The United States is undergoing a profound change as the number and fraction of older persons in the population rise. Similar changes are occurring around the world as an inevitable consequence of longer life and fewer births per woman. At the same time, the meaning of chronological age has changed. In terms of health and vitality, many “old” people today are functionally similar to younger persons living a few decades ago. A rising proportion of those in their seventh or eighth decades of life does not necessarily mean that more people are unable to work or care for themselves. In this important sense, the part of population aging due to longer life is not a problem; rather, it reflects revolutionary changes in health and longevity of the population (the part of population aging due to lower birth rates is a different matter).
Nonetheless, many of our policies, institutions, and behaviors are structured by chronological age, with age 65 having played a particularly important role in the last half century or so. If Americans continue to retire around age 65, then longer life and population aging will indeed be costly. As the share of the population 65 and over grows, there will be fewer workers to support them. Although we expect productivity to continue to grow and living standards to rise, if the population does not work longer and retire later, we will have to set aside a greater share of our earnings to provide for old age. Since changing this demographic trend is not a realistic option, we cannot continue our economic behaviors and policies as in the past. We must adjust to the changing demographic realities, including lower birth rates, a workforce that is growing more slowly, and longer lives.
Our options are few: We must consume less, work more, or both. There are four basic approaches for adapting to the new economic landscape created by an aging population, and for providing the resources to support the consumption of households in their later years:
- Workers save more (and consume less) in order to prepare better for their retirements.
- Workers pay higher taxes (and thus consume less) in order to finance benefits for older people.
- Benefits (and thus consumption) for older people are reduced so as to bring them in line with current tax and saving rates.
- People work longer and retire later, raising their earnings and total output.
Higher saving rates for the working age population would reduce their current consumption but enable them to rely less heavily on public benefits for pensions and health care in the future. Of course, higher saving rates for the young will not help pay for the benefits of the current elderly. (For some basic definitions, see Box 2-1.) Raising payroll taxes would reduce consumption by the working age population, making it possible to pay more benefits to the current and future elderly. Alternatively, costs could be shifted to the elderly by cutting their benefits or raising their taxes as a condition for receiving the benefits. A final option is to increase the size of the
There is increasing recognition that the term “elderly” is an inadequate generalization that obscures the diversity of a population age group that spans more than 40 years. Nevertheless, for purposes of comparison—over time, across countries, and sometimes over different variables—some chronological demarcation of age categories is necessary. This report uses the term “elderly” to refer to people aged 65 and over.
The “developed” and “developing” country categories used in this report correspond directly to the “more developed” and “less developed” classification used by the United Nations. Developed countries comprise all nations in Europe (including some nations that were part of the former Soviet Union) and North America, plus Japan, Australia, and New Zealand. The other nations of the world are classified as developing countries. While these two broad categories commonly are used for comparative purposes, it is increasingly evident that they often do not accurately reflect developmental differences between nations.
labor force, possibly by raising labor supply during the traditional working years but more likely by extending time in the labor force and retiring later.
The fundamental issue that society faces is how to allocate the increased costs of population aging across these four sources. Each of the options has different implications for which age groups and which generations will bear the costs and enjoy the benefits. If benefits of the elderly are cut now, then this group will be permanently worse off because they will not recoup this loss in a later period. If, instead, benefits for the elderly are preserved by raising taxes, then current workers will consume less. They may gain later on from the higher benefits once they themselves are old, but how much depends on economic growth. The same will be true of later generations. If we ask current workers to save more, this would help reduce taxes and benefits in the long run, but it will not help pay the rising costs of benefits for the current elderly.
The longer our nation delays making changes to the benefit and tax structures associated with entitlement programs for older individuals, the larger will be the “legacy liability” that will be passed to future generations. The larger this liability, the larger the increase in taxes on future generations of workers, or the reduction in benefits for future generations of retirees, that will be required to restore fiscal balance. The same is true for raising retirement ages. The remaining chapters in this report discuss these and related issues in more depth. The purpose of this chapter is to present a coherent view of the primary findings discussed in the following chapters, where supporting evidence and references may be found.
Of course, population aging has other consequences that are not explicitly addressed in this report. It has many social effects, from impacts on families and household structure through needs for the redesign of transportation networks and the built environment. But here our focus is more narrowly on the long-term consequences for the macroeconomy.
