Statement of
Douglas Holtz-Eakin
Director

The Cost and Financing of
Long-Term Care Services

before the
Subcommittee on Health
Committee on Energy and Commerce
U.S. House of Representatives

April 27, 2005

Mr. Chairman and Members of the Subcommittee, thank you for the opportunity to be here today to discuss the cost and financing of long-term care (LTC) services. A Congressional Budget Office (CBO) report from April 2004, Financing Long-Term Care for the Elderly, examines these issues in greater detail. Long-term care is the personal assistance that enables people with impairments to perform daily routines such as eating, bathing, and dressing. Such services may be provided at home by family members and friends; through home and community-based services such as home health care, personal care, and adult day care; or in institutional settings such as nursing or residential care facilities.

In my statement today I want to make the following points:

• With the aging of the baby-boom generation, the United States' elderly population is expected to grow rapidly over the next several decades. The surge in the number of seniors will increase the number of people with impairments and, in turn, the demand for long-term care services.
  
  
• The resources devoted to long-term care services are already substantial. CBO estimates that spending on such care for the elderly (including the value of donated care) totaled over $200 billion in 2004--or approximately $24,000 per senior with impairments. In reporting estimates of LTC spending, CBO chose to include the value of donated care because it is an integral part of long-term care, even though measuring it accurately is difficult.
  
  
• Currently, donated care is the largest source of financing for long-term care costs, followed by the combined public programs--Medicaid and Medicare--and out-of-pocket expenditures. Private long-term care insurance is a small portion of the current financing.
  
  
• Financing patterns for long-term care are heavily influenced by the rules governing public LTC programs. Those rules create incentives that discourage people from making their own financial preparations and encourage them to rely on government assistance. If left unchanged, those incentives will add to the financial demands that government programs for retirees are already facing as a result of demographic changes and rising health care costs.

Demographic Trends

The oldest members of the baby-boom generation become eligible for early retirement under Social Security in 2008. According to estimates by the Bureau of the Census, the number of elderly people (those age 65 and older) in the United States will increase by two and a half times between 2000 and 2050. The share of the population claimed by the oldest seniors, those age 85 and older--and those most likely to use long-term care--will reach about 5 percent by 2050, more than triple the 1.5 percent share they had in 2000 (see Figure 1). By comparison, the proportion of the population accounted for by working-age people (ages 20 to 64) will grow by only about 35 percent by 2050.

Figure 1.

Source: Congressional Budget Office based on Bureau of the Census, U.S. Interim Projections by Age, Sex, Race, and Hispanic Origin, Table 2a, "Projected Population of the United States, by Age and Sex: 2000 to 2050" (March 2004), available at www.census.gov/ipc/www/usinterimproj/natprojtab02a.pdf.

Although the number of the oldest seniors will rise, declines in the prevalence of functional impairment could offset some of the effects of that increase. Impairment among seniors appears to have waned significantly during the 20th century. From 1910 to the early 1990s, the overall prevalence fell by about 6 percent per decade. From the early 1980s to the present, the prevalence of impairment may have fallen even faster, according to research findings from the National Long-Term Care Survey. In contrast, some types of impairment, such as those requiring the use of a cane to walk, have been increasing. Impairment among people under age 65 may also be increasing, which could eventually lead to higher future rates of impairment among seniors. In fact, one recent study projects that the currently declining trend in the prevalence of impairment among seniors will reverse in the future, leading to greater rates of institutionalization than those that exist today.⁽¹⁾ As those conflicting trends suggest, projecting the prevalence of impairment in future years and basing estimates of spending on those projections are both difficult and subject to a high degree of uncertainty.

Demographic changes may affect the composition of LTC financing in the future as well. Smaller families, lower fertility rates, and increasing divorce rates may make donated LTC services less common in the future. The size of the average family has declined, reducing the number of adult children available to care for their elderly parents. Family size fell from 3.8 members in 1940 to 3.1 members in 2000; if current trends continue, it will decline to 2.8 people by 2040. At the same time, the rate at which women participate in the labor force will probably continue to grow, at least until 2010, further reducing the availability of donated care. Those family-related trends, in sum, could further stimulate the demand for formal, or paid, services.
 

Sources of Long-Term Care Financing

Long-term care is financed with both private resources and public programs (see Figure 2). Private resources include donated care, out-of-pocket spending, and private insurance. Public programs include primarily Medicaid and Medicare, although the Department of Veterans Affairs and the Social Services Block Grant program also fund long-term care.

Figure 2.

Source: Congressional Budget Office.

