A number of factors will converge to shape the magnitude, scope and nature of the future demand for long-term care including changing demographics and the health and functional status of these populations, the availability of family and other informal caregivers, the financial status of future generations and the degree that they engage in advanced planning for long-term care needs, and the availability and cost of institutional and community-based alternatives. It is important to note, however, that the crystal ball is invariably inaccurate. Projections must be examined with caution-a major medical advance in the treatment of Alzheimer's disease, for example, or a pandemic that wiped out a large subpopulation, could significantly influence the size and character of the long-term care population in the future.
The Impact of Population Aging
The 21st century will be marked by an unprecedented increase in the size of the elderly population as the large baby boom generation ages. While most elderly people are not disabled, the likelihood of needing long-term care services increases with age. The 65 and over population is expected to increase substantially between now and 2040,. The use of different mortality, fertility and immigration assumptions, however, creates widely disparate estimates (Friedland & Summer, 1999) ranging from 59 million to 92 million elders living in the United States in 2040 (U.S. Bureau of the Census, 1996). Estimates for the 85 years or older population, those most likely to need long-term care, range from 8.3 million to 20.9 million in 2040. Older women will continue to outnumber men, although in recent years women have not seen the significant decline in deaths from heart disease that have benefited men (Treas, 1995). If trends such as these continue, the gender gap in longevity may shrink somewhat.
The change in the number and proportion of the elderly population will be affected further by the racial and ethnic diversity in the U.S. In 1995, 85 percent of the 65 and over population was non-Hispanic white; eight percent was black, five percent was Hispanic and the remainder was other races. By 2050, the distribution of elderly Americans by race and ethnicity is projected to be 67 percent non-Hispanic white, 10 percent black, 16 percent Hispanic, almost seven percent Asian Pacific Islander, and less than one percent other races (Figure 3) (Friedland & Summer, 1999)
Increased Longevity: Quantity Vs. Quality
Part of the projected increase in the number and proportion of elderly in the 21st century is due to an increase in life expectancy at age 65. Males can expect to live an additional 17.6 years in 2020 and 18.5 years in 2030; this is up from 15.5 years in 1995. The comparable estimates for females are 20.6 years in 2020 and 21.2 years in 2030, up from 19 years in 1994 (U.S. Bureau of the Census, 1996).
But the extent to which these added years will be quality years free of disability is the subject of some debate. Studies (Colves & Blanchet, 1983; Crimmins et al., 1997) indicated that life expectancy at birth in the U.S. was 72.7 years in 1976 and 75.4 years in 1990. But the life expectancy free of disability was 56 years and 58.9 years respectively, leaving the other 16.7 years and 16.5 years spent with a limitation in activity. Some researchers argue that medical advances have increased life expectancy but have not changed the onset of illness (GAO, 1994). They predict that declining death rates may actually increase long-term care need if, for example, more people live to develop age-related disabling conditions such as dementia or live longer with existing disabilities.
Projecting the size of the long-term care population in an aging America has become a subject of particular interest to researchers. One recent analysis using 1994 National Health Interview Survey data applied to U.S. Census Bureau projections estimates that the 65 and over population with activity limitations will rise from 12 million in 1994 to 22 million in 2020 to 28 million in 2030 (Rice, 1996). Kunkel and Applebaum (1992) have emphasized the need to use a dynamic model, where the rates of disability and mortality are not necessarily assumed to remain constant over the next half century. In estimating the size of the elderly long-term care population using various scenarios, they note that by the year 2040-when the baby boomers reach their 80s and 90s-the number of older Americans with disabilities will have increased significantly. Their estimates of the over-65 population expected to have a long-term disability, however, range from 14.8 to 22.6 million people depending on the assumptions built into the model.
Manton and colleagues (1997) report that disability rates among the elderly may be declining (Figure 4). They estimated that the prevalence of disability among older Americans (as measured by limitations in ADLs) for 1994 (21.3 percent) was 3.6 percent lower than the 1982 standardized rate (24.9 percent). According to these researchers, the number of elderly Americans with a serious chronic disability increased from 6.4 million in 1982 to 7.1 million in 1994. But if the elderly disability rate had remained at the 1982 levels, the 1994 number would have been 8.4 million. This research team estimated that the 1982-94 disability decline reduced the institutionalized population by 400,000, saving $17.3 billion in spending on nursing homes in 1994. Using the constant 1982 disability rate, the estimated number of disabled elders in 2030 would be 17 million; with the constant estimated 1996 disability rate, the number would be 14.6 million; and based on a 1.5 percent annual decline, the number would be 8.5 million.