Population Aging Is Global
Populations are aging all over the world, but the pace and timing vary by country and region. In the wealthier nations, the aggregate median age has risen from 29 years in 1950 to 39 in 2010 and is projected to reach 48 by 2050. In the developing countries, the median age now is 27, but they too are aging, and their median age is projected to reach 37 by 2050. This aging of populations will influence the U.S. economy through globalized markets for goods, capital, and labor. U.S. wages, profit rates, and capital per worker depend in part on national factors, but also on these international ones. The old age dependency ratio is the ratio of the population
aged 65 and above to the population aged 20 to 64.1 The global old age dependency ratio is currently much lower than in the United States because of the preponderance of less-developed nations in the world population. For international markets, however, levels of income also matter. And when we take gross domestic product (GDP) into account (Chapter 8), the global old age dependency ratio is very similar to that of the United States. Projections suggest that this will remain true through 2050.
On average, people tend to accumulate assets over their lifetimes, and the elderly hold more assets than the young, particularly if we count pension plan assets. For this reason, population aging tends to boost wealth per capita and per worker, perhaps reducing its rate of return. To capture higher returns, investment may flow from older countries to younger ones. Population aging can be expected to alter international capital flows, and the patterns may be complex.
Birth, Mortality, and Population Growth
As mortality rates have fallen in the United States, the average length of life has risen from 47 years in 1900 to 78 today, and it is expected to continue to rise in coming decades. The committee projects U.S. life expectancy to reach 84.5 years by 2050 (Chapter 3). The average person living now is much more likely to survive until age 65 or 70 and to live more years thereafter. This is aging at the level of the individual.
A second and less obvious cause of population aging is a decline in the birth rate. With lower birth rates, younger generations are smaller relative to older generations, and this raises the average age of the population. There is a bulge in the U.S. population resulting from the baby boom generation, people born after the Second World War between 1945 and 1965, when there were three or four births per family on average. The baby boom generation is now beginning to reach age 65; members of this cohort are leaving the labor force and being replaced by smaller cohorts. This shift will mean that the number of retirements will be increasing sharply in the next two decades, just as the number of workers is leveling off. The number of retired persons per working person is expected to rise continuously in the years ahead. In this context, it is important to stress that while the baby boom generation has made the phenomenon of population aging more noticeable, the coming demographic transition is not just about the baby boom cohort—it is fundamentally about longer-run factors. Population aging is a broad, more pervasive trend that is here to stay.
1This will be the definition throughout the report, unless otherwise noted. The concept is based on chronological age and is subject to the limitations noted earlier in this chapter and elsewhere.
The United States is not alone in experiencing population aging. The entire world has embarked on this journey. The high-income countries are farthest along in the process, but many middle-income and lower-income countries will have slowing growth and aging populations in the coming decades because of their rapid and deep fertility declines. The United States is among the youngest of the rich nations because others generally have lower birth rates and lower immigration rates (immigrants are younger on average than the native-born population). Population aging through 2050 will be relatively modest in the United States compared to that in Japan and European nations.
While the future is uncertain, everyone who will reach age 65 by 2050 has already been born, as have many of the younger people who will be in the workforce then. As discussed in Chapter 3, the ratio of people aged 65 and over to people aged 20–64 will likely rise by 80 percent between now and 2050, and there is only a 1 in 40 chance that the increase will be less than 60 percent. Accordingly, we can be virtually certain that very substantial population aging is going to occur. Given the current age patterns of work and consumption, the biggest growth will be in age groups in which people currently consume a great deal and work very little. A growing fraction of the population will need to live either on its own savings or on public transfers. This is what makes population aging challenging for individuals and for public policy.
An important feature of rising life expectancy concerns the quality of life of the older years. As more people enter their seventh, eighth, and ninth decades, are these years likely to be healthy or unhealthy? Active or disabled? Engaged with family and society, or isolated and depressed?
If rising life expectancy merely added years of unhealthy, disabled, and isolated lives, the aging trend could be viewed with mixed feelings by older individuals and would probably become costly to society at large. Fortunately, the evidence shows that for most people, added years of life are largely active ones. Until very recently, rates of disability at older ages declined along with mortality rates. However, a number of recent studies have found that these declines ceased during the past decade, clouding the prognosis for future improvements in functional impairments.
Regardless of the time trends in disability rates by age, as the population ages, more people will be disabled and in ill health. The characteristics of the adult population will be changing in other ways as well, with a rising proportion of Hispanics and Asians, an increase in the level of education,
and growing obesity. Some of these changes in demographic composition will tend to raise disabilities on average, and others to lower them.
How will all these changes affect the ability of the adult population to work in the coming decades? An analysis commissioned by the committee indicates that, taking into account all these changes in the population, the proportion aged 20 to 74 who will be able to work will decline very slightly between now and 2050, from 91 percent to 89 percent (Chapter 5). Most people will have plenty of healthy years still available at the time they retire. Later retirement is both a realistic policy option and an available individual choice.