Private Sources

Most seniors with impairments who reside in the community, including those with severe impairments (unable to perform at least four activities of daily living, or ADLs), rely largely on donated care from friends and family. And many people who pay for care in their home also rely on some donated services.

The economic value of donated care is significant, although estimates of it are highly uncertain. In 1998, the Department of Health and Human Services estimated that replacing donated LTC services for seniors with professional care would cost between $50 billion and $103 billion (in 2004 dollars). Another analysis, in 1997, estimated the value of donated care for people of all ages who had impairments--measuring it as the forgone wages of caregivers--at $218 billion.⁽²⁾

Out-of-pocket spending in 2004 accounted for about one-fifth of total LTC expenditures, or roughly $5,000 per senior with impairments (see Table 1). The federal government subsidizes a portion of out-of-pocket spending through the tax code. Taxpayers with impairments (or taxpayers who have dependents with impairments) may deduct LTC expenses from taxable income along with other medical and dental costs, but only the portion of total medical costs (LTC, medical, and dental expenses) that exceeds 7.5 percent of adjusted gross income.

Table 1.

Private insurance for long-term care is a relatively recent development and pays for only a small amount of care at present. Few elderly people currently have private coverage--no more than 10 percent.⁽³⁾ However, that source of financing is growing--although the precise extent of the growth is difficult to measure accurately. The data on private LTC insurance generally capture payments that insurers make directly to providers but do not always pick up insurers' reimbursements to policyholders for covered services that policyholders initially pay for out of pocket. Thus, estimates of LTC insurance payments--and of out-of-pocket spending--should be interpreted with caution because the former may be underestimated and the latter overestimated.

In 1995, private insurance paid about $700 million for LTC services for seniors, or 0.8 percent of all such expenditures. In 2004, such spending totaled about $6 billion, CBO estimates, or about 3 percent of total expenditures. According to America's Health Insurance Plans, the number of policies written yearly increased from about 300,000 in 1988 to more than 900,000 in 2002 (see Figure 3). About 9.2 million policies were sold from 1987 through 2002; roughly 72 percent of them are still in force.

Figure 3.

Source: America's Health Insurance Plans, Research Findings: Long-Term Care Insurance in 2002 (Washington, D.C.: AHIP, June 2004), p. 15.

A typical LTC insurance policy pays the cost of nursing home care and home and community-based care but specifies a maximum daily benefit (such as $100 or $150) and may impose other limits. Policies with so-called inflation protection increase the dollar value of their benefits by a contractually specified percentage each year, usually 5 percent. Although some policies offer coverage for an unlimited period, most commonly cover services for a shorter time, such as four years, or until benefit payments for a policyholder reach a preestablished maximum lifetime amount. Policyholders typically become eligible to collect benefits when they reach a specific minimum level of impairment, usually defined as being unable to perform two or three ADLs or having a cognitive impairment significant enough to warrant substantial supervision.

Premiums for LTC insurance reflect the cost of services and the risk that policyholders will require long-term care as they age. In 2002, the average annual premium for a typical policy with no inflation protection or nonforfeiture benefit was $1,337 if the policy was purchased at age 65; with those two added features, the premium rose to $2,862. Premiums were three to four times higher if the policy was purchased at age 79 (see Table 2). The lower premiums offered to younger people reflect the lower risk of their requiring LTC services at younger ages and the expectation that younger policyholders will pay premiums over a longer period than will people who purchase coverage when they are older. Thus, the average annual premium for the same policy with inflation protection and a nonforfeiture benefit purchased by a 40-year-old would be only $1,117 and by a 50-year-old, $1,474.

Table 2.

In fact, fixed premiums are a key feature of LTC insurance policies--that is, the premiums do not increase as the policyholder grows older or as his or her health deteriorates, even though the risk of requiring services rises. Instead, insurers calculate premiums to ensure that the premiums' total, paid over the life of a policy, plus the interest that accrues from investing them will be sufficient to cover both the claims of the policyholder and insurers' profits and overhead costs. However, insurers reserve the right to increase premiums for a specific group, or rating class, of policyholders--such as all policyholders in a state--if new data indicate that expected claims will exceed the class's accumulated premiums and their associated investment returns.

Government Programs

Medicaid is the biggest government source of payment for long-term care. Jointly funded by the federal and state governments, Medicaid is a means-tested program that pays for medical care for certain groups of people, including seniors with impairments who have low income or whose medical and long-term care expenses are high enough that they allow those seniors to meet Medicaid's criteria for financial eligibility. Within broad federal guidelines, the states establish eligibility standards; determine the type, amount, duration, and scope of services; set the rate of payment; and administer their own programs. The share of each state's Medicaid expenditures that is paid by the federal government is determined by a statutory formula; nationwide, the federal share of the long-term care portion of Medicaid spending is about 56 percent.

