Skip Navigation

Reform of the Social Security program is slated to be a major item on the agenda of the 106th Congress. One aspect of the program likely to receive attention is the age at which workers first become eligible for retirement benefits. Under current law, workers may retire and begin collecting permanently reduced Social Security benefits as early as age 62. (For full retirement benefits, they must now wait until 65; that age will gradually increase to 67 in the next quarter century.) A number of proposals call for raising the earliest eligibility age to at least 65. Opponents of such proposals worry that raising the eligibility age could cause severe hardship for people who would have trouble delaying retirement and who lack pensions and other means of support.

This Congressional Budget Office (CBO) paper, prepared at the request of the Committee on Ways and Means minority, provides material relevant to that debate. In particular, it contains new information about the characteristics, circumstances, and financial resources of men and women who began receiving Social Security benefits at age 62 or 63 in recent years. In accordance with CBO's mandate to provide objective and impartial analysis, this paper contains no recommendations.

Ralph Smith of CBO's Health and Human Resources Division prepared the paper under the direction of Joseph Antos. Susan Labovich did the computer programming for the analysis. Amy Weber provided research assistance. A number of people at CBO provided helpful comments and suggestions, including Paul Cullinan, Robert Dennis, Theresa Devine, Arlene Holen, Noah Meyerson, Benjamin Page, Kathy Ruffing, Paul Van de Water, and Bruce Vavrichek.

Christian Spoor edited the manuscript, and Liz Williams proofread it. Ronald Moore prepared the paper for publication. Laurie Brown prepared the electronic versions for CBO's World Wide Web site.
 

June E. O'Neill
Director
January 1999
 
 

SUMMARY AND INTRODUCTION

CURRENT LAW

ISSUES IN RAISING THE EARLIEST ELIGIBILITY AGE

WHEN DO WORKERS USUALLY RETIRE?

WHO TAKES EARLY BENEFITS?

• Characteristics of Early Retirees
• Who Might Be Dependent on Early Benefits?

CONCLUSION

APPENDIX: THE SURVEY OF INCOME AND PROGRAM PARTICIPATION
 

TABLES
1.	Social Security Status of People Age 62 or 63 in the Survey of Income and Program Participation in the Early 1990s
2.	Selected Characteristics of Social Security Beneficiaries and Nonbeneficiaries Age 62 or 63 in the Early 1990s
3.	Income and Net Worth of Social Security Beneficiaries and Nonbeneficiaries Age 62 or 63 in the Early 1990s
4.	Major Sources of Income for Men Receiving Early Social Security Benefits
5.	Major Sources of Income for Women Receiving Early Social Security Benefits
6.	Selected Characteristics of Men Receiving Early Social Security Benefits
7.	Selected Characteristics of Women Receiving Early Social Security Benefits
8.	Alternative Estimates of Dependency
A-1.	Comparison of SIPP and SSA Estimates of Social Beneficiaries Age 62 or 63 in the Early 1990s
FIGURES
1.	Labor Force Participation Rates of Men and Women, by Single Year of Age, 1997
2.	Labor Force Participation Rates of Men and Women 55 or Older, 1948-1997

 

SUMMARY AND INTRODUCTION

In just 10 years, the first wave of baby boomers will reach age 62 and become eligible for early Social Security retirement benefits. If their behavior follows the pattern of workers who turned 62 during the past quarter century, many of them will claim permanently reduced benefits at that age, or shortly thereafter, rather than waiting until they are eligible for full benefits a few years later.

Because of the aging of the baby-boom generation and projected gains in life expectancy, the number of Social Security beneficiaries is expected to increase much faster than the number of workers paying Social Security taxes. The program's trustees project that total spending (in 1998 dollars) will more than double by 2032 under the current benefit structure. By then, however, the revenues earmarked for Social Security spending will be sufficient to pay only three-quarters of the program's expected costs. That projected shortfall--together with pressure on the federal budget from increased spending on health programs for the elderly--could ultimately result in budget deficits that could threaten the nation's economic prosperity.

Both the Congress and the Administration will most likely try to resolve those problems well before the baby boomers begin drawing benefits. One approach that is often advocated for addressing the Social Security shortfall is "raising the retirement age." That phrase can have two different meanings, however. It is often used to describe a policy that would raise the age at which workers become eligible for full retirement benefits--the "normal retirement age." That age is already scheduled to increase from 65 to 67 by 2022. Some proposals would speed up that transition and then further increase the age to keep pace with future gains in life expectancy. Workers could still begin drawing Social Security benefits at 62 or any later age, but they would get less than they would under current law.

"Raising the retirement age" could also mean raising the earliest eligibility age for Social Security retirement benefits. Several Members of Congress and various groups have recommended doing that as part of their plans for alleviating future pressure on the economy and the federal budget. The recent proposal by the National Commission on Retirement Policy, for example, would raise the earliest eligibility age by two months per year starting in 2000, until it reached 65 in 2017, and would continue raising it to keep pace with increased life expectancy thereafter.⁽¹⁾

Raising only the normal retirement age could lower monthly benefits without reducing the number of years that benefits could be received.⁽²⁾ For example, when the normal retirement age increases from 65 to 67, workers who choose to start collecting benefits at age 62 will receive 70 percent of their full benefits rather than 80 percent. Similarly, workers will be able to receive about 87 percent of their full benefits at age 65 or wait two years to receive 100 percent. If the normal retirement age was increased beyond 67, workers collecting benefits before then would incur larger reductions.