An important question for individuals and policy makers is whether those who do not work are out of the labor force owing to health impairments or because of a combination of economic forces and personal choices about work and leisure. At present, 83 percent of the population aged 65 and over is not working. Studies reviewed in Chapters 4 and 5 indicate that, although many older people are out of the labor force, this mainly reflects their work-leisure choice and their economic situation rather than a deterioration in their health status. Of those who are out of the labor force in their 60s, the majority do not report having even a minor impairment, let alone a major disability. Even in their early 70s, a substantial fraction of people not in the labor force have no functional impairment.
One of the important implications of these results is that the nation needs to rethink its outlook and policies about working and retirement. In many countries, including the United States, the age of 65 has conventionally been considered the “normal” retirement age, and this chronological age has been incorporated into many public policies and private attitudes. The committee believes that age 65 is an increasingly obsolete threshold for defining old age and for conditioning benefits for the elderly. Trends in education, health, mobility, and physical and mental disability suggest a need to reconceptualize stages of the life span and reconsider notions about work, saving, leisure, and retirement.
The bottom line is that the nation has many good options for responding to population aging. On the whole, America is strong and healthy enough to pay for increased years of consumption through increased years of work, if we so choose. Alternatively, we will be healthy enough to enjoy additional active years of retirement leisure if that is our decision, individually or collectively.
THE CHANGING ECONOMIC LANDSCAPE
As fertility, mortality, and health status have changed over recent decades, so have patterns of economic activity over the life cycle. At the be-
ginning of the twentieth century, life expectancy among men was 47 years, and two-thirds of those who survived to 65 were still working.2 The average retirement age of men declined steadily during most of the twentieth century, by 10 years or more, while life expectancy rose dramatically and the health of older people improved. In the early 1990s, the average male retirement age leveled off. More recently, the average retirement age has risen for both men and women, although it remains far lower than a century ago. The pattern for women was dramatically different. Starting after the Second World War, women’s labor force participation rates rose, from 33 percent to 60 percent in 2000. The net effect of these changes in mortality and labor supply is that in 1962 the ratio of retired years to working years in the individual life cycle was 0.35; by 2010 it was 0.41, and given the committee’s projections, it will rise to 0.52 in 2050.
Age patterns of consumption have also changed strikingly since 1960, particularly if we include health care (Chapter 3). In 1960, consumption after age 60 was lower than at younger adult ages. In that era, although government provided some income support, primarily through Social Security, there was very little public provision of health care. Medicare and Medicaid were introduced in the 1960s, and by 1980, the government provided substantial direct support for public and private consumption of food, health care, and other items. By 2007 the age-consumption profile had been radically transformed: The elderly consumed substantially more than younger people. The combination of declining labor earnings at older ages with sharply increasing consumption means that population aging has become more costly, both for individuals and for society as a whole.
ECONOMIC PROSPECTS IN THE COMING YEARS
The central thread of the committee’s charge was to evaluate the long-run impacts of the changing demographic structure on America’s living standards, with particular attention to issues such as saving, economic growth and productivity, income security in the older ages, and the ability of government to provide current and prospective levels of support for the elderly.
Our work must acknowledge the substantial impact of the recent financial crisis and the subsequent economic downturn. This report was begun during the middle of the worst U.S. business downturn since the 1930s. While the committee was concerned mainly with the long-run impacts of
2Much of the increase in life expectancy at birth was due to declining infant and child mortality, the effect of which is much like that of higher fertility. Declining mortality at adult ages has a different effect, raising the ratio of years spent in old ages to years spent in working ages, as is discussed in detail in Chapter 3.
demographic changes, it is aware that the recent economic downturn will leave a lasting imprint on America’s economic structure. It is likely to raise the federal debt by at least 35 percent of GDP. It may lead to some deterioration in labor market skills and to an increase in the population reliant on public support programs. Accordingly, even if and when the nation has returned to full employment, the scars left by the recession will likely make the transition to a sustainable fiscal future more difficult.3
The Macroeconomics of an Aging Population
Disregarding for now business cycle movements and changes in health status, per capita income depends on two factors: average productivity per person employed and the fraction of the population employed. Productivity, measured as net national product per person employed, grew at 1.56 percent per year over the 1960–2010 period. During the same period, workers as a fraction of the population grew at 0.31 percent per year owing to demographic change, a sharp expansion in the supply of women’s labor, and changes in the supply of men’s labor. Taken together, these generated growth in output per capita of 1.88 percent per year (Table 2-1).
