited to $200 million in fiscal year l998 and increases to $400 million in fiscal year 2002, after which this benefit will need to be reauthorized if it is to continue. This block grant program serves potentially eligible persons on a first-come, first-served basis. Due to limited funds, only a small portion of those who fall within the income and resource eligibility standards can be served.

Table 5
1998 Annual Income Eligibility for Buy-in Programs*

* All numbers for the income eligibility for buy-ins are $240 above the relevant poverty guideline due to a $240 annual unearned income disregard to which all applicants are entitled. The asset limits for all three programs are $4,000 for individuals and $6,000 for couples. A house, a car of limited value, and a few other resources are not counted toward the asset limit.

MEDICARE BUY-IN PARTICIPATION: WHY THERE IS A PROBLEM

The Medicare buy-in program was created to protect low-income Medicare beneficiaries from the financial hardships of cost-sharing requirements. However, almost a decade after enactment of the program, between 3.3 and 3.9 million of the low-income senior citizens and disabled individuals who are eligible for QMB and SLMB benefits are not receiving them. Between 1.9 and 2.4 million low-income senior citizens and disabled individuals are eligible for, but not receiving, the QMB benefit. An estimated 1.4 million low-income senior citizens and disabled individuals are eligible for, but not receiving, the SLMB benefit.

This means that, over the course of a year, the federal government is deducting $525.60 per person?or $1,051.20 per couple?out of the Social Security checks of low-income senior citizens and disabled individuals who are entitled to be spared from these costs. This amounts to between 1.8 and 2 billion dollars a year.14 In addition, many people who are entitled to, but not receiving, QMB benefits are paying substantial portions of their incomes on physician and hospital bills (i.e., Medicare deductibles and copayments) even though, by law, these individuals are not responsible for these costs.

Hundreds of thousands of additional low-income seniors and disabled individuals who are supposed to get buy-in protection through the block grant program established under the Balanced Budget Act of 1997 are not getting such benefits. Nationally, there are an estimated 1.6 million persons with incomes between 120 and 135 percent of poverty who potentially qualify for the new block grant buy-in program known as QI-1. Because the total funding available for this program is $262.5 million for calendar year 1998, only approximately 499,000 of these potentially eligible low-income persons are entitled to benefits in l998.15

As of June l998, only 4,723 individuals were enrolled in the QI-1 program?less than one percent of the population who could potentially be served by the program with the available block grant funds in calendar year l998. This means that the QI-1 program has not been reaching the additional 494,000 persons who are suppose to receive buy-in protection through this program.16 Based on this low enrollment, as much as $260 million will be deducted from the Social Security checks of low-income Medicare beneficiaries who should be receiving the new benefits.17

Other studies support the findings in this report. In December l996, the Urban Institute estimated that over 3.7 million Medicare beneficiaries who were eligible for the QMB and SLMB programs were not receiving these benefits.18 The Urban Institute's estimate for non-participation in the QMB program was 2 million. In l993, Families USA Foundation reported that 1.8 million senior citizens were eligible for, but not receiving, QMB benefits.19

A major reason for low participation in the Medicare buy-in program is lack of knowledge about the program?on the part of both beneficiaries and social service workers. Although Medicare beneficiaries must visit a Social Security office to enroll in Medicare, they are not allowed to apply for buy-in benefits at that office. Instead, they must make a separate trip to a welfare office. Senior citizens report how difficult it is to find someone in welfare offices or Social Security offices who knows about the buy-in program.20

Low-income senior citizens and disabled individuals must also overcome obstacles to apply for buy-in benefits. In many states, individuals must apply in person at the local welfare office. Even if the state allows applications by mail, the application form is difficult to understand and fill out without professional assistance. The applicant is also required to gather extensive documentation. These bureaucratic hurdles stymie low-income beneficiaries' ability to apply for the benefits to which they are entitled by law.

RECOMMENDATIONS

To guarantee access to entitled benefits, the Social Security Administration (SSA) and the Health Care Financing Administration (HCFA) in the Department of Health and Human Services must work together to ensure that eligible individuals know about Medicare buy-in benefits, can easily apply for them, and actually enroll in the program. To date, federal and state efforts to fully implement the Medicare buy-in program have been slow and insufficient.21 Only a concerted national effort spearheaded by SSA and HCFA will result in participation rates in all states close to 100 percent.

Authority exists in current law for both SSA and HCFA to take a more direct and active role in increasing participation in the Medicare buy-in program. Most of the recommendations described below are permissible under current law. However, the fact that enrollment levels have remained so low since the inception of the buy-in program suggests that legislative or administrative modifications may be required to actually make a difference in buy-in participation rates.

