Aging into the 21st Century

Many factors affect the need for special services by the elderly, participation in programs for the elderly, and the type and amount of support to which elderly persons have access. The list is long and the interrelations are complex. The tables presented offer data on only some of these important variables:

• Marital status
• Children
• Living arrangements
• Household status
• Education
• Labor force participation and economic dependency
• Income and poverty

Living with a spouse is a primary factor contributing to the support and independence of the elderly. The majority of elderly men, but only a minority of elderly women, are currently married and living with their spouses-a pattern that is expected to continue over the next several decades.

Projections of the Bureau of the Census show in Table 6a divide marital status broadly into "never married" and "ever married." Within "ever married" are two categories, "married, spouse present" and "other"-a large category that includes widowed, divorced, and married but not living with spouse. According to projected estimates of the marital distribution of the 1995 population, about 5 percent of elderly men have never married, about 71 percent are married and living with their spouses, and about 24 percent are widowed, divorced, or married and not living with their spouses. The percent of never-married women is somewhat higher than for men, but the striking difference is in the distribution between "married, spouse present," and the "other ever-married" statuses. A little more than one-third of the women are married and living with their spouses, while nearly three-fifths are in the "other" group-most being widowed, and the rest either divorced or married but not living with spouses. According to the census projections, by 2010 the proportions will not experience significant change.

SOURCE: U.S. Bureau of the Census (1996b).

Table compiled by National Aging Information Center

In contrast to the overall elderly population, the percentages of men and women aged 75 and over, married and living with spouses, are substantially lower, and the percentages reported in the "other" category are substantially higher, with most being widowed. Only three-fifths of the men and one-fifth of the women 75 years and over are currently married and living with spouses. Nearly three-fourths of the women were previously married, but are unmarried now. Age and sex differences in mortality, in particular the premature mortality of men, contribute significantly to the sex differences in the marital distribution in later life. Because differences in mortality are expected to continue but converge somewhat, the marital distribution of elderly people is also expected to change. By 2010, the "married spouse present" status will tend to rise for men and women by nearly two-thirds and one-quarter, respectively.

The projections of the marital status of the population prepared by the SSA are consistent with those of the Bureau of the Census through the year 2010, the terminal year of the census projections (see Table 6b). The SSA projections presented in this report extend to 2050. Because the SSA figures on the percent married include absent spouses, they are somewhat higher than those of the Bureau of the Census. In 1995, about three-quarters of the men, and two-fifths of the women 65 and over, were married. After 2010, the proportion of married men is expected to decline substantially, from 73 percent in 2010 to 67 percent in 2050. The proportion of married women is expected to rise from 39 percent in 1995 to 44 percent in 2030 and then decline to 42 percent in 2050.

Alternative II of Social Security Administration.)

SOURCE: Social Security Administration (1995).

¹ Married includes those married with spouse absent.

By ages 85 and over, many married persons have died, leaving numerous widows and some widowers. Consequently, in 1995, the proportion married is sharply lower in this oldest age group, 51 percent for men and 14 percent for women. The proportions married at ages 85 and over for both men and women are expected to fluctuate between 1995 and 2050, with small net gains by 2050.

There are interesting variations on the basis of race in the proportions of married elderly persons expected in the years to 2000 (Himes, 1992). Paralleling the marital distribution of the general population, over three-quarters of white men aged 65 and over were married in 1995 and will be married in 2000. About two-fifths of white females aged 65 and over were married in 1995 and will be married in 2000. The patterns for the older aged are about the same, but the marital proportions are much lower. According to Himes (1992), for these ages, somewhat over one- half of the white men, and one-tenth to one-eighth of the white women are married. The figures for blacks are much lower than for whites for both of these age ranges.