The committee’s projections indicate that the fraction of the population in the workforce will decline by about 0.24 percent per year over the period 2010–2030 as the baby boomers retire, and then decline much more slowly at 0.02 percent per year for the next two decades, 2030–2050. Thus, if productivity continues to grow at its recent rate of 1.56 percent per year, output per capita will grow much more slowly, at only 1.32 percent per year, from 2010 to 2030, followed by growth at 1.54 percent from 2030 to 2050. So population aging will tend to slow the growth in income per capita by about 0.55 percentage points per year over the next two decades relative to past growth since 1960. But this calculation reflects only the changes in the labor force as a share of population and ignores the subtler effects of the aging of the labor force itself and the effect of population aging on the consumption side as the proportion of higher-consuming older persons rises.
The weighted support ratio takes into account these additional effects of future population aging. It is the ratio of the hypothetical labor supplied by the population to consumption by the population, assuming that per capita labor income and consumption at each age remain the same as they
3It is also important to acknowledge the post-2008 decline in housing prices. While the decline has left many households with much lower housing equity than they expected to have, it is not clear whether this will translate into a major long-term macroeconomic effect. Also unclear is whether housing’s contribution to personal assets or consumption is changing because of the aging of the U.S. population. The committee does not believe there is evidence that the value of housing will decline because of demographic-related shifts in household composition.
TABLE 2-1 Average Annual Growth Rate in Output and Workers in Three Time Periods
|Period||Output per Worker (% per year)||Workers per Population (% per year)||Output per Population (% per year)|
SOURCES: Output is net national product from the Bureau of Economic Analysis (BEA). Workers are persons engaged from the BEA. Hours worked per worker are from the Congressional Budget Office (2012, Table 2.3). Labor force and population from projections made by the committee (see Appendix A).
The projected weighted support ratio falls by 12 percent from 2010 to 2050. Accordingly, if consumers relied solely on labor income, consumption would be 12 percent lower in 2050 (than if there were no population aging) and would therefore decline by 0.33 percent per year over that 40-year period. It is likely that the economy will experience rising productivity during this period. Suppose that productivity growth continues at its average annual rate over the past half century, 1.56 percent. Even with the declining weighted support ratio, consumption would still be expected to rise at around 1.2 percent per year (i.e., 1.56–0.33 percent per year). In other words, the committee expects that average living standards will continue to rise over the coming decades, but that population aging will make the rise somewhat slower than would otherwise be expected.
Viewed in this way, the impact of population aging on consumption looks modest. Employment has grown faster than population over recent decades, but this trend is likely to reverse, and there will be a slower growth in average incomes due to demographic trends. However, there might be other offsetting factors, either positive or negative, that could change the growth in living standards. This report examines these offsetting factors in detail.
It is important to note that while the macroeconomic effects of aging might be modest for the economy as a whole, particular subgroups of the
4The methods for constructing the age profiles of labor income and consumption are described in Lee and Mason (2011). Labor income is pretax salary and wages, including fringe benefits, plus two-thirds of self-employment income, averaged across all men and women at each age (including zeros). Consumption includes private household consumption expenditures allocated by age, including education, health care, and imputed services of owner-occupied housing. It also includes in-kind public transfers, most importantly public education, Medicare, and Medicaid.
population could be much more vulnerable. For instance if the burden of adjustment were to be targeted on public programs such as Social Security, Medicare, or Medicaid, or if the costs were born largely by workers or largely by retirees, the overall impact would be quite substantial for the subgroup most affected. Similarly, if the economic cost of population aging were to be expressed in terms of cuts in programs such as Social Security, Medicare, or Medicaid, the impact on the elderly would be substantial.
Potential Changes in Productivity as an Offset to Population Aging
Several factors may offset or amplify the decline in the number of workers per capita and the related changes in consumption that are noted above. Such factors could include changes in underlying productivity growth, in labor force behavior, and in government policies including those influencing the growth in public and private health care costs.
Analyses of productivity growth generally separate the determinants of labor productivity growth into (1) those generated by increases in the quantity and quality of inputs and (2) those generated by technological change and other improvements in efficiency.
Changes in Other Inputs to the Production Process
The first factor—increases in productivity due to higher inputs—involves a wide variety of different forces that could be affected by demographic aging. These might include increases in the quantity and quality of private and public capital, changes in net ownership of foreign assets, and improved education, training, and skill of the labor force. Net income will also be influenced by rates of return and levels of risk for both domestic and foreign assets.