1. Notification: SSA and HCFA must assume responsibility for identifying and notifying beneficiaries eligible for the buy-in. Under existing legislative authority, SSA and HCFA can?and should?notify individuals about their potential eligibility for the buy-in at a number of different points. All notices sent to beneficiaries should include an explanation of the buy-in benefit and information about eligibility requirements and where to apply.

When individuals first phone or go to SSA to apply for Social Security and/or Medicare benefits, SSA should provide them with information, both orally and in writing, about the Medicare buy-in program.

SSA and HCFA should include prominently-placed information about the Medicare buy-in, along with the name and number of the appropriate state agency to contact, in the initial award letter sent to new beneficiaries informing them about their Social Security and Medicare benefits.

SSA and HCFA should include a well-placed description of the Medicare buy-in in the annual notice sent to all Social Security beneficiaries informing them about the annual cost-of-living increase in their Social Security benefits.

As required by law, HCFA should send an annual notice to all Medicare beneficiaries explaining the Medicare program that includes a well-placed description of the buy-in.

HCFA should send to all new Medicare beneficiaries with Social Security incomes up to the eligibility levels for the buy-in program a letter informing these individuals that they may be eligible for the Medicare buy-in program.

2. Streamline the application process: HCFA must ensure that a simplified application form is made available to potential buy-in participants. Key barriers to participation in the buy-in are the complexity and length of the application together with documentation requirements. In l993, Secretary Shalala of the Department of Health and Human Services advised then-Senator Donald Riegle that her Department was developing a simplified application form for the Medicare buy-in. Although the form was developed and sent to the states, no information is available regarding its use. HCFA should ensure that a simplified form is used by state offices. HCFA should also allow a grace period for eligibility errors that would allow states to enroll individuals immediately upon self-declaration of the truth of information provided in the application.

3. Change the effective date for QMB eligibility: Unlike most Medicaid benefits, QMB benefits are not retroactive. Rather, entitlement begins in the month after eligibility has been determined. Because Medicaid has 45 days to process an application, this delay can easily cause QMBs to lose buy-in benefits for a month or more. It is recommended that the effective date for eligibility for QMBs be changed to three months retroactive from the date of application. This would bring QMB eligibility in line with that of SLMBs and other Medicaid beneficiaries and prevent the undue hardships that QMBs experience when delays occur in eligibility. This would require a legislative change.

4. Directly enroll buy-in participants at Social Security offices: SSA should directly enroll Medicare beneficiaries into the buy-in program at Social Security offices. This single change would make the greatest difference in increasing participation in the Medicare buy-in program. Taking applications at Social Security offices and training Social Security personnel to assist Medicare beneficiaries with completing these applications would make it possible for potentially eligible persons to apply for the buy-in at the same time as they apply for Social Security and Medicare.

Social Security offices already have experience with enrolling beneficiaries in the Medicaid program. In 32 states, Social Security offices determine Medicaid eligibility for SSI beneficiaries.22 Under the same legislative authority (Section 1634 of the Social Security Act), SSA offices can determine eligibility for the Medicare buy-in, under an agreement with each state. However, to date, SSA and individual states have not developed such agreements. As a result, SSA is not directly enrolling potential beneficiaries in the buy-in program.

This suggests that legislative or administrative modifications are needed requiring that SSA work with HCFA to develop and implement a system for directly enrolling all potential Medicare beneficiaries in the buy-in program at Social Security offices. Such modifications would require SSA workers to tell potential applicants about the buy-in program, assist them with a simple application form, and then forward the applications to the appropriate state Medicaid office.

CONCLUSION

The financial protections in the Medicare buy-in program are crucial for the security of low-income Medicare beneficiaries. It is the federal government's obligation to ensure that low-income Medicare beneficiaries receive the financial protections that Congress intended when it enacted the Medicare buy-in program.

The Balanced Budget Act of l997 introduces a new level of urgency to the need for the federal government to take aggressive action to protect some of our nation's poorest citizens. As a result of this Act, the Medicare Part B premium is scheduled to increase more rapidly than it would have increased under prior law. Unless significant change occurs, the federal government will take an even larger bite out of the meager Social Security checks of low-income Medicare beneficiaries in the future.

