Income Security and Health Care Financing Programs
The financing of retirement and of health care for an aging population presents immense challenges. These are being met with different strategies around the world.
INCOME SECURITY FOR AN AGING POPULATION
Carroll School of Management
The U.S. population is aging, longevity is increasing, and the need for retirement income is growing. Yet at the same time that the need for retirement income is increasing, the retirement income system is contracting. Retirement income may be thought of as a three-legged stool consisting of public government funds (Social Security), employer-sponsored retirement plans (defined benefit or defined contribution plans), and personal savings. All three legs of this stool are wobbly.
The cost of Social Security is increasing. With a pay-as-you-go system, more retirees, and fewer workers, the cost rate goes up—this is well known. Social Security will replace less of an individual’s preretirement income in the future. Under current law, a decline in the replacement rate from 39 percent in 2002 to 28 percent in 2030 is anticipated. Thus, the first leg of the retirement income stool is not so secure.
The second leg, employer-sponsored retirement, is also not solid. In
the private-sector workforce, fewer than half the people working have any type of employer-sponsored retirement plan. This figure has changed little over the past 30 years. The plans themselves, however, have changed substantially from defined benefit plans to defined contribution plans. Defined benefit plans promise a specific monthly benefit upon retirement; defined contribution plans—such as 401(k) plans—are characterized by contributions from the employer or employee (or both), and the balance of the account is available to the employee upon retirement. In the early 1980s, 62 percent of those with pension coverage had a defined-benefit-only plan. By 2007, 63 percent of those with pension coverage had a defined-contribution-only plan. Such plans shift all responsibilities to the individual. With a 401(k) plan, the individual must decide whether to join the plan, how much to contribute, how to invest those contributions, and how to change the portfolio over time. And as Munnell observed, “people make mistakes every step of the way.” About 20 percent of people who are eligible fail to participate in 401(k) plans. Of those who do, only 8 percent contribute the maximum; 92 percent do not. Decisions regarding investment are also faulty, with 42 percent of people failing to diversify and putting all their funds into stocks or all into bonds. Furthermore, people do not change the investments as they age. When the stock market fell, people age 55 to 65 held two-thirds of their assets in equities and did not roll over their investments. In sum, people do not have much money in their 401(k) plans. In 2007, the average 401(k) plan had $60,000 in the account, considerably below the potential.
The final support of the three-legged stool is personal savings, which is currently completely inadequate for retirement needs. And with the booming housing market of the 1990s, as many people borrowed on their homes and spent beyond their incomes, the savings rate went negative.
The solution to the retirement income challenge is straightforward and similarly threefold: people should remain in the workforce longer, make better use of retirement assets, and save more.
Working longer, Munnell argued, “is really the most powerful thing people can do to have a secure old age.” An individual retiring at age 62 receives 75 percent of the Social Security benefits that would come at the full retirement age which is moving toward 67, while an individual who continues working until age 70 gets 132 percent. Social Security benefits, in Munnell’s view, are “backbone income”: they are lifelong, adjusted for inflation, and unavailable in the private sector. Working longer also gives the assets in a 401(k) plan more time to grow. Finally, working longer shifts the ratio between working years and years in which support will be required.
There are several obstacles to working longer. Employers’ perceptions of older workers are a problem. In survey responses, employers say they
value experience but are less positive about older workers’ ability to learn new tasks and less sure of their physical health, stamina, and continued presence in the job. Such perceptions make employers disinclined to train or promote older workers.
Furthermore, clearly not everyone can work at advanced ages. Estimates of disability-free life expectancies indicate that people in the bottom income quartile at age 50 can expect 14 years of disability-free life. It would hardly be possible to recommend to such individuals that they continue working for another 17 years. In response to later questions on this topic, Munnell proposed that one approach might be setting different ages for full retirement based on life experience, perhaps using a person’s average indexed monthly wage. She acknowledged that the disability issue is very hard and requires very careful management. Nonetheless, she remarked, “We are taking over General Motors. We are taking over banks. We could figure this out. This is not impossible to solve and to try to make changes that are helpful for the bulk of the population and also protect the vulnerable population.”