Medicaid generally pays for services provided both in nursing facilities and in the home, although the specific benefits that the program provides differ from state to state, as do patterns of practice, the needs and preferences of beneficiaries, and the prices of services. In total, Medicaid's expenditures for long-term care for elderly people since 1992 have grown at an average annual rate of about 5 percent (see Figure 4). CBO estimates that in 2004, Medicaid's payments for institutional care for seniors, including both state and federal expenditures, totaled about $36.5 billion. Accounting for about 40 percent of total expenditures on nursing facilities, Medicaid's payments cover the care of more than half of all elderly nursing home residents.⁽⁴⁾

Figure 4.

Sources: Personal communication by Brian Bruen of the Urban Institute, and the Congressional Budget Office's estimates.

Medicaid's expenditures for home and community-based services (HCBS), which include home health care, personal care services, and spending under HCBS waiver programs, are much smaller than its spending for nursing homes--HCBS expenditures constitute only about 23 percent of total Medicaid LTC spending. (Under the waiver programs, states have the option of providing people with impairments with enhanced community support services not otherwise authorized by the federal statutes.) Since 1992, Medicaid spending for home-based care for seniors has grown faster than spending for institutional care, rising by about 11 percent annually, on average, compared with about 3 percent growth for care in nursing facilities.

Many people who are not eligible for Medicaid while they live in the community become so immediately or shortly after being admitted to a nursing facility because of the high cost of institutional care. (Nursing home costs in 2004 averaged about $70,000 annually for a private room.) According to a 1996 study, about one-third of discharged nursing home patients who had been admitted as private-pay residents became eligible for Medicaid after exhausting their personal finances; nearly one-half of current residents had similarly qualified for coverage.⁽⁵⁾ Medicaid coverage is especially common among nursing home patients who have been institutionalized for long periods.

Medicare, the nation's health insurance program for the elderly, covers care provided in skilled nursing facilities (SNFs) and at home, but its benefits are designed primarily to help beneficiaries recover from acute episodes of illness rather than to provide care for long-term impairment.⁽⁶⁾ Medicare covers up to 100 days per spell of illness for SNF care, and the stay must be preceded by a hospitalization lasting at least three days. In contrast, Medicare's home health benefit, while originally conceived to finance short-term rehabilitation, has evolved into what some observers have described as a de facto LTC benefit. To be eligible for reimbursement under the home health benefit, the beneficiary must be homebound and require intermittent care provided by a licensed professional, such as a registered nurse or physical therapist. If those conditions are met, Medicare will cover services provided by a home health aide, in addition to skilled care; aide services are the assistive services that typify long-term care.

By CBO's estimate, Medicare's LTC spending for seniors in 2004 totaled about $16 billion for care in skilled nursing facilities and $18 billion for home health care (see Figure 5). Although the program's outlays for those categories grew rapidly from the late 1980s to the mid-1990s, expenditures actually declined near the end of the past decade. A combination of factors was responsible, including changes to reimbursement methods imposed by the Balanced Budget Act of 1997, increased federal activities to counter providers' fraud and abuse of the program's payment systems, and delays in processing claims. CBO projects steady growth in spending for SNF and home health care over the 2006-2015 period, averaging approximately 5 percent annually.

Figure 5.

Source: Congressional Budget Office.

Issues in Controlling Federal Long-Term Care Spending

CBO has projected that total LTC expenditures for seniors (including the value of donated care) will rise from about $195 billion in 2000 (2.0 percent of gross domestic product, or GDP) to $540 billion (in 2000 dollars) by 2040, or 2.3 percent of GDP.⁽⁷⁾ That estimate of a relatively modest increase in use of long-term care services incorporated the assumption that the prevalence of impairment would decline at a rate of about 1.1 percent per year. If impairment levels instead remain about the same as they are today, use of services will rise faster, to $760 billion by 2040, or about 3.3 percent of GDP. Demand for care could be even higher if, as some researchers believe, the prevalence of impairment actually increases in the future.

The current mix of financing for long-term care, in which a significant share of financing comes from government programs, adds to the pressures that the federal budget will experience with the aging of the baby-boom generation. Contributing to the strains that government LTC programs will face are incentives created by those programs that diminish the attractiveness of using private resources--especially private insurance--as a means for seniors to finance their care. Changes in those incentives might encourage more people to make their own preparations for financing their care rather than rely on governmental assistance.