Individuals eligible for buy-in protection can benefit greatly from the extra income they retain from this critical program. Low-income Medicare beneficiaries should not have to choose between food, shelter, and needed health care. The burden for QMB enrollment should shift from these individuals to the federal and state agencies administering the program. With a concerted effort by the SSA and HCFA, the promised protection can become a reality.

REFERENCES

1 This estimate of between 1.8 and 2 billion dollars for calendar year 1998 was determined by multiplying the number of low-income senior citizens and disabled individuals who were entitled to benefits as Qualified Medicare Beneficiaries (QMBs) and Specified Low-Income Medicare Beneficiaries (SLMBs) but who were not receiving benefits in March l998?between 3.3 and 3.9 million persons?by the annual amount for the Part B premium ($525.60).

2 Families USA Foundation, The Health Cost Squeeze on Older Americans (Washington, DC: Families USA Foundation, February l992); Special Committee on Aging, United States Senate, Medicare and the Health Costs of Older Americans: The Extent and Effects of Cost Sharing (Washington, DC: U.S. Government Printing Office, April l984).

3 AARP Public Policy Institute and The Lewin Group, Out-of-Pocket Spending by Medicare Beneficiaries Age 65 and Older: 1997 Projections (Washington, DC: American Association of Retired Persons, l997).

4 The Medicare buy-in program connects the two largest public health programs in the country?Medicare and Medicaid. Medicare is an entirely federal program that provides a standardized package of health services to the aged and certain disabled individuals. Eligibility is not based on income or resources. Medicaid is a combined federal-state program. Although general program requirements for Medicaid are dictated by federal law, Medicaid is operated and administered by the states. Eligibility for Medicaid is based on having income and resources below specified ceilings.

5 The annual income standard for the Medicare buy-in program is actually $240 above the federal poverty guideline because applicants are entitled to disregard $240 of unearned income annually.

6 Two additional buy-in programs exist that are not the subject of this report. Qualified Disabled and Working Individuals (QDWIs) are individuals with incomes up to 200 percent of the federal poverty guideline who are eligible to have the Medicaid program pay their Part A premium ($309 a month in l998). Individuals eligible for the second type of buy-in program are called Qualified Individuals-2s (QI-2s). These are Medicare beneficiaries with incomes between 135 and 175 percent of poverty who are eligible to have the Medicaid program pay a portion ($1.07 a month) of their Part B premium.

7 These estimates are presented as ranges rather than precise figures because it is not possible to determine the exact number of Medicaid beneficiaries whose incomes fall below the poverty guideline.

8 Although the buy-in program covers the eligible institutionalized population, these estimates exclude institutionalized persons.

9 Families USA Foundation, op. cit.; Special Committee on Aging, op. cit.

10 AARP Public Policy Institute and The Lewin Group, op. cit.

11 Congressional Budget Office, Economic and Budget Outlook: Fiscal Years l999-2008 (Washington, DC: January l998).

12 Patricia B. Nemore, "Variations in State Medicaid Buy-in Practices for Low-Income Medicare Beneficiaries" (Washington DC: National Senior Citizens Law Center, November l997).

13 The Balanced Budget Act of l997 also pays the portion of the Part B premium attributable to changes in the Medicare home health benefit ($1.07 a month in l998) for those with incomes up to 175 percent of poverty. This report, however, does not address access to buy-in benefits for this category of beneficiaries.

14 See footnote 1.

15 The Balanced Budget Act of l997 does not stipulate what proportion of the block grant funds states should spend on the two categories of beneficiaries: those between 120 and 135 percent of poverty who quality to apply for payment of their Medicare Part B premiums and those with incomes up to 175 percent of poverty who quality to have the portion of their Part B premiums attributable to changes in the Medicare home health benefit paid for ($1.07 a month in 1988). In this report, it is assumed that all of the state funds would be used for beneficiaries in the first category. This is a reasonable assumption given the very small size of the benefit for those with incomes up to 175 percent of poverty.

16 The Balanced Budget Act of l997 appropriated $200 million in fiscal year l998 and $250 million in fiscal year 1999 for this block grant program. The fiscal year runs from October 1 to September 30. To determine the number of beneficiaries who could potentially be served as QI-1s by the new block grant program in calendar year 1998, we added a quarter of the appropriation available for fiscal year 1999 ($62.5 million) to the $200 million appropriated for fiscal year 1998. The total amount available for calendar year 1998 is $262.5 million. This calendar year adjustment is based on discussions with Miles McDermott, Technical Director for Division of Financial Management in the Center for Medicaid and State Operations, HCFA. To determine the number of beneficiaries who could potentially be served as QI-1s by the new program, we divided the $262.5 million by the annual cost for the Part B premium for an individual in l998 ($525.60).

