include long-term care insurance in the cost of joining the community. Either a separate policy is sold, or residents of the community pay, on a fee-for-service basis, for the services they need.

Finally, new mechanisms are becoming available to free up elderly people's resources to finance long-term care. Home equity conversions would allow elderly people to convert what is frequently their largest asset-their house-into income. Banks would provide individuals annuities or lines of credit, to be repaid from the sale of the house at the individual's death. Riders on life insurance policies have also become available that would allow individuals to convert their death benefits to long-term care benefits under certain circumstances.

Each has some potential to provide elderly people access to their savings or assets, although the process of trading assets for income is not without complications and raises a number of issues of consumer protection.72 But access to savings is not the same as insurance. Insurance is designed to protect income and assets, not convert them. Insurance is also intended to cover the costs of service needs. Cash income derived from either housing or life insurance may bear little relation to these needs.

Recognizing the limits to the individual insurance market aimed at the elderly, some insurers have recently turned to employers as a better way to spread the risk and to market their products. Job-based policies have the potential to spread the financial risk among low-risk as well as high-risk purchasers, and allow the financing of insurance over the working years of the employee. Further, they have lower marketing and administrative costs. Premiums for individual policies tend to be 20 percent to 30 percent higher or more for the same coverage sold through an employer-sponsored plan.73

While such plans were unheard of three years ago, the number of employers offering this coverage has grown from a mere two in 1987 to 47 in 1989, with another 64 to become effective in 1990. In addition to employees, most of the plans were offered also to retirees and their spouses, current employees' spouses, and the parents and parents-in-law of the employees. Of the 51,368 persons enrolled in employer-sponsored long-term care plans as of the end of 1989, half were active employees. The rest were retirees and their spouses or immediate relatives of the employee.74

The development of job-based coverage is encouraging, but except for a few pilot programs negotiated by the United Auto Workers with Ford and Chrysler,

the premiums for the insurance are still paid entirely by employees. Furthermore, even in the larger firms offering coverage, availability of coverage depends on each individual's health status.

Enthusiasm over the prospects of restructuring employee compensation to include long-term care insurance must be tempered by the realization that employers are preoccupied with controlling their health care costs and accounting for postretirement medical benefits. Fear of creating an unintended liability-analogous to employers' experience with postretirement medical benefits-may inhibit employers from expanding benefits to include long-term care. In addition, too few employees are affected by long-term care needs at any one time to constitute a critical mass agitating employers to change their employee compensation to add this benefit. It must be remembered that even for more common employee benefits, such as health plans and pensions, coverage is not universal. In 1986, over 40 percent of employers did not provide health insurance to their employees.75 In 1986, less than half of all full-time employees were covered by pension plans. 76

Most important, even if employers offer and employees purchase benefits, job-based coverage—like individual coverage-does not guarantee adequate protection. Unlike health coverage, where people pay in every year to cover costs in that year, for the most part long-term care coverage means paying in over a number of years for coverage most likely to be used after retirement. Therefore, ensuring the adequacy of long-term care benefits, like ensuring the adequacy of pensions, means addressing such issues as portability of coverage across employers (including movement from an employer who offers insurance to an employer who does not) and some return on employees' contributions even if participation should lapse.

In sum, for today's disabled population, private long-term care insurance is not an option. Insurers do not sell policies to persons who are already impaired. The current market is relatively small-at the most 5 percent of the elderly population held policies in 1989. Furthermore, premiums would absorb an enormous share of income for most elderly, and policies would typically still leave many costs uncovered. The greatest potential for private insurance to play a major role in the financing of long-term care thus lies in the future.

Future Prospects for the Market

The emergence of job-based coverage, combined with expected improvements in the average income of

the elderly population, makes some observers optimistic that private insurance coverage will grow until it fills a considerable part of the need for long-term care protection.

The extent of long-term care insurance in the future depends on people's behavior: how many people will purchase long-term care insurance during their working years, how many will continue paying premiums for the rest of their lives, and how many people who do not buy such insurance before retirement will be willing and able to purchase it after they retire.