One important topic is how population aging might affect total wealth. This includes most significantly the nation’s stock of capital: Its homes, factories, equipment, software, and inventories. Additionally, in an increasingly globalized economy, a substantial amount of the nation’s wealth and obligations are international. At the end of 2010, America’s total wealth (domestic and foreign) was $46.9 trillion, which was 3.6 times its net national product.5
Population aging might affect total holdings of wealth, the composition of assets (e.g., houses, consumer durables, and corporate capital), and society’s propensity to hold risky assets. Economic theory suggests that
5Wealth here refers to current-cost fixed assets from the Bureau of Economic Analysis (BEA) plus net investment position at book value from the BEA. This is a slightly different calculation than found in Chapter 7, which uses flow of funds data.
many households accumulate savings for precautionary reasons and for retirement. If the proportion of the population at older ages increases, then if other things are unchanged, an average household will have higher wealth relative to income. For example, given the actual distribution of wealth by age in 2007, the projected population age distribution in 2050 would suggest a substantial 25 percent increase in national net worth per person aged 20–64 through the mechanical effect of population aging alone.6 In addition, if a longer life span increases the years spent in retirement, then households will need to accumulate more assets to maintain a given standard of living after retirement (abstracting from Social Security), which could further raise national net worth. These two factors (the need to save more for a longer retirement and changes in the age distribution) tend to lead to higher desired wealth holdings relative to income. This could raise asset income and might also raise the productivity of U.S. workers.
One important additional factor in a nation’s total wealth holdings is government debt. Private households that buy government bonds view them as part of their wealth holdings, and these bonds may inhibit or crowd out holdings of capital and other assets in their portfolios. To the extent that households in aggregate do not raise their saving to prepare for the increased taxes that are required to pay the interest, government debt can crowd out national wealth. Over the period 2006–2011, the debt-output ratio in the United States rose by over 30 percentage points.7 If the rising debt reduced capital and other wealth, then national income would be lower by this amount times the rate of return on wealth. While this would be a worrisome development, most of the recent rise in federal debt has been due to the deep recession and steps to stimulate the economic recovery; to date it has not been caused primarily by population aging or by policies related to the elderly population. Going forward, however, population aging will raise substantially the costs of various public programs, including Social Security, Medicare, and Medicaid, under current benefit structures. Depending on the policy response, aging might increase debt, as discussed in a later section of this chapter.
Although analyses of demographic aging often focus on tangible and financial wealth, the impact on human capital should also be noted. Human capital refers to the useful skills of the population acquired through formal and informal education, training, and on-the-job experience. Economic
6This calculation is based on the distribution of net worth by age group of household head in 2007 and the committee’s population age distributions for 2010 through 2050, taking into account household headship rates by age. See Appendix A for sources and details.
7Data on debt refer to debt held by the public and are from the Congressional Budget Office (2012). Government debt would be much larger if it included the implicit net liabilities of Medicare and Social Security (see Chapter 9).
studies indicate that human capital is just as important as tangible capital as a driver of economic growth.8
Increased investment in the human capital of our children offers several advantages in the context of population aging. Higher educational attainment is associated with longer life, better health throughout life, and less disability in old age. It also raises labor productivity and helps to augment the relatively smaller number of workers with higher quality workers, which also helps to prevent declining returns to capital as populations age. Education also improves financial literacy.
To date there is little research on the impact of aging on investment in human capital. In the past, most human capital was accumulated in the early stages of the life cycle. Both formal education and on-the-job learning are greatest up to 30 years of age. The other important aspect of human capital is that it can serve as a brake on changes in returns on assets. As will be discussed below, slower population growth may lead to a lower rate of return on assets. To the extent that returns on human capital remain high, this will tend to slow the decline in the returns on tangible capital since increased quality of labor would partially offset decreased quantity.
Improved Efficiency and Technology
The second ingredient in productivity growth comes, over the long run, from the generation and diffusion of new scientific, technological, and engineering knowledge as well as other gains in efficiency. Most of the factors at work here operate independently from our nation’s demographic structure. However, one age-related driver for technological change and innovation is the tendency for scientific output of innovators in science, engineering, and many other fields to rise steeply in the 20s and 30s, peak in the late 30s or early 40s, and then trail off at older ages. Yet it is the global trends in basic science and invention that increasingly determine a nation’s long-run technological competitiveness. While having a young population can help drive invention and innovation, population aging has very little effect on technological change across societies. Other factors such as income levels, education, institutions, and economic incentives to innovate tend to dominate the actual distribution of scientific and technological output.
As the population ages, so does the labor force. Will the aging of the labor force make it less productive? The committee has investigated this possibility (Chapter 6) and concludes that the impact is likely to be very small. While there may be some adverse productivity impact from a rising
8For example, Barro, Mankiw, and Sala-i-Martin (1995) conclude that human capital accumulation predicted from their model is roughly comparable in size to physical capital accumulation in the United States.
number of older workers, this is likely to be offset by the favorable impact of the decline in inexperienced workers. Taking earnings as an indicator of productivity by age, population aging in the United States will have a negligible effect on average labor productivity in the coming decades. Nonetheless, there are many other potential impacts of a changing age distribution for which empirical evidence is lacking.