Turning to the use of retirement assets, people need to use the assets they have more effectively. When people get to retirement, Munnell noted, “it is hard to tell them they should have eaten better and exercised more and saved more. They are what they are, so we need to make sure they use what they have effectively.” One important challenge is annuitization: people need an orderly way to decumulate retirement balances. Annuitization yields the highest income for life, yet people dislike it. Munnell suggested that ways be devised to “coax, cajole, default, or mandate” that some portion of people’s financial assets be annuitized. Tapping housing equity efficiently is another challenge. For many people, their house is their major asset; leaving it to heirs is a luxury that very few can afford. There needs to be some financial innovation so that people can access their housing equity.
The third component of improving retirement income is to increase savings, and Munnell has a simple imperative—“We need more.” Without increased personal savings, standards of living in retirement will deteriorate. Munnell’s research suggests that 43 percent of current households will not be able to maintain their preretirement living standards when they reach age 65.
In terms of the three legs of the existing retirement income stool, the situation is particularly acute for those who depend on Social Security. Can benefits be cut? If the population is divided into income terciles, Social Security payments are crucial at the bottom, really important in the middle, and significant even for the top. Thus to suggest that shortfalls in the Social Security system be fixed partly by tax increases and partly by benefit cuts is not realistic. “When I look at the retirement income
landscape,” Munnell observed, “I think that we cannot afford to cut back on benefits at all.”
Even if people work longer and Social Security benefits are not diminished, private retirement plans will be inadequate. When 401(k) plans were first devised, they were intended to be supplementary to Social Security and defined benefit plans. “They were never intended for this job,” Munnell said, yet they have become many people’s sole form of private retirement income.
A whole new tier of savings for retirement is necessary. For people with or without a 401(k) plan, it would provide 15 or 20 percent of replacement income. It would be contributory by either employee or employer, and it should be mandatory, with no access until retirement. This new tier of savings would be invested in something very low cost, and the portfolio would change as people age. It would be paid out as an annuity. Munnell acknowledged that although she has many ideas, she does not have a plan. Nevertheless, she asserted that the imperative is to increase savings by whatever means.
The solution to income insecurity for an aging population, then, is threefold: work longer, use assets more effectively, and save more. With that agenda, Munnell proposed the following key questions:
Are people healthy enough to work longer? Are there large components of the population for whom working longer is not the right answer? What can be done about them? How can a system be devised that gets most people who can work to work and yet protects those who cannot? Is there adequate employer demand for older workers? What are the actual levels of productivity of older workers?
What new financial instruments would help people make more effective use of their retirement assets?
How can savings be increased?
HEALTH CARE FINANCING
Kenneth M. Langa
Department of Internal Medicine
University of Michigan
There are three basic questions regarding health care financing for an aging population: (1) What is happening to the prevalence and onset of disability?, (2) How should older adults be cared for generally and at the end of life?, and (3) Who will care for the growing number of older
adults? Langa reviewed the areas of consensus, controversy, and ignorance regarding these questions.
Disability is measured in terms of the capacity to perform activities of daily living (ADLs) and instrumental activities of daily living (IADLs) independently. The risk of disability increases with age, and the United States has an aging population. What are the trends in disability prevalence? There is a general consensus that the prevalence in disability among the aged has declined over the past 20 years. There is also some evidence of significant declines in cognitive impairment, although longitudinal data are more limited. These declines are probably the result of increases in education, through multiple direct and indirect causal pathways. Decreases in cardiovascular disease have also contributed to decreases in disability prevalence, as have improvements in health care (such as cataract surgery, joint replacement, and treatment for hypertension and cholesterol). Healthy behaviors are also helping to compress disability, so that onset of disability is later when it does occur. Disability rates will be higher for those in the bottom income quartile. This may be due, in part, to differences in early education, which has long-term consequences including influence on the choice of jobs, ways of using the brain, environments, and social interactions. Better education earlier in life for those in the bottom quartile could be an important component of addressing disparities in health and disability later in life. A final factor regarding trends in disability is the impact of rising obesity and diabetes, including among children. The large increase in the prevalence of obesity and diabetes may reverse some of the disability declines that have occurred over the past 20 years.