Most policies are marketed as lifetime decisions. That is, the price does not reflect the current risk of needing long-term care but, rather, is based on an average lifetime risk. Consequently, the price of a policy purchased by a nonelderly person is low relative to the risk later in life, but high relative to the risk of needing long-term care in the earlier years. Some insurers are considering policies that refund part of the premium paid but, in current policies, once payments cease the policy no longer has any value. With no refunds and without any inflation protection, insurance purchased at age 30 costs a tenth of what the same policy costs at age 65.77

Will a 50-year-old, who might not need any longterm care for 25 or more years, be willing to buy a policy that will pay $50 a day for a nursing home stay and $25 a day for home health care? The answer depends on the potential purchaser's assessment of several risks, including the risk that his or her expectations about inflation and the future cost of long-term care will be wrong, that he or she will not be able to afford to continue to pay the premium, or that the insurer will not honor the policy (either because the company is no longer in business or because that type of policy was cancelled). It also depends on the extent to which employers are willing to subsidize premiums of workers and, if so, how government regulates the insurance market. Both are uncertain.

Will a retiree continue to pay premiums after retirement or buy coverage for the first time? The future behavior of the elderly will depend heavily on their incomes and assets. Successive groups of elderly people will stop working with more retirement income and income from savings. Therefore, the average income of the elderly, adjusted for inflation, is projected to increase.78

Despite improvements over time, however, there will still be elderly for whom long-term care insurance is simply too expensive. Among workers expected to retire between the year 2012 and 2021, for

example, about 18 percent will have retirement incomes (pensions, retirement accounts, social security) below 125 percent of poverty.79

About 80 percent of unmarried elderly women and 60 percent of unmarried elderly men will have incomes from all sources of less than $20,000 (1988 dollars) in 2030-leaving premiums for long-term care insurance in excess of 5 percent of income for more than 70 percent of the unmarried elderly in 2030 (if they do not purchase insurance until after age 64). For 48 percent of all elderly (married or not), a premium of $720 a year is projected to be within 5 percent of their gross income in 2030. For another 23 percent, such a premium would be within 10 percent of their income. 80 In today's market, however, a $720 policy would provide only limited coverage.

The uncertainty of income, assets, consumer preferences, and the quality of the products offered makes it difficult to estimate the scope of long-term care coverage in the future with any precision. Estimates prepared for the Commission indicate that by the year 2018 from 24 percent to 48 percent of the elderly might actually have coverage, assuming a less comprehensive policy (less home care, lower payments per day of nursing home care) than the typical "good" policy reported by the HIAA.81 Estimates depend heavily on how many employers are projected to provide access to long-term care insurance policies.82 The top of the range-48 percent is the proportion of elderly for whom premiums would be within 5 percent of income.

"During the early 1960s it was considered intolerable that only half of the elderly had some form of acute care insurance. Based on our work, private long-term care insurance will be lucky to reach that level of market penetration."

-Joshua M. Wiener, senior fellow,
The Brookings Institution

The development of private insurance for long-term care may alleviate financing problems for some portion of the future elderly and nonelderly disabled population. But the adequacy and scope of that protection cannot be ensured without government oversight and support. Furthermore, even if private insurance protection grows substantially, two critical questions will remain:

• How to protect the currently disabled population as private insurance expands,

How to protect the population that will inevitably be left out.

PROSPECTS FOR THE FUTURE

Growth in the numbers of people likely to need long-term care makes improvements in the nation's financing of this care not only important but imperative to the well-being of all Americans. The size of the population over age 65-and, more important, over age 85-is increasing. As a result, the demand for long-term care will almost certainly rise substantially. Even among the younger population, needs can be expected to grow. Experience so far raises serious doubts about the adequacy of current long-term care mechanisms to respond to greater need.

Changing Needs

Growth in the numbers of older people reflects several demographic trends, including longer lifespans for older people and the aging of the baby boom generation. Thirty-two million Americans are now age 65 and over. That number is expected to double between now and 2030, even on the conservative assumption that improvements in longevity slow to half their historical rate. In roughly the same period, the population age 85 and older is expected to increase almost fivefold (from 2.5 million to as many as 12 million).83

people, is likely to mean greater need for long-term care. Exactly how much greater is uncertain. Medical advances could reduce people's need for care, but longer lifespans could also increase this need. If disability rates remain what they are today, by 2030 the number of elderly persons needing help with basic tasks is expected to increase from about 7 million to 13.8 million and the number requiring nursing home care, from 1.5 million to 5.3 million (see Figure 3-9).84

Growth in the elderly population is the major but not the only source of potentially increased long-term care needs. The use of high technology and new medical breakthroughs may continue to extend the lives of more mentally retarded, developmentally disabled, and physically disabled persons. Increased survivorship of low birthweight children, greater longevity for children with terminal chronic illness, and earlier detection of chronic health problems have doubled the proportion of disabled children under 17 over the last 25 years-from less than 2 percent in 1960 to more than 5 percent in 1984.85 This growth will undoubtedly continue in the future.