ADAPTING TO THE CHANGING DEMOGRAPHIC
AND ECONOMIC ENVIRONMENT
The Need for Action Despite the Uncertainty of Projections
This description of the consequences of population aging is based on forecasts of changes between now and 2050. Yet any forecast is subject to uncertainty, as described in detail in Chapter 3. The fact remains, however, that even after taking account of such uncertainty, the old age dependency ratio is virtually certain (97.5 percent probability) to rise by nearly 60 percent between now and 2050. Moreover, the committee stresses that the mere fact that projections are uncertain does not mean that the government should postpone responding to an anticipated fiscal imbalance.9 Individual taxpayers and beneficiaries are averse to uncertainty, in the sense that they want to avoid the possibility of a future loss even when balanced by the possibility of an equal future gain. In the same way, they would prefer to avoid the risk of larger future benefit cuts or tax increases even if there is an equal possibility that smaller benefit cuts and tax increases will be needed. For this reason, uncertainty means that action should be accelerated rather than delayed, to lessen the likelihood of very large benefit cuts or tax increases in the future. With uncertainty, it is desirable to have a lower debt-to-GDP ratio than we otherwise would, through additional public saving beyond what would be needed to respond to an expected fiscal imbalance. Of course, the desirability of doing this hinges critically on our ability to resist future political pressures to spend any budget surpluses that might accumulate.
At the beginning of the chapter, the committee discussed the four adaptive mechanisms that America might use to deal with an aging population: We can reduce the relative living standards of the elderly, work longer, increase saving while working, and/or transfer more from working ages to the elderly. Each of these mechanisms is discussed in more detail below.
9The committee refers here to a “fiscal imbalance,” though by law benefits cannot exceed accumulated reserves plus current revenues. When it refers to an “imbalance” it means that, according to current projections, revenues plus reserves will be insufficient to pay the currently scheduled level of benefits.
Reducing Consumption of the Elderly
A first policy option would be to reduce people’s consumption when old, relative to their earlier years. If household decisions were well informed and made independently, this might happen naturally as people weigh the relative priority of consumption when young versus old. A reweighting of priorities could lead to a reduction in consumption in older years, or to working longer, or to saving more in younger years (thereby reducing consumption in those years). So households themselves might change their life-cycle consumption patterns if they fully anticipate the need to build a nest egg to sustain them over a longer retirement period. A more complex problem has to do with government transfer programs, where adjustments in taxes and benefits paid must take place via the political process. To date it has proven quite difficult for policy makers to achieve political consensus around the question of how to restore the nation’s Social Security and Medicare systems to solvency. Yet doing so is essential if the nation is to provide an institutional context in which households can make retirement plans.
Age at retirement is central to population aging and its economic consequences. Raising retirement ages is one key alternative to reducing the consumption of leisure and enhancing people’s ability to stretch their assets over their lifetimes. The average retirement age for men declined substantially in the United States throughout most of the twentieth century. Although this trend stopped in the early 1990s and then reversed, men still retire at a much younger age than in the past, despite their better health and much longer life. Women’s average age at retirement has moved parallel to men’s over recent decades, but it stabilized and began to rise somewhat later. The committee foresees a continued rise in the labor force participation rate of older Americans. Based on evidence reviewed in Chapter 5, the committee concludes that the potential for work is much greater than is reflected by the proportions of elderly actually working.
Health Status and Retirement
On average, Americans today are retiring in much better health than was true three decades ago (Chapter 5). A substantial proportion of older persons who are not working have no major impairment to their health status. This suggests that, if people choose to work longer for either economic reasons or because of personal preferences, their health status will mostly not be an obstacle. Further, changes in job mix coupled with a general de-
crease in the physical demands of most jobs bode well for continued labor force participation at older ages.
Will There Be Enough Jobs for the Young If People Work Longer?
Some have expressed concern that if older members of the population work longer, they will take jobs away from the young. Yet this did not happen in the past nor has it occurred in other nations (Chapter 5). The committee thus concludes this is not a substantive concern for the United States in the future. In normal times, except for deep business cycle recessions, the overall number of jobs is determined primarily by the size of the labor force. If anything, an increase in older workers is predicted to slightly increase the wage rates of young workers.
Are Current Workers Saving Adequately for Retirement?