Direct Approaches to Limiting Federal Spending for Long-Term Care

One approach to relieving the pressures on federal finances would be to directly reduce the role of Medicaid and Medicare, the programs responsible for the bulk of government-financed care. The most commonly discussed options are tightening the financial qualifications for people applying for Medicaid coverage and reducing Medicare's coverage of home health care.

Medicaid's spending for long-term care could be constrained by making it more difficult for middle-income people to qualify for coverage by spending down their resources. The intent of Medicaid's current rules is to restrict applicants to those who are destitute. Yet despite that intention, many applicants manage to protect a significant portion of their personal wealth and still qualify for Medicaid coverage by taking advantage of certain rules regarding the disposition of assets, a practice known as Medicaid estate planning. Strengthening the rules to reduce the use of such strategies would delay the point at which some people became eligible for benefits and would prevent others from qualifying. It could also discourage some people from going through the application process. However, it is unlikely that imposing those additional restrictions would have more than a modest impact on Medicaid's expenditures.⁽⁸⁾

Medicare's home health care benefit is relatively generous. Once a person meets the physical qualifications for coverage, there are no copayments or other coinsurance requirements. A modest cost-sharing requirement for beneficiaries could decrease the program's LTC expenditures because beneficiaries would probably reduce the amount of care they used in response to that kind of financial incentive.

Challenges in Encouraging Private Financing of Long-Term Care

Future federal spending on long-term care could be lessened by encouraging people to rely more on private resources for their LTC needs. Out-of-pocket spending and donated care already account for a very substantial share of LTC services, but private long-term care insurance currently finances very little such care. CBO estimates that the proportion of LTC spending that private insurance pays will rise to about 17 percent in 2020; that share would be less than the shares of either Medicaid or Medicare. Several factors underlie the limited rise that CBO projects for the use of private insurance. Some factors affect the availability and quality of insurance: they include issues related to administrative costs, the instability of premiums, adverse selection, and the inability to insure against certain risks unique to long-term care. A final factor--the interaction of private insurance and Medicaid--is critical in the way it affects demand for private insurance.

Administrative Costs. Administrative costs contribute a substantial amount to LTC insurance premiums because most policies are sold individually rather than as group (employer-sponsored) policies.⁽⁹⁾ The costs of marketing to and enrolling individuals are about double those for groups, for which fixed administrative costs may be spread over more people.

On average, administrative costs as a percentage of premiums are likely to fall in the future as group policies make up a larger share of the private LTC insurance market. In 2002, group policies constituted nearly one-third of new LTC policy sales.⁽¹⁰⁾ (By comparison, nearly 90 percent of people with private health care insurance hold group coverage.⁽¹¹⁾) But group policies are accounting for an increasing share of the LTC insurance market, a trend that is likely to continue if more employers offer LTC coverage as an employee benefit. If employers offer such a benefit, any part of the premiums for their employees' LTC coverage that they pay for, like their contributions for regular health insurance, is not included in employees' taxable income.

Instability of Premiums. Although LTC insurers typically offer premiums that do not automatically increase as the policyholder grows older or experiences deteriorating health, state insurance regulators allow insurers to increase premiums for all holders of a given type of policy in a state (known as a rating class) if they find that they have miscalculated the expected cost of their claims. Some insurers have boosted premiums several times for that reason, leading many policyholders to cancel their coverage and in all likelihood deterring some potential purchasers from acquiring LTC coverage.⁽¹²⁾ However, premiums may be stabilizing: a survey of top-selling LTC insurance carriers by the Health Insurance Association of America observed fairly steady premium levels from 1997 to 2001 after a sustained decline in average premiums from 1990 to 1996.⁽¹³⁾

Policyholders can obtain some protection against large jumps in premiums by purchasing nonforfeiture benefits with their policy. That feature enables policyholders who cancel their coverage to recoup from the insurer at least some of the premiums they have paid. Nevertheless, although policyholders might get a proportion of their premiums back, they do not receive the associated returns on the investment of that money.

Adverse Selection. The relative newness of the market for LTC insurance and the still fairly small number of policies being sold suggest that the market may be affected by adverse selection. People who purchase LTC insurance have greater expectations than nonpurchasers of using services in the future, and those greater expectations are not captured in the information that insurers collect as they enroll purchasers of their policies. If insurers believed that adverse selection was occurring, it might lead them to set premiums higher than a policyholder's health status would suggest so as to incorporate the greater likelihood that that policyholder would use the insurance. In turn, the higher premiums might deter people who would purchase coverage if the premiums reflected their relatively lower expectations of using LTC services.