Geographic Diversity
While the estimates of the population aging are usually presented for the nation as a whole, the number of elderly and its impact on the demand for long-term care vary tremendously across states. As the baby boomers start to reach retirement age in 2011, the size of the 65 and over populations is projected to increase in all states (Campbell, 1997). California and Florida continue to rank first and second, respectively, in having the largest number of elderly residents. By 2025, Texas will rank third, passing New York and Pennsylvania.
In 1995, Florida had the largest proportion of elderly (19 percent), and Alaska the smallest (5 percent) (Table 2). By 2025, Florida is expected to remain the oldest state with more than 26 percent of its population age 65 or older. Alaska will rank as the youngest state with 10 percent elderly.
The rapidity of growth in the elderly is further illustrated by the fact that only 5 states had at least 15 percent of their population age 65 or older in 1995. By 2025, 48 states will reach or exceed that proportion. Only Alaska (10 percent), California (13 percent), and the District of Columbia (14 percent) will not meet or exceed the 15 percent level.
Changes in the concentration of elderly are affected by several factors that have different implications for long-term care demand in the future. The migration of working-age people affects age concentration more than elderly migration (Treas & Longino, 1997). First, there is the natural increase in the size of the non-elderly population which affects the proportion of elderly in a state. Utah, for example, had a relatively low proportion of elderly in 1994 (9 percent), due to an unusually high fertility rate and many young newcomers. The net-migration of the non-elderly population also influences the proportion of elderly in different ways. Iowa, for example, ranks fifth in the proportion of elderly (15.2 percent), due in large part to the out-migration of younger people. Consequently, elderly Iowans may face significant long-term care problems because of the shrinking of their potential informal caregiver pool. In contrast, Alaska has a very small proportion of elderly because many of the newcomers are young.
Elderly tend to remain in place relative to the nonelderly (Treas & Longino, 1997), although in general, the U.S. is a country of "non-movers." In 1994, 94.4 percent of those 65 or older compared with 81.7 percent of the under 65 were stable (Table 3).
The movers in both age groups were most likely to remain in the same county. Interstate migration rates have remained stable over four decades, with this pattern concentrated in the early years of retirement (Longino, 1998). Almost 2 million people age 60 or older moved from one state to another during the five-year period ending in 1990. Florida was the highest ranking state, receiving 23.8 percent of the in-migrants, followed by California (6.9 percent), Arizona (5.2 percent), and Texas (4.1 percent) (Table 4).
The 1990 census data show a changing pattern of retiree out-migration. While it is generally assumed that the major origin states are all in the "rust belt," in 1990, fewer migrants were leaving New York, Illinois and Ohio than in 1980. And surprisingly, some sunbelt states such as California and Florida were rising to the top of the list of sending states (Table 5).
If the rate remains stable and most of the movers are recent retirees, the number of elderly making long moves should grow very little. The median age among current interstate movers is in the mid-60s. If that median age increases over time, this could have an impact on the demand for long-term care, particularly for the states where an influx of retirees occurs. It is also important to recognize that the young retirees "age in place" in states like Florida and Arizona, placing additional burdens on long-term care resources over time. A countervailing trend is the "return migrant" phenomenon in which older, widowed and less healthy elderly return to their state of origin (e.g., disabled elderly who have moved to Florida as young retirees returning to the Northeast when they need long-term care). Litwak and Longino (1987) referred to amenity-seeking elderly as young retirees moving for certain features such as better climate. In contrast, assistance-based migration often follows the onset of a disability. It is projected that during the 20 years following 2030, amenity migration will decline and be supplanted by an assistance-based migration era. These patterns will influence both the future financing and design of long-term care systems in the U.S., including the amount of formal services that may be needed to replace a diminishing caregiver pool in certain states. It also underscores the point that, as is true today, the costs of long-term care will not be distributed evenly across the states.