With a larger number of people in retirement and living longer, it would be useful to determine whether they will have adequate living standards during their retirement years. Over the last half-century, much of the worst poverty among the elderly has been reduced, in part owing to government transfer programs. Using updated measures, the poverty rate of the elderly is similar to that of the rest of the population;10 it is much below that of the nonelderly if in-kind government transfers (such as Medicare) are taken into account.
Looking prospectively, the committee has considered whether saving during the worklife will be sufficient to support an adequate living standard in retirement, taking into account expected future government benefits. This question raises several issues concerning the structure of saving (particularly the changing nature of employer-sponsored pension plans), the rate of return on pension savings, and the prospective impact of demographic aging on rates of return.
Changing Nature of Pension Plans
About half of the U.S. workforce is covered by an employer-sponsored retirement plan, and this has been true for the last half-century. But the structure of pension plans has changed dramatically over this time. In the 1970s, most employer-sponsored pension plans were of the defined benefit (DB) variety, where payouts were based on an employee’s earnings history,
10The standard poverty measure shows lower poverty rates among the elderly, but the new Census Bureau Supplemental measure makes several adjustments that change the picture completely (see Short, 2011).
length of service, and retirement age. Initially, defined benefits were mainly paid in the form of lifetime annuities (though more recently lump sums have in some circumstances been permitted). Capital market risk in DB plans is primarily borne by the plan sponsors, though the participants bear some of that risk if an employer files for bankruptcy and the plan is so underfunded that the reinsurer cannot guarantee full benefits. Today, employer plans in the corporate sector have mostly converted to defined contribution (DC) pensions (e.g., 401(k) or 403(b) plans). Participants must generally decide how much to contribute (sometimes with an employer match) and where to invest the funds, thus bearing capital market risk more directly. At retirement, the benefits are usually paid out as a lump sum rather than as a life-long benefit stream.
The changing nature of pensions has several implications. One is that, in DC plans, individuals may have difficulty determining whether their saving is adequate for their retirement needs. Another is that investment decisions have been shifted to participants who may be unable or unwilling to make informed investment choices. And in DC plans, most individuals are faced with the difficult decision of whether and how to annuitize their assets upon retirement, which traditionally was not the case in DB plans.
In the context of the macroeconomic impacts of aging, a key feature of the movement to DC plans has to do with different incentives to retire. While DB plans traditionally embedded strong incentives for people to retire early, there is no such encouragement in most DC plans. The increased prevalence of DC plans is perhaps the main reason for the increase in the labor force participation rate of older workers since the mid-1990s.
Changing Rates of Return
Some analysts have raised concerns that an aging population will reduce future rates of return or reduce the price of assets, a topic investigated in Chapter 8. The committee concludes that rates of return on investment are likely to have only modest effects on most retirees’ financial status because their asset holdings are small, though rates of return will influence savings adequacy for those with greater assets. There are a number of ways that population aging might affect asset returns and prices. Population aging will likely lead to higher U.S. wealth holdings per capita and per worker, which could drive down average returns on capital. Projections of global population aging suggest that these trends will be similar around the world. But population aging could also lead to increased government debt, which would tend to raise returns to capital. Globalization of financial markets implies that broader forces—particularly the overall macroeconomic environment, the business cycle, shifts in global savings and investment patterns, and the rise of high-savings countries such as China—are likely
to dominate the pattern of capital market returns in the coming years. Compared to these broader forces, the effects of U.S. population aging on rates of return are likely to be very small. Though major changes in asset prices have occurred and are likely to occur in the future, demographic forces are generally too predictable and move too slowly to cause major financial market dislocations such as those of the 2007–2009 period.
Overall Adequacy of Consumption in Retirement Years
Studies of U.S. retirement saving adequacy produce different answers depending on the methods used, with research suggesting that between one-fifth and two-thirds of the older population have undersaved for retirement (see Chapter 7). Some common themes emerge. First, there is good evidence indicating that low- and lower-middle-income households accumulate few financial and pension assets for retirement. For those households, Social Security, Medicare, and Medicaid are a central part of maintaining retirement living standards. To the extent that benefits paid by these government programs might be reduced in the future, the living standards of the affected retiree households will fall.
Second, the quality of financial decisions, and therefore financial literacy, will play an increasingly important role in how well households fare in their retirement years, particularly in light of the continued trend to DC pensions. Households will need to decide how much more to save and how to structure their portfolios during their working years. They will need to decide when it is economically prudent to retire, taking into account personal, macroeconomic, and political uncertainties. When they do retire, they will need to decide whether to annuitize their accumulations and, if so, how much and with what annuity options. The many households whose wealth rests mainly in their home ownership will need to decide whether and how to use those assets to finance retirement consumption. These are very complicated questions, and financial professionals give varying advice. The committee is concerned that our nation is poorly prepared to move into this changed financial landscape, and it finds substantial value in boosting financial literacy.