17 This estimate of $260 million was determined by multiplying the number of beneficiaries who are entitled to participate in the programs based on current funding levels but who were not receiving benefits?494,000 individuals?by the annual amount for the Part B premiums.

18 Marilyn Moon, Crystal Kuntz, and Laurie Pounder, Protecting Low-Income Medicare Beneficiaries (Washington, DC: The Urban Institute, l996).

19 Families USA Foundation, The Medicare Buy-in: A Promise Unfulfilled (Washington, DC: Families USA Foundation, March l993). See also General Accounting Office, Medicare and Medicaid: Many Eligible People Not Enrolled in Qualified Medicare Beneficiary Program (GAO/HEHS-94-52, January l994).

20 See Families USA Foundation, op. cit.

21 See Families USA Foundation, op. cit. and GAO, op. cit.

22 See Patricia B. Nemore, op. cit.

TECHNICAL APPENDIX

by Lisa Maria B. Alecxih, Steve Lutzky, Scott Scrivner, The Lewin Group

In this technical appendix, we present the methodology used to estimate the number of people eligible for, but not participating in, the Medicare buy-in programs. Within our definition of buy-ins are three groups: 1) Medicare beneficiaries with full Medicaid benefits that have their Part B premiums, Medicare coinsurance and deductibles, and other non-Medicare services, such as prescription drugs, covered by Medicaid; 2) Medicare beneficiaries with incomes at or below the poverty level and assets at or below two times the Supplemental Security Income (SSI) criteria that have their Part B Medicare premiums, coinsurance and deductibles covered by Medicaid, referred to as Qualified Medicare Beneficiaries (QMBs); and 3) Medicare beneficiaries with incomes between 100 and 120 percent of the poverty level and assets at or below two times the SSI criteria that have their Part B Medicare premiums covered by Medicaid, referred to as Specified Low-Income Medicare Beneficiary (SLMBs). We also provide an estimate of the number of Medicare beneficiaries eligible for federally paid Part B premiums as a result of the 1997 Balanced Budget Act (BBA) because their incomes are between 120 and 135 percent of the poverty level and their assets are at or below two times the SSI criteria.

In order to compute the number of people eligible for, but not participating in, the Medicare buy-in programs, we estimate both the number of people who are eligible for the buy-in and the number of people who actually receive the benefit. Because we are only estimating the number of non-participants in the community, we adjust the number of buy-ins subtracting those individuals in institutions. We then subtract the number participating from the number of eligibles to calculate the number of eligibles who do not participate. In the remainder of this section, we describe the methodology used to estimate the number of individuals meeting the asset and income requirements for the buy-in programs, present the technique used to estimate the number of buy-in participants, and discuss the limitations of our methodology.

1. Estimating the Buy-in Eligible Population

In order to estimate the number of people eligible for the buy-in benefit, we must identify those people who meet the buy-in income and asset eligibility criteria. We estimated the number of Medicare beneficiaries who fall within the different income criteria using data from the March 1997 Current Population Survey (CPS). These income criteria for eligibility differ from the U.S. Bureau of the Census definition of poverty.1

1 The SIPP topical module on assets and liabilities contains detailed information on individual and joint assets. Like other surveys of asset information, however, some sample members have missing asset information. Asset amounts for these persons are imputed by the Bureau of the Census.
2 State-level data was used if sample sizes met a 95 percent confidence interval test.

In addition to the income criterion, Medicare buy-ins must meet an asset eligibility criterion: countable assets must not exceed $4,000 for single individuals and $6,000 for married couples. The buy-in asset criterion has several main asset exclusions, including: the home that a person or couple lives in; the value of one automobile if it is used to provide necessary transportation, or $4,500 of the value of the automobile if it is not used in such a manner; up to $1,500 of burial spaces; home furnishings; and the cash surrender value of an individual's life insurance policy.

Because the CPS does not include information on assets, we used Survey of Income and Program Participation (SIPP) data to build assumptions about the percentage of individuals meeting the income criteria, but not meeting asset criteria in individual states.2 The most recent available SIPP data are the 1993 data from the 1992 cohort. We produced separate estimates of the percent of individuals who meet the buy-in income criterion but do not meet the asset criterion for single and married persons. We used this percentage, hereafter called the SIPP asset adjustment factor, to adjust the CPS estimates.