One recent study suggests, however, that although adverse selection does exist in the LTC insurance market, it may not be producing higher overall claims costs.⁽¹⁴⁾ According to that study, the higher costs of policyholders with greater-than-average expectations of using services in the future are offset by the lower costs of policyholders who are averse to risk and whose probability of using services in the future is actually lower than the average for the population at large. Because of the market's youth, there are no clear data to resolve the question of adverse selection.

The Inability to Insure Against Certain Risks. Private LTC insurance may be unattractive to some consumers because it does not, in general, insure against the risk of significant price increases for long-term care. Most policies promise to provide contractually specified cash benefits in the event that a policyholder becomes impaired. To protect themselves against LTC price inflation, consumers can purchase a rider to their policy under which the policy's benefits grow at a specified rate each year (usually 5 percent); however, such riders offer no protection against additional costs if prices rise at a faster pace. Concerns about price increases of that kind are not unjustified: Medicaid's average reimbursement rates for nursing facilities grew at an average annual rate of 6.7 percent from 1979 to 2001.⁽¹⁵⁾ Over a 20-year period, a nursing facility benefit of $100 per day in today's dollars would grow to $265 per day with an annual inflation protection rider of 5 percent. But the benefit would need to grow to $366 per day to keep up with a 6.7 percent annual growth rate, should costs continue to grow that fast in the future.

An additional risk is that a policy could become obsolete at some point in the future. LTC services, and the private insurance policies that cover such care, are steadily evolving as the LTC insurance market matures. That fluidity may give some consumers pause, and indeed, one prominent rating agency recommended in 2000 that people purchase LTC coverage no earlier than age 60 to avoid the problem of obsolescent coverage.⁽¹⁶⁾ Some consumers might also be reluctant to purchase LTC insurance if they believed that changes in public policy at some point could render their coverage obsolete.

The Availability of Medicaid. The availability of Medicaid poses a substantial disincentive for people considering the purchase of private long-term care insurance. Although Medicaid in general serves people with very low income and assets, it also provides assistance to people with impairments who exhaust all of their private sources of financing for their long-term care. Even people who have set aside significant savings may eventually become eligible for Medicaid assistance. In that way, Medicaid serves as an alternative form of insurance for people who do not have private coverage and who are impaired for a significant period. Indeed, Medicaid's impoverishment requirement may discourage people from saving because the less they have, the more quickly they will qualify for coverage. It also creates an incentive for people to give away or hide their assets so that they can qualify for Medicaid.

There are substantial drawbacks to Medicaid coverage for long-term care. As a means-tested program, Medicaid requires eligible applicants to rely on out-of-pocket spending until they use up all of their savings. In addition, because Medicaid generally pays lower fees for services than those paid by private payers, beneficiaries may not receive the same quality of care that private policyholders receive. In some states, moreover, Medicaid might not be as flexible in the types of services it covers as private insurance would be; a person who has private coverage would probably have a broader choice of providers and types of care than a Medicaid beneficiary would have.

Those drawbacks to Medicaid's coverage are balanced by features that some people might consider advantageous. Medicaid is free from the perspective of the beneficiary. In addition, Medicaid has a defined-benefit structure--that is, it covers a specified set of services. Private insurance, by contrast, only ensures that a policyholder will have a specified monetary benefit to pay for care. It does not guarantee that the money will be sufficient to pay for desired services.

Although Medicaid's coverage differs in some respects from that of private insurance, it may nevertheless reduce the demand for private policies. Indeed, one recent study found that the availability of Medicaid constitutes a substantial deterrent to the purchase of private insurance, even for people at relatively high income levels.⁽¹⁷⁾ Medicaid's rules for financial eligibility affect people's decisions to purchase private LTC insurance as well as how much insurance they buy because the rules offer a low-cost alternative (by allowing people to qualify for the program's benefits) to making personal financial preparations for possible future impairment. People who buy private insurance or accumulate savings substantially reduce the probability that they will ever qualify for Medicaid's benefits, thereby forgoing the value of the government-provided benefits that they might otherwise have obtained. Thus, the availability of Medicaid raises the perceived cost of purchasing private insurance or of saving. That increase is small for relatively wealthy people who have little likelihood of ever qualifying for Medicaid coverage, but it can be substantial for others.
 

Conclusion

Currently, elderly people finance LTC services from various sources, including both private resources and government programs. Incentives inherent in the current financing structure have led to increased reliance on and spending by government programs and may have discouraged people from relying on private resources (savings, private LTC insurance, and donated care) to prepare for potential future impairment. The demographic changes projected for the coming decades will bring increased demand for long-term care and heightened budgetary strains.