In addition, we are just beginning to recognize the long-term care consequences of the AIDS epidemic. A recent study in five communities points to much unmet need for long-term care already among AIDS patients in both institutions and the community. 86 The demand for care will undoubtedly grow as more persons are diagnosed with HIV-positive infections.

Capacity to Respond

The capacity of families and programs to provide needed care is already severely strained. It seems highly unlikely that service availability will keep up with the growing service needs.

First, availability of family care, on which people depend if they are to remain at home, is likely to diminish.87 Specifically, the recent decline in fertility rates suggests that there will be fewer adult children potentially available to provide care. The continued increase in female labor force participation, coupled with the trend toward delayed childbearing, also can be expected to lead to fewer adult children potentially available to provide care. Daughters, in particular, will be less able and willing than the current generation of adult children to care for their elderly parents because of the competing demands of employment and child care. Larger numbers of younger elderly, between 65 and 70, may offset the loss of caregivers due to employment. But the prospects are uncertain.

Second, future elderly will be financially better-off, facilitating their ability to purchase care. But projected income improvements are heavily concentrated among the younger elderly, rather than the older population most likely to need care. Between 1990 and 2020, real incomes of people 65 to 74 are expected to double. Among the population 85 and older, however, real incomes are expected to increase only

"[T]he increasing population frailty and the increasing need for women to be in the labor force... is a very fragile basis for the nation's long-term care policy.... Of these caregivers, over 50 percent are working, and a large proportion of them already make work sacrifices themselves, limiting their own future coverage and social security and benefits, and are contributing substantially in a number of cases from their own financial income to try to support their elderly relatives."

-Dr. Carroll Estes, director, Institute of Health and Aging, University of California, San Francisco

17 percent.88 The capacity of people who most need the care to pay for it will remain decidedly limited.

Whether services will be available if people are able to pay is also an open question. Use of paid care in the home is very limited today, even among people with higher incomes. Among the likely reasons for limited use is the difficulty individual consumers face in identifying reliable providers of long-term care services. Traditional health care providers are often reluctant to provide and organize the array of services that disabled people require. Suppliers of homemaking, transportation or other services that are part of long-term care can be reluctant to serve a disabled population. Without government involvement, the private marketplace seems unable to develop a delivery system to deliver needed services efficiently to the disabled population.

As currently structured, existing programs are unlikely to provide the requisite public leadership. Medicare's focus on people who need skilled care, even with more liberal definitions, precludes extended longterm care assistance. In theory, Medicaid benefits might allow broader service, at least for people poor enough to qualify. In practice, however, federal rules

and expenditure concerns in most states are likely to continue to limit service. Without more federal assistance, even states now building long-term care programs with heavy reliance on their own funds are likely to be unable to keep up with the need for care.

Although most states have little investment in home care, almost all are heavily involved in nursing home support. But that involvement by no means guarantees sufficient support to meet growing needs. Research indicates that unless low-income people buy private coverage, even substantial expansion of private insurance will have only modest effects on the future need for publicly financed nursing home care.89 Insurance will be expensive; the better its protection, the more it will cost. Low-income people will almost certainly find themselves without private coverage. Job-based long-term care coverage is unlikely to reach large numbers of middle-income workers in their retirement years. Many of these people, like the current elderly, will have to depend on public programs for nursing home care.

It is unlikely that a welfare-based program will attract the substantial resources necessary to meet these needs. If Medicaid were to provide coverage 30 years from now on the same terms it does today, its expenditures-measured net of general inflation-would be almost triple what they are now, even if private insurance were to expand dramatically.90 These estimates

assume that nursing home costs will increase faster than inflation. Even if costs are held to general infla tion rates, greater use alone would increase real Medicaid spending by about 75 percent.

"What we have now is a perverse [Medicaid] system. It needs to be changed from a nonsystem that kicks people when they're down to an insurance program that helps people plan for their later years for themselves and for their loved ones."

- Senator Max Baucus

Experience over the last decade suggests that Medicaid will not achieve the coverage or spending levels necessary to support nursing home use at current levels. Nursing home use has not kept pace with growth in the elderly population and the estimated need for long-term care, and in many places the need for nursing home care exceeds the supply of beds. Under these circumstances today's Medicaid patients have trouble finding nursing home care. Without a change in public policy, more and more Americans will have difficulty getting the care they need, in nursing homes as well as at home.