Back to top

Children, after spouses, are the next most important potential source of social support for elderly people. About one-third of white women and nearly one-quarter of black women aged 65 and over were married and had at least one child in 1995. Himes (1992) projected little change in these proportions from 1995 to 2000. Forty-seven percent of white elderly women were unmarried and had at least one child, and this proportion is expected to increase to only 48 percent by 2000. Fifty percent of black unmarried women aged 65 and over had at least one child, and that figure is expected to reach 52 percent by 2000. Linking the Himes projections for 2000 with projections made by Siegel and Ghadialy (1993b) for 2010, we can anticipate a large increase in the percentage of elderly white women with at least one child in the current decade, from 79 percent in 1990 to 85 percent in 2000, and then a small rise to 86 percent in 2010. The large increase from 1990 to 2000 reflects not only the continuing decline in the mortality of children and young adults but also the rise in fertility and marriage during the baby-boom era. The figures for elderly white men are very similar to those for elderly white women. The proportions of black men and women (especially black men) who have at least one child are much lower than for whites. In the final analysis, these figures suggest that a great majority, but by no means all, of elderly women will have a child available for support in 2010.

Two groups-elderly unmarried men with no children, and elderly unmarried women with no children-are small and declining. The percentage of unmarried elderly white women with no children declined from 17 percent in 1985 to 13 percent in 1995, and is expected to decline to 12 percent in 2000. For black women it is expected to be 20 percent in 2000, down from 26 percent in 1985 and 22 percent in 1995 (Himes, 1992).

Married elderly men and women without children may be able to depend on a spouse for support, but only as long as the spouse does not die or become disabled. This group is smaller than the elderly unmarried group and its numbers have been declining even more rapidly. Nine percent of white men and 12 percent of black men will belong to this group in 2000; for women the figures are 4 and 5 percent, respectively.

For persons aged 85 and over, the percentages of men and women married with children are much lower than for the age group 65 to 84 (Himes, 1992). The percentages of both married and unmarried persons with at least one child are projected to increase markedly between 1990 and 2020, while the percentages for those without children are projected to decline significantly. This pattern is consistent with the higher survival rate for women who have children, but the trend is probably more affected by the general rise in marriage and fertility since the 1930's. Only about 8 percent of the 85-and-over age population will be unmarried and without children in 2020.

Back to top

As stated previously, nearly three-quarters of elderly men are married and live with spouses, but only one-third of elderly women are married and live with spouses. About 31 percent of all elderly lived alone in 1990, and four-fifths of these were women (Table 7). Since most elderly men are married and most elderly women are not married, elderly women are more likely to live alone. One in eight elderly live with other relatives, not including a spouse. There has been a significant shift toward solitary living in recent decades. The Lewin/ICF projections show little change in the proportion of elderly persons living alone between 1990 and 2020, but it is still possible that a substantial increase could occur. Many elderly persons choose to live alone if their health and finances permit. This pattern reflects the desire of most elderly to be independent. Some of those who live alone have a child or children living nearby. Others have a child or children living at a distance who regularly keep in touch with them.

SOURCE: 1990 data-U.S. Bureau of the Census (1991). Projections-Lewin/ICF (1990) estimates based on data from the Current Population Survey and the Brookings/ICF Long Term Care Financing Model.

Table compiled by National Aging Information Center

Solitary living increases with advancing age. In 1990, 47 percent of those aged 85 and over lived alone; projections show that 45 percent will live alone in 2020. The proportion of Hispanics living alone in 1990 (22 percent) was considerably lower than for whites (31 percent). While roughly 31 percent of all elderly lived alone in 1990, according to the Lewin/ICF projections, only about 8 percent lived alone and had no living children (Table 8). The figure is expected to remain at about this level until at least 2020. The figure for blacks and Asian and Pacific Islanders is substantially higher (11 percent) than for whites (8 percent), as a result of the higher percentages of blacks without living children (Himes, 1992). Furthermore, only about 2 percent of all elderly lived alone and had no living children or living siblings.

White, Hispanic, and Black and Others are defined as mutually exclusive categories.)

SOURCE: Lewin/ICF (1990) estimates based on data from the 1984 Supplement on Aging (SOA); the Brookings/ICF Long Term Care Financing Model; and estimates by author.

¹Percent living alone with no living children out of the population 65 years and over living alone (in each race/Hispanic group).

²Percent living alone with no living children out of the total population 65 years and over (in each race/Hispanic group).

Table compiled by the National Aging Information Center

SOURCE: Lewin/ICF (1990) estimates based on data from the 1984 Supplement on Aging (SOA) and the Brookings/ICF Long Term Care Financing Model.