The Future of Informal Caregiving
There are a number of factors that will influence the size and nature of the supply of informal caregivers in the 21st century. The most important predictor of a strong informal network is being married. In 1995, 73.4 percent of males aged 65 or older were married; the projected estimates for 2030 and 2050 are 70 percent and 66.8 percent, respectively (Table 6) (NAIC, 1996). The pattern for males aged 85 and over is somewhat different. Between 1995 and 2030, the proportion of married males in the old-old category is projected to increase from 50.5 percent to 56.5 percent. By 2050, however, that number is projected to decline again to 52.4 percent. Elderly females are significantly less likely to be married today and in the future, but the percentage with spouses increases over time. In 1995, 39.4 percent of females aged 65 or older were married; the comparable estimates for 2030 and 2050 are 44.1 percent and 41.5 percent, respectively. The percent of very elderly females in the married category increases from 13.6 in 1995 to 15.3 percent in 2030 to 16.6 percent in 2050.
One gross measure of the availability of informal caregivers is the ratio of the population in the average caregiving range, ages 50 to 64, to the population aged 85 and older. In 1990, that ratio was 11 to 1; by 2050, there will be only four potential caregivers for every elderly person (RWJF, 1996). This estimate, however, does not include the vast number of elderly spouses (particularly wives) with primary caregiving responsibilities, or the increasing number of young-old children caring for their old-old parents.
Another important predictor of the strength of the future caregiver population is the projected family structure of older persons (Himes, 1992). In 1990, 41.5 percent of white males age 85 or older were married with at least one child and another 11.4 percent had spouses but no children (Table 7). Another 37 percent were unmarried with at least one child, leaving 10.2 percent the most likely to have no informal supports. By 2020, it is projected that 4.8 percent of white old-old males will have neither spouse nor child. In contrast, among white females aged 85 or older in 1990, only 7.3 percent were married with at least one child, 2.4 percent were married but childless, and 68.3 percent were unmarried with at least one child. Twenty-two percent-one in five very elderly females-had no close family members as potential caregivers. By 2020, the situation improves dramatically for women age 85 and over; with only 8.7 percent projected to be unmarried and childless. The potential informal caregiver pool is much smaller for black than for white elderly, but the potential availability of family members does increase over time for this population. Among black males aged 85 or older in 1990, 17.6 percent were unmarried and childless; the comparable projection for 2020 is 9.2 percent. Similarly, the percentage of very elderly females without spouse and children is projected to decrease from 31.5 percent in 1990 to 14 percent in 2020.
One subpopulation of elderly may be most likely to use formal services-individuals who live alone and have no living children or siblings. Researchers have estimated that 1.2 million people aged 65 or over will be in that status in 2020, up from 682,000 in 1990 (NAIC, 1996). The 288,000 people who are 85 or over and living alone without close kin are the most likely candidates for formal long-term care service use.
The projections of family structure tell only part of the story. Researchers have observed, for example, that while daughters take on greater parent-care roles, their efforts adjust to both parental need and the need of others, including their own children (Wolf et al., 1997). As women delay childbearing, they will be more likely to try juggling significant childrearing and elder care responsibilities, resulting in a potential diminution in the amount of caregiving hours provided to the disabled parent. On the other hand, recent research also suggests that a given adult child's own care efforts are reduced, but on much less than a one-for-one basis, as the efforts of siblings are increased. The more children a frail parent has, the greater the volume of help from children the parent will receive. The Depression-era mothers, this cohort of elders aged 85 or older, had relatively low fertility rates. At present, and for the near future, the parents of the baby boom generation are passing through old age with, on average, a rich pool of family resources as measured by simple numbers of surviving offspring and in-laws. During the first half of the next century, however, this situation will reverse as we expect that older people will have few surviving children, on average.