The committee concludes that Americans are likely to face greater economic difficulties in retirement than they did in the recent past. Retirement insecurity and saving inadequacy are likely to increase rather than recede, and these will be exacerbated by the need to navigate a more complex financial and macroeconomic structure.11
11It is useful to note that the meaning of retirement has changed over time, and that what might be regarded today as a problem or potential problem (e.g., too extended a retirement, too much retirement consumption, and difficult financial choices in old age) evolved from
While future uncertainties are large, there are many ways to improve Americans’ retirement security. These include encouraging people to work longer, raising retirement ages, improving insurance protection and long-term care, and putting Social Security and Medicare on a sound long-term footing. Additionally, while financial innovation has been rapid in some areas, it has been relatively slow to develop products that will help most households manage their savings and transition to retirement. The nation needs to improve private-market solutions such as more saving, better financial literacy, and enhanced financial, long-term care, and annuity products.
Changing Public Policies and Transfer Programs
As noted earlier, the committee concludes that the overall macroeconomic consequences of population aging in the United States are likely to be modest. However, because the government plays a particularly important role in financing consumption and health care for the elderly, many of the consequences of population aging will be focused on specific government programs rather than spread across the economy. For these programs, population aging will have a major effect on costs. Population aging already has led to projected shortfalls in the finances of Social Security, Medicare, and Medicaid and is likely to lead to increasing government budget deficits in the future (Chapter 9). The consequences for Social Security are predictable, and they can be relatively easily addressed by benefit formula changes and increases in contributions. Programs providing health care and long-term care, notably Medicare and Medicaid, are a different matter. Health care costs per eligible person have been growing substantially faster than per capita income for decades, and if this pattern continues, it will interact with population aging to drive up public health care expenditures substantially. Recent reforms have attempted to address this problem and could lead to fundamental change in delivery, quality, and cost of care, but their impacts are as yet unclear.
Government programs are particularly important because it is largely through them that policies could influence retirement ages and perhaps also employers’ demand for older workers. Moreover, they could alter changes in consumption for both workers and the elderly. Changes to these programs will also go far in determining how the costs of population aging will be shared across generations and age groups in the future. A critical need in the near future is to put Social Security and government health programs on a secure footing and to reduce uncertainty and mistrust by announcing
something radically different. For insightful studies of the history of retirement, see Costa (1998) and Haber and Gratton (1994).
future changes well in advance so that people can structure their plans accordingly. Additionally, policies need to be developed that encourage better adaptive behavior on the part of workers and companies.
Aside from transfer programs, many other public programs could help smooth the transition to an aging society. For instance, programs that make the workplace friendlier for older workers might encourage them to stay at work or to take part-time employment (Chapter 5). Much analysis remains to be done to develop financial instruments such as improved reverse mortgages as well as better long-term care and annuity products. And not least, the nation will need to find ways to improve financial literacy for people who increasingly will manage their own retirement finances from the first days of joining the labor force to the later days of deciding how and when to spend money in retirement.
The main message of this report is that the committee sees population aging as posing serious but not insurmountable challenges for the United States. From an economic point of view, a larger fraction of national output will need to be spent on consumption by the elderly. Yet an aging population also offers many opportunities, as long as sensible policies can be implemented with some lead time to allow companies and households to respond.
The committee has outlined several ways that the nation could adapt and adjust to the changing demographic structure. Encouraging later retirement is one useful route. It is unlikely to be hindered by poor health or disability. Such a trend could be helped through changes in the structure of pension and health programs and through a continued shift toward DC plans.
Additionally, public policies could facilitate greater saving during the working years as well as provide better financial instruments to turn financial and housing assets into income streams that protect people against the vagaries of longevity. Improving the nation’s financial literacy will also be critical in a world where people face a larger number and more complex array of choices about working, saving, and retirement.
Moreover, because of the prospect for a sharply rising federal debt, the nation will need to balance the priorities of different programs as well as provide additional revenue streams devoted to programs for the elderly. There is little doubt that there will need to be major changes in the structure of federal programs, particularly for health.
The committee concludes that our nation must take action soon, rather than simply continuing with the same policies and practices as in the past. Population aging, unlike other aspects of the future, is a certainty, and many
adjustments will have to be made. No single feasible policy adjustment is likely to be an acceptable or a sufficient response to the challenge and opportunity of population aging. An aging society need not have lower living standards, slower growth in innovation and productivity, or inefficiently high tax rates. But delaying decisions on how to adapt to our aging demographic structure will make the transition more difficult and costly.