Table compiled by National Aging Information Center

Estimates of the proportion of elderly who live alone, with no living children or siblings, among all those who live alone, vary from 11 percent (Supplement on Aging, National Health Interview Survey, 1984) to 7 percent in 1990 (Lewin/ICF estimate based on the same data; see Table 9). The proportions of elderly with no living children or siblings are higher for persons living alone than for persons living in married couples or with others, and increase with advancing age. Approximately 12 percent of persons aged 85 and over living alone have no living children or siblings. According to the Lewin/ICF series, the proportions in all age groups and living arrangement categories are not expected to change by 2020.

Back to top

Changes in marital status and living arrangements result in changes in the numbers and types of households. "Headship ratios" (that is, households per 100 population) increase rapidly with increasing age up to about age 50 because of new family formations and remarriage of divorced persons (Table 10). They continue to increase thereafter in spite of rapidly rising spousal mortality rates as the surviving widowed men and women continue to maintain their own homes. As a result, the peak headship ratio occurs at the highest ages, shown in Table 10 as 64 households per 100 population at ages 75 and over in 1995. Census data indicate, however, a decline within the oldest age groups as householders move in with close relatives or into group quarters when health fails or income falls too low. The projections of the Bureau of the Census, released in 1996, imply modest changes in headship ratios by 2010, with slight gains at the middle and older ages.

SOURCE: U.S. Bureau of the Census (1996b).
Table compiled by National Aging Information Center

Households maintained by elderly persons (Table 11) are mostly either households maintained by married couples (45 percent) or households maintained by a woman with no relatives present (35 percent). Married couple households are much less common, and female-headed nonfamily households are much more common at the older ages than at younger age groups. This shift is evident even within the older ages. As the age of the householder increases from 65 through 74 to 75 and over, the proportion of married couple households decreases and the number of female-headed nonfamily households increases. Male-headed family and nonfamily households are much less common among the elderly. In fact, the former are rare. The Bureau of the Census expects no dramatic change in the distribution of types of households among the elderly between 1990 and 2020.

SOURCE: U.S. Bureau of the Census (1996b).
Table compiled by National Aging Information Center

One-fifth of all households are maintained by a person or persons aged 65 and over, and this proportion is expected to increase slightly between now and 2010, as is expected for the overall proportion of elderly in the population. These proportions can be very different for different types of households. Nearly one-half of all nonfamily households headed by a woman had an elderly head in 1995, but only one-sixth to one-seventh of the other types of households were headed by an elderly person (Table 12). With advancing age, the proportion of female-headed nonfamily households maintained by elderly persons increases, while the proportion of the other household types maintained by elderly persons tends to decrease. Changes in these proportions between now and 2010 are expected to be negligible or small.

SOURCE: U.S. Bureau of the Census (1996b).
Table compiled by National Aging Information Center

Back to top

The proportion of high-school graduates among the elderly will increase sharply in the decades to come as the more educated younger cohorts age. In 1995, 64 percent of the elderly were high-school graduates (Bureau of the Census, 1996c), and the proportion is expected to increase to over three-quarters in 2010 and to nearly seven-eighths in 2020 according to Siegel (1993a). The proportion of high school graduates remains approximately the same within a particular cohort as it ages. Consequently, the population aged 75 and over will have about the same percentages of high school graduates in a given future year as did the overall elderly population a decade earlier, and with a lag of 20 years, the population aged 85 and over will show about the same percentages of high school graduates as did the overall elderly population a few decades earlier.

In spite of these increases in education overall, the elderly remain the least educated age group in our society. In addition to the 36 percent of the elderly who have not completed high school, an additional small proportion of elderly, especially Hispanic elderly, have limited facility in the English language, even though they may have completed high school.

Back to top

Trends in labor force participation among the elderly will have an important impact on economic support ratios and dependency. They affect:

• the proportion of total income derived from earnings and the proportion allocated to savings;

• the accumulation of credits for annuities, pensions, Social Security benefits, and Medicare;

• the balance of workers and nonworkers; and

• preemption of time and energy of adult children for support of dependent parents.