Labor force participation of women, both those likely to be caring for elderly parents and elderly spouses, will also influence the availability of informal caregivers in the future. Female labor force participation rates, particularly among the age cohorts most likely to be caring for elderly parents, since the 1970s. Among those aged 45 to 54, the rate is projected to increase from 75.4 percent in 1996 to 79.9 percent in 2006; the comparable figures for the 55 to 64 year olds are 49.6 percent and 55.8 percent, respectively (Fullerton, 1997). The nature of informal assistance may change, including an escalation in long-distance caregiving. An estimated 3.3 million baby boomers are currently providing long-distance care and this number will more than double over the next 15 years (National Council on Aging, 1997a).
One of the uncertainties of the 21st century is the extent to which new family structures will affect the size and character of the pool of informal caregivers. With increases in divorce and remarriage, many elderly individuals are finding themselves with a range of step in-laws and step-grandchildren. It is not clear, however, whether this trend will increase or decrease the availability of caregivers. On the one hand, divorces may estrange families and prove to be a deterrent to caregiving. On the other hand, the new configurations of families provide a potential expansion of the pool of caregivers in the future.
The Economic Status of the Future Elderly
One of the major determinants of the future market for long-term care is the extent to which future generations of elderly will be able to afford these services. One measure of the economic status of the elderly population and their potential use of formal services is educational attainment; more highly educated individuals are likely to be wealthier and also have a predilection for purchasing care. The 65 and over group will clearly be more educated in the 21st century (Besl & Kale, 1996). In 1997, 40 percent of those aged 75 and over had less than a high school education; only 13 percent had a bachelor's degree or higher. In contrast, among those 45 to 54 years old, only 13 percent had less than a high school education; 28 percent had at least a bachelor's degree (Day & Curry, 1998).
The future educational profile of the elderly will be quite different from the observed 1990 profile (Figure 5) (Hobbs & Damon, 1996). In 1990, nearly half the elderly population had not completed high school; in 2030, more than four out of five elderly (83 percent) are projected to have completed high school or more. Those 65 and over with at least a bachelor's degree are estimated to increase from 11 percent in 1990 to almost one in four (24 percent) in 2030. The proportion of college educated males is expected to increase from 14 percent to 26 percent. The increase for females is even more dramatic, up from 8.5 percent to 22 percent. Future improvements in the levels of educational attainment among the elderly will be slower for blacks and Hispanics than for whites.
Today, only one in every eight older people has incomes below the official poverty line (Smith, 1997), although after adjustments for the value of non-cash benefits such as food stamps and Medicaid and the implicit rental value of housing, only one in every 20 of the elderly is poor today. This should not obscure the reality that many elderly remain economically vulnerable. For example, one in every four nonmarried older women is poor and among elderly widows, poverty rates run as high as 40 percent. Seven out of ten older black single women live below the poverty line.
There is an enormous amount of inequality among older people, far more than exists between them and the rest of the American population (Smith, 1997). While the median household income among the elderly aged 70 and over is $15,624 in 1996 dollars, the comparable figures for those over age 85-the population most likely to need long-term care-is $9,439 (Table 8). The average household over age 70 has less than $9,000 in financial assets (defined as total wealth minus housing and other real property assets); those in the bottom 10 percent have no financial assets at all while those in the top five percent have more than $300,000.
Home ownership is an important form of wealth for the elderly. It is also an important part of the long-term care system, in that most people prefer to remain in their own homes rather than being institutionalized. In 1997, 79.1 percent of those aged 65 or over owned their homes; among the 75 and older group the figure is 75.8 percent (U.S. Bureau of the Census, 1998a). The highest ownership rates were among the 55 to 64 year olds, where four out of five were home owners. In contrast, only two-thirds of the 35 to 44 year olds were homeowners in 1997. This trend has potential implications for several aspects of long-term care in the future. Since one of the deterrents to spending down to Medicaid is the loss of one's house, the fact that a larger proportion of the future elderly population may not own their homes may enhance the appeal of Medicaid as a viable financing option. This trend also suggests that there will be less opportunity for future elderly individuals to use the equity in their homes to purchase care (although the use of reverse annuity mortgages and other such mechanisms have never really taken off in the United States). Finally, a decrease in home ownership among the senior population may create a greater demand for residential care alternatives in the future.