Trends in disability are important to study, because they have large implications for both costs of long-term care and for the ability of older workers to remain in the labor force. Tracking trends in disability also helps to identify healthy behaviors and other risks and protective factors in the environment.
Regarding the right amount of care for older people and those at the end of life, there is particular evidence of regional variations in Medicare expenditures. Per capita Medicare expenditures vary widely across regions. Growth in Medicare expenditures has also exhibited striking differences. In Miami, for example, Medicare expenditures are far higher and growing faster than in Oregon. The data are properly adjusted for characteristics of the population, and thus the results reflect actual differences in utilization rates, not differences in the people or the burden of disease in these regions. The patients in the high-use regions are not sicker, nor do they prefer more intensive care in general or at the end of life. Furthermore, they do not have better health outcomes in many ways that can be measured. Rather, physicians’ behavior in high-use areas is
key. Physicians in high Medicare expenditure regions make more referrals to subspecialists, more frequently hospitalize older patients at the end of life, more frequently use intensive care units for older patients, and less frequently discuss palliative care with patients and families.
Why is determining the right amount of care for older adults important? First, given the size of the Medicare budget, even small changes in per capita growth in Medicare expenditures will have major implications for the program’s future solvency. Second, the so-called gray areas of decision making—determining whether and how much to intervene—is where health care money goes. As Langa observed, “I would argue that geriatrics is almost entirely defined by gray areas of decision making.” Good geriatricians need to handle ambiguity and complexity. In treating older patients, he suggested, there is a “huge potential, especially with full insurance coverage, for things to expand if we don’t figure out what should be done rather than just being able to do everything we can.”
The tremendous regional variation in Medicare expenditures suggests that the decisions made in these discretionary gray areas vary widely and that further refinement is needed, for example, in the guidelines for appropriate care for frail adults with multiple coexisting chronic diseases. The guidelines for treating a 50-year-old in this situation might be very different from those for an 80-year-old. Further attention should also be given to determining the appropriate goals of chronic disease care in terms of independence or length of life, for example. More work is needed to define and measure outcomes. Another essential challenge is to figure out how to organize and pay for services in a way that will reward more conservative practice. The current system of compensation for procedures performed rather than for actual health outcomes generates perverse incentives. Finally, better tools for prognosis and for encouraging communication between doctors and patients and their families are essential. This is true not just at end of life. End-of-life care and costs should be thought about from an earlier age, prior to crisis or any kind of impairment.
The next question is who will provide the care for an aging population. The pool of informal nonprofessional caregivers in the United States is shrinking. The population over age 65 is increasing, and the population of 44- to 55-year-olds—the potential caregivers—remains essentially flat. Looking at informal caregiving costs for various chronic diseases, the largest costs were for dementia, amounting to 29 percent of costs in 2000. When other aspects of brain ill health, such as depression and strokes, are added, this accounts for 54 percent of informal caregiving costs. In informal settings, caring for someone with dementia can be literally a full-time job. Thus, brain health will be key in terms of the costs and hours involved in informal caregiving.
In the formal sector, there are coming shortages in geriatricians and
direct care workers. The United States currently has 7,000 geriatricians, and trends suggest that number will decline by 2030, when projected need will rise to 30,000 geriatricians. Current shortages in direct care workers for home care services and nursing home care are also significant and expected to grow.
Considering the future supply of informal and formal care is important because it focuses attention on potential interventions to increase supply. It also highlights the importance of making the most of what is available, particularly improving the performance of geriatricians and other care providers in the gray areas of decision making regarding care for older patients. Improved training, clearer guidelines, and different pay incentives could both improve outcomes and lower costs. There is also a great potential to coordinate the formal and informal care systems, especially through such technology as telemedicine and e-consultations. This would extend the caregiving network and could improve outcomes while decreasing costs. Research is also now being conducted to understand the determinants of the well-being of informal caregivers and ways to intervene to maintain it. Supporting informal caregivers with new technologies could make them more effective and efficient. The training of geriatricians and other caregivers also needs further study to increase recruitment and retention and improve decision making.