Labor force participation drops sharply for both men and women with increasing age from ages 55 to 64 (and even earlier) through the oldest ages (Table 13). Two-thirds of the men and one-half of the women now work at ages 55 to 64, but relatively few men or women are still working at ages 75 and over. According to the latest labor force projections from the Bureau of Labor Statistics, labor force participation ratios for men at these ages will either remain unchanged or decline slightly, but those for women will continue to rise. (See also SSA series, 1992, and Urban Institute, 1989.) By themselves, these patterns do not determine a particular pattern of change in the median age at retirement for either men or women, but they suggest a halt in its historical decrease and, possibly, a stabilization. A declining trend in median age at retirement has been evident over the last 4 decades for both men and women, but the trend has leveled off somewhat since the 1970's (Gendell and Siegel, 1992 and 1996).

SOURCE: Unpublished data provided by H. N. Fullerton, Jr., U.S. Department of Labor, Bureau of Labor Statistics, Office of Employment Projections. See also: Fullerton, H. N. Jr. (1995). The 2005 labor force: Growing, but slowly. Monthly Labor Review, 118 (11), 29-44.

¹Hispanics may be of any race.

Table compiled by the National Aging Information Center

The projected shifts in labor force participation, median age at retirement, and longevity, along with past and projected shifts in fertility, foreshadow pronounced changes in the relative numbers of nonworkers and workers in the next half century. For example, the ratio (per 100) of nonworkers 55 years of age and over to all workers is expected to rise from 31 in 1995 to 35 in 2010 and then to 52 in 2030 (Siegel, 1993a). Currently, there are more than three workers for every older nonworker, but it is projected that in 2030 there will be less than two workers for every older nonworker. Most of these older nonworkers are women. The economic support ratio for women will fall from over 5 workers of both sexes for every female nonworker aged 55 and over in 1995 to 3.25 workers for every female nonworker aged 55 and over in 2030.

Historically, dependent or nonworking members in a household have mostly been children. After 2010, these so-called dependents will more often be elderly. Children and even many adults under age 55 who do not work generally have to be supported. The proportion of children, virtually all of whom are not working and are therefore dependent, is expected first to decline and then to increase within narrow limits, and to be nearly equaled by the number of elderly by 2030. Moreover, there is a large nonworking population aged 16 to 54, which, together with the nonworking population aged 55 to 64, exceeds the number of nonworkers 65 and over in the period up to 2010. By 2030, there will be far more nonworkers aged 65 and over (39 to 40 per 100 workers) than nonworkers aged 16 to 64 (30 per 100 workers).

In planning the financial operations of the Social Security system, the SSA has regularly charted the ratio of the number of beneficiaries under Old Age Survivors and Disability Insurance (OASDI) to the number of covered workers. The ratio of beneficiaries to 100 covered workers was 30-31 in 1995 in the SSA projection series II (Table 14). Compared with the level in 1995, the ratio is expected to increase sharply to 49 by the year 2030 and then to 51 in 2050. Alternatives I (Low Cost), II (Intermediate), and III (High Cost) represent different OASDI cost rates and demographic shifts. The series vary from high fertility and small reductions in death rates (Low Cost) to low fertility and substantial reductions in death rates (High Cost). The projected dependency ratios range from 43 to 56 in 2030; thereafter, the Intermediate series and the High Cost series, but not the Low Cost series, continue to increase.

SOURCE: U.S. 104th Congress, 1st session (1995).

Table compiled by National Aging Information Center

There is a strong correlation between the cost of the OASDI programs and changes in population age patterns. Hence, the pattern of the annual cost rates of the system is similar to the pattern of the annual ratios of beneficiaries to workers. The OASDI cost rates are expected to rise rapidly after 2010 because the number of beneficiaries is expected to rise more rapidly than the number of covered workers. This will occur because the relatively large number of persons born during the period of high fertility rates after the end of World War II through the mid-1960's will reach retirement age, while the relatively small number of persons born during the subsequent decades, 1965 to 2005, will comprise the labor force.

Back to top

The income situation of the elderly, on average, is relatively favorable, and the extent of poverty is less than among the rest of the population (U.S. Bureau of the Census, 1996d). Yet, there is a wide dispersion in the distribution of income and assets among the elderly population. A closer look reveals a pronounced economic disparity among subgroups within the elderly population. Many elderly fall either below the poverty level or have a yearly income that is less than 200 percent of the poverty level. Women, blacks, persons living alone, very old persons, those living in rural areas, and especially persons with a combination of these characteristics are subject to living in poverty to a disproportionate degree (Siegel, 1993). However, more and more women in the older age groups as well as in the younger ones have been entering the labor force. As a consequence, these women are accumulating credits for annuities and pensions, as well as Social Security benefits and Medicare coverage, in their own right. Failure to have a pension has been linked directly to the probability of falling into poverty in old age.