Assuming reasonable rates of growth in the economy, it is projected that boomers will have higher real incomes in retirement than today's retirees (Manchester, 1997). However, it is also important to emphasize that the generation now beginning to retire (people born between 1930 and 1940) has had a unique set of economic experiences (Friedland & Summer, 1999). They were in the labor force during a period of strong economic growth in the 1960s, saw their home values increase dramatically in the 1970s, and may have begun saving for retirement early enough to benefit from the high interest rates in the late 1970s and the stock market booms of the early 1980s and mid 1990s. They were more likely than workers in previous generations to participate in insured pension plans with vested rights to pension benefits. People 20 years younger did not directly benefit from these circumstances. The growth in real wages slowed during their entry into the labor force, they paid higher real prices for homes and did not benefit from the 1980s stock market expansion. They are less likely than the preceding generation to have defined-benefit pension plans with employer contributions. At the same time, they had more schooling, less children and are more likely to have a working spouse than their predecessors. These circumstances may provide financial advantages in retirement not experienced by the current generation of retirees.
Contrary to the common wisdom, most baby boomers have relatively few financial nonhousing assets. According to data from the 1993 Survey of Income and Program Participation, those with incomes at the 80th percentile had only $25,300 in financial assets (Figure 6) (Rother, 1997). For those at the 90th percentile, the average financial assets were $66,000. Furthermore, the financial assets held by the typical boomer were worth only $1,000, and only one fifth of boomers had more than $25,000. The lowest one-quarter reported negative financial assets-their liabilities exceeded their financial assets. With respect to future financial security, a study conducted by Lewin-VHI for AARP in 1994 estimated that three-fourths of baby boomers will have all three major sources of retirement income-Social Security, pensions, and assets-in 2030. However, from 49 percent to 60 percent (depending on the Gross Domestic Product growth assumptions) of retirees in 2030 will depend on Social Security for half or more of their income.
The incomes of the baby boom generation are more unequal than those of preceding generations. They have been subject to the economic pressures of slow wage growth and increased income inequality since the 1970s because they entered the labor market when wage inequality was more pronounced. These inequalities will continue into old age, raising the spectrum of more economic disparity among future generations of elderly. Given the correlation between disability and lower financial status, the very people who need long-term care may be less likely to afford these services in the future relative to the current cohort of elderly.
The baby boomers' perceptions of future long-term care needs may influence the way in which they prepare financially. A recent survey of Americans aged 42 and over (Greenwald, 1998) found that only 12 percent feel it is very likely that they or their spouses will require long-term care; only 11 percent feel it is very likely that they will require assistance with daily activities. There also continues to be a lack of knowledge about who pays for long-term care. Almost three out of four respondents to a recent survey (National Council on Aging, 1997b) believe that Medicare is the primary funding source for most older persons' services, and 48 percent report having done little or no long-term care planning.
These trends will converge to influence the magnitude, nature and scope of the demand for long-term care in the future. It is not clear how all of the interactions will play out, but several major themes have emerged from this discussion.
- The aging of the United States and, in particular, the growth in the proportion of the 85 and over population, will place increasing demands on our fragmented long-term care system. The extent to which declining disability rates may mitigate some of this effect is uncertain.
- Long-term care will continue to be a woman's issue from both the care recipient and caregiver perspectives.
- The face of the long-term care population will become more racially and ethnically diverse in the next half a century, suggesting the need for policies and programs that are more culturally sensitive and appropriate.
- Aging is not a homogenous phenomenon across the country; some states and substates have already reached or surpassed the one-in five elderly people projected to live in the United States thirty years from now. Changing migration patterns of both the young and the old will continue to produce states and regions that are much older and have greater demands for long-term care than others.
- Elderly cohorts of the future will be more highly educated and, on average, more financially secure than today's cohort. This improvement, however, belies the subpopulations who will be financially unprepared to bear the costs of long-term care in the next century.
Outline
| I. Introduction
| II. Defining Long-Term Health Care
| III. The Three Legged Stool of Long-Term Care Policy
| IV. Trends in Long-Term Health Care Delivery
| V. Workforce Issues
| VI. The Future of Long-Term Care Demand
| VII. The Future Supply of Long-Term Health Care Services
| VIII. Sinking or Swimming Into the Future?
| IX. Conclusion
| X. References
| XI. Author