Langa grouped the research agenda into three questions:
What are the trends in the prevalence of disability, and will disability be more compressed toward the end of life as the society ages?
What is the right amount of care for older patients in general, and older patients near the end of life in particular?
What caregiving resources, both paid and unpaid, will be available for older adults with disability and chronic disease?
AN INTERNATIONAL PERSPECTIVE
Department of Economics
University of Hawaii at Manoa
Different societies provide for the economic needs of the elderly via a mix of public resources, private assets, and family transfers. Relying on data from the National Transfer Accounts Project, Mason traced patterns in health consumption and its finance, the labor income of older individuals, and the funding of the life-cycle deficit—that is, the gap between consumption and labor income.
Japan, which has the oldest population in the world and is among the most rapidly aging, may be used to illustrate trends that are also occurring elsewhere. Consumption, which rises at the end of life, is a trend that is accelerating. The increase comes from spending on health care. These trends raise two questions: What steps can be taken to provide for and control health care spending at the end of life? And can policies raise the labor income of the elderly by increasing the age of retirement and increasing the productivity of older workers?
Turning to the composition of health care spending, data on publicly and privately funded health consumption from age 55 to over age 90 in several countries can be compared. Public spending on health in the industrialized countries is high and steeply increasing with age. In this, the United States is similar to Germany and Japan, but the United States stands out from other industrialized nations in its large private spending on health. In this, it more closely parallels some developing countries, such as Uruguay, Chile, and Brazil. Other industrial nations, such as Austria or Finland, have less private spending on health than the United States.
The same data sets yield information on labor income as a percentage of consumption for elderly people in several countries. On average, for those age 65 and older, 17 percent of their consumption is funded by their labor income. Countries with a higher average are the Philippines (38.9 percent) China (35.5 percent), and Kenya (34.2 percent). However, given the multiple differences between these societies and the United States, these patterns do not provide much guidance for the nation.
By contrast, among high-income industrial economies, a far smaller portion of consumption by older people is funded by their labor income. In Austria, the figure is 1.8 percent of consumption, in Germany it is 3.2 percent, and in the United States it is 15 percent. While this is far better than Austria or Germany, Mason observed, the developing countries are “starting from a very bad position, and it is going to take a revolution in employment amongst older workers and what they are paid in order to make much headway about the gap between what they consume and what they produce.”
Thus, for the foreseeable future, closing the life-cycle deficit with labor income will not be feasible. It will be covered instead by three types of reallocations. One is asset-based reallocations, in the form of either asset income or dissaving. In actuality, Mason said, this is almost entirely asset income; dissaving is rarely observed. Another form of reallocation is net public transfers, in the form of goods and services provided through the public sector, cash transfers such as pensions, and lower taxes for the elderly. Finally, there are private transfers, either net interhousehold or net intrahousehold transfers.
Mason then discussed the share of the life-cycle deficit over age 65 that is funded by private assets, public transfers, and private transfers in a number of countries. Taiwan is the sole economy that comes close to funding the life-cycle deficit equally from the three sources. No country relies solely on family transfers, although they are a relevant component of funding the life-cycle deficit in Asian countries. Most countries depend primarily on a mix of personal assets and public transfers. Those that rely most on public transfers include Austria, Finland, and Germany, as well as Japan and Costa Rica. The United States is unique among highly industrialized countries in relying more on private assets and less on public transfers. This is a result of its employment-based pension system. Developing countries rely far more on private assets to fund the life-cycle deficit. The Philippines is particularly notable for its reliance on private assets, but this may not be supporting a very high level of consumption.
This support system also changes as people age further. In general, asset-based reallocations decline and transfers increase with age. Public transfers tend to increase in Austria, Germany, and the United States, whereas family transfers increase in Costa Rica, Japan, Mexico, South Korea, and Thailand.
Mason proposed three questions regarding the funding of the lifecycle deficit via asset-based reallocations, public transfers, and private transfers:
Which of these reallocation systems work? Which are sustainable? What incentives and disincentives does each system create for the timing of retirement, the quantity of saving, and investment in human capital?
Do the reallocation systems vary in their risk-sharing features?
Which reallocation systems are least able to adapt to population aging?