In 1994, the median income of households maintained by persons aged 65 and over was $18,095, while the median for all households was $32,264 (U.S. Bureau of the Census, 1996d). This 44 percent difference reflects a real decline of household income from its peak at earlier ages. Yet, it overstates the effective difference because of the smaller size of older households, the greater noncash benefits received by the elderly, and probably a greater underreporting of money income by older persons (Siegel, 1993).

Projections of income to 2020 or 2030, in comparison with current estimates for 1990, have been prepared by Rivlin and Weiner (1988), the Urban Institute (1989), and Lewin/ICF (1990). These sets of data reflect the decline of income and the increase in poverty with advancing age in the older age groups. They also reflect the considerable deficit in income of unmarried persons in comparison with married couples. The projections themselves are based on rather optimistic assumptions of the trend in per capita income and are designed to serve as a foundation for projecting demand for public services and entitlements. Rivlin and Weiner projected an 85 percent increase in real median income for "elderly families" between 1986 to 1990 and 2016 to 2020, and a 17 percent increase for families maintained by persons aged 85 and over. Family income is defined as joint income for married persons and individual income for unmarried persons. The Urban Institute also projected increases in real median family income. The income of married couples would increase by 57 percent between 1990 and 2010 and by 50 percent between 2010 and 2030. The income of unmarried men, beginning at a level 54 percent lower than for married couples, would increase by 51 percent and 55 percent in those periods, and the income of unmarried women, beginning at a level 61 percent lower than for married couples, would rise by 35 percent and 59 percent in those periods.

According to the Lewin/ICF series, in 1990, about 17 percent of the elderly living in the community (as opposed to nursing homes) had incomes less than 100 percent of the official poverty level, and about 48 percent had incomes less than 200 percent of the official poverty level (Table 15). Because of the assumptions of rising per capita income, these percentages are projected to drop sharply to 7 and 25 percent, respectively, by 2020. Since considerably more elderly who live alone live in poverty, compared to those who live with others, we will focus on persons aged 65 years and over who live alone and are poor. In 1990, the poor, defined for this purpose as having incomes under 125 percent of the official poverty level, constituted 40 percent of the elderly living alone but only 21 percent of the elderly living with others (Table 16). The poor comprised only 35 percent of those aged 65-74 living alone, but 49 percent of those aged 85 and over living alone. The poverty ratio was notably lower for whites than for Hispanics and "blacks and other races." The proportion of the poor among the white elderly living alone was 36 percent, among the Hispanic elderly living alone 65 percent, and among the black and other elderly living alone 69 percent. The poverty ratio for those living with others was far more favorable in all age and race categories, except ages 85 years and over. Because of the underlying economic assumptions, it is expected that the percentages of those living in poverty among the elderly living alone in 2020 will be well below those of 1990, by more than half in several categories.

SOURCE: Lewin/ICF estimates based on data from the 1984 Supplement on Aging (SOA), the Current Population Survey, and the Brookings/ICF Long Term Care Financing Model.

¹ Impairment is defined as having difficulty with at least one of five Activities of Daily Living (ADLs)-eating, bathing, dressing, transferring, and toileting.

² The current estimates of the U.S. Bureau of the Census are: Less Than 100 Percent, 12; 100-149 Percent, 14; 150-199 Percent, 13; and 200 Percent or More, 61. Source: U.S. Bureau of the Census (1991b).

x = not available

Table compiled by National Aging Information Center

SOURCE: Lewin/ICF (1990) estimates based on data from the Current Population Survey and the Brookings/ICF Long Term Care Financing Model.

¹ White, Hispanic, and Black and Other are defined as mutually exclusive categories.

Table compiled by National Aging Information Center

Back to Previous | Main | Next >

Visitor’s Guide | Contact Us | Feedback | Privacy Notice | Disclaimer | Accessibility | Sitemap | A–Z Index | FOIA | Plain Writing Act | USA.gov | HHS.gov