We computed SIPP asset adjustment factors at the state level where possible. We were able to calculate state-specific adjustment factors for approximately 20 states. We also computed eligibility for singles and couples separately where possible. For certain states, sample sizes were not sufficient to yield reliable state-level estimates of the asset adjustor and/or the percentage of persons in poverty. In these cases, we assigned the mean adjustor value for the Census Division in which the state is located. For some states, the sample size for couples did not support Census Division Level estimates. In these cases we used the national estimate to adjust couples. Also, in order to include the most representative set of asset adjustment factors, we bound the adjustment factors at one standard deviation from the mean. This minimizes the effect of disproportionately high or low adjustment factors.

Three states, California, Colorado and Massachusetts, have state SSI payments for community-based elderly which, when combined with federal SSI payments, exceed the poverty guidelines. As a result, including only those people with incomes less than the poverty level in the eligibility estimation would understate the actual number of eligibles for these states. To account for this, we defined all Medicare eligibles who received SSI payments as buy-in eligible.

2. Estimating the Number of Buy-in Participants

State-level estimates of Medicare buy-ins were provided for March 1998 by the Health Care Financing Administration's (HCFA) Office of Information Services. We also obtained a national estimate for June 1998 of 4,723 beneficiaries bought in under the new BBA provisions.3

3 Personal Communications with David Evans of HCFA, June 12, 1998.
4 C. McKeen Cowles, The 1997 Nursing Home Statistical Yearbook (San Francisco: Cowles Research Group, 1998).

These estimates are based on state-reported information of the number of Medicare beneficiaries for which the state used Medicaid funds to pay for Part B premiums. These estimates, however, include the institutionalized, a population that is not included in the CPS estimates. In order to exclude the institutionalized population from our estimates, we estimated state-level institutionalization rates for the Medicaid population and adjusted the number of buy-ins accordingly using the following methodology:

We calculated the percent of the total institutionalized population with Medicaid in each state as the primary payor using figures from the 1997 Online Survey Certification and Reporting (OSCAR) data calculated by Cowles.4

We then excluded individuals who do not receive Medicare from the estimates of individuals for whom Medicaid is the primary payor (such cases cannot be buy-in cases) using 1996 Medical Expenditure Panel Data-Nursing Home estimates calculated by the Agency for Health Care Policy and Research. In this case we used a national adjustor to exclude 4.2 percent of individuals.

Because the majority of the states we contacted reported that they only automatically buy-in the nursing home population that receives Medicaid because they are indigent and do not automatically buy-in the medically-needy population, we provided two sets of estimates. The first set of estimates only subtracts medically-needy Medicaid nursing facility residents from the number of buy-ins. The second set removed all Medicaid nursing facility residents. We estimated the percentage of Medicaid nursing facility residents who were indigent versus medically-needy using the Long-Term Care Financing Model. We were not able to make state-level assumptions for this adjustment.

We bound these estimates of the percentage of buy-ins in nursing facilities to one standard deviation to eliminate the effect of very high or low estimates. We present these adjustors in Exhibit 1.

We subtracted this percentage from the number of buy-ins reported in March 1998.

3. Calculating the Numbers of Individuals Eligible, But Not Enrolled

We present the range of the estimated number (rounded to the nearest 1,000) of individuals eligible for, but not enrolled in, the buy-in programs. The number eligible for, but not receiving, benefits is the estimated eligible population minus the number of buy-ins reported after adjusting for those buy-ins in institutions. We also estimate the number of eligibles with incomes between 120 percent and 135 percent of poverty, though sample sizes do not support state-level estimates. We find that nationally there are 1,557,000 persons who meet this income criteria and have qualifying assets.

4. Limitations of Our Methodology

SIPP asset data is imputed for a substantial number of individuals because data are missing. This may affect the validity of the SIPP adjustment factors.

SIPP asset data is based on 1993 data, while we base income on 1997 CPS data. The SIPP asset adjustment factor may be skewed to the extent that the relationship between income and assets changed between 1993 and 1997.

The sample sizes do not support the use of state-level SIPP asset adjustors for certain states. In these cases, we use the average adjustor value for the Census Division in which the state is located. These adjustors may not be reliable if a state differs substantially from the region in which it lies.

Our adjustments to account for the portion of the buy-in population that is in institutions was based on data from other sources. To reach our final numbers, we had to base two minor adjustments on national numbers that we applied evenly across the states. Thus, these adjustments do not capture differences across states.

Exhibit 1
Adjustment Factors Used in this Analysis

For a print copy of this report, contact Families USA at 202-628-3030.

[Return to top]