Combining increases in the normal retirement age with increases in the earliest eligibility age could achieve similar savings by reducing the length of time that benefits would be received rather than the monthly amount of those benefits. Also, if raising the earliest eligibility age encouraged people to work longer, their need for retirement income would be reduced and the nation would have more resources.

An important issue in the debate over such proposals is whether raising the earliest eligibility age would present a serious hardship to many of the workers who would otherwise have claimed benefits at age 62. Opponents argue that a significant portion of those retirees would not have adequate resources without their Social Security benefits and that continuing to work would not be a practical alternative for them.

This paper addresses that issue by examining the income and characteristics of men and women who began receiving reduced Social Security benefits in recent years. It tries to discern what might have happened to those retirees if they had not been able to claim benefits until a later age. The analysis is based on information collected in the early 1990s by the Census Bureau--as part of the nationwide Survey of Income and Program Participation--on some 3,000 people age 62 or 63.

Although the survey provides a good basis for examining the retirees who would have been affected by a change in Social Security eligibility rules in the early 1990s, it must be used with caution when applied to a future change, for several reasons. Perhaps most important, the people who retired in the early 1990s had worked in an environment in which they could count on being able to receive Social Security retirement benefits at age 62 if they met the eligibility criteria. Thus, they could factor in those benefits when deciding about how much to try to save and how much to work. Although the eligibility age for Social Security benefits is not the only factor that goes into those decisions, research suggests that it is an important factor, at least for determining when workers retire. With a higher eligibility age for early benefits, it is reasonable to expect that people would have behaved differently. (Indeed, in an earlier era, when Social Security benefits for retired workers were not available until age 65, a much larger fraction of men continued to work until then.) Similarly, with a higher earliest eligibility age, employers might well have altered the terms of the pension plans they offer.⁽³⁾

Other differences exist between the people who were in their early 60s in the survey and future members of that age group. First, people in that age category in the 21st century are likely to be healthier and better educated, and a larger portion of the women will have had many years in the labor force. Second, they might well face a labor market in which opportunities are more plentiful than was the case in the early 1990s. The period in which the survey was taken happened to be one of above-average unemployment and downsizing.⁽⁴⁾ Third, in the 21st century, a smaller fraction of workers are likely to require physical strength to do their job.

Despite those caveats, several findings stand out about the resources of early Social Security beneficiaries and their ability to delay retirement. On the one hand, most of the workers who took early retirement benefits in the early 1990s had pensions and other income that would have been sufficient to keep them at least out of poverty had they not been able to start collecting benefits until age 65. Moreover, most of those workers did not appear to have health limitations or other problems that would have prevented them from working until a later age if necessary.

On the other hand, a significant minority of the people who took early benefits had very low outside income and might also have had serious problems working until a later age. For example, roughly one in 10 had non-Social Security income below the poverty threshold and said that they had a work-limiting disability.
 

CURRENT LAW

Under current law, the age at which a worker becomes eligible for full Social Security retirement benefits (the normal retirement age, or NRA) is at least 65, depending on the worker's year of birth. For those born before 1938, the NRA is 65. For somewhat younger workers, it increases by two months per birth year, reaching 66 for those born in 1943. The NRA remains at 66 for workers born from 1944 through 1954 and then begins to increase in two-month increments again, reaching 67 for workers born in 1960 or later.

Workers who retire and begin receiving benefits as early as age 62 incur a permanent reduction in the amount they receive each month.⁽⁵⁾ That reduction equals 5/9 of 1 percent for each month that the worker is short of the normal retirement age of 65. Thus, workers who start drawing benefits as soon as they turn 62 incur a maximum reduction of 20 percent (5/9 x 36). As the normal retirement age increases from 65 to 67, that maximum reduction will increase as well.

Spouses of retired workers are also eligible for reduced benefits when they turn 62, and widows of workers are eligible for reduced benefits at 60 (younger if they are caring for a minor). The maximum permanent reduction for people taking such auxiliary benefits before age 65 is 25 percent for spouses and 28.5 percent for widows.

The size of the early-retirement reduction for workers is intended to be "actuarially fair" in the sense that the total value of the reduced monthly benefits that an average worker could expect to receive between age 62 and death is similar to the total value of the full monthly benefits he or she could expect to receive over that time by delaying initial benefits until age 65. For example, ignoring discounting, a worker who retired at age 62 and expected to live 15 more years would be equally well off receiving reduced benefits of $800 per month for 15 years or unreduced benefits of $1,000 per month for 12 years (starting at age 65). In either case, those benefits would total $144,000. Because of discounting, however, the break-even point for life expectancy is higher than 15 years. Using the Social Security Administration's (SSA's) inflation-adjusted discount rate of 2.8 percent, a 62-year-old worker would earn the same total amount from early reduced benefits or delayed full benefits if he or she lived for about 17 more years (to age 79). However, although the life expectancy for 62-year-old men is approximately 17 years, for women it is about 21 years. With that longevity, a 62-year-old woman would theoretically accrue greater total benefits by waiting until normal retirement age to begin collecting them.⁽⁶⁾

The size of the actuarial reduction may encourage some workers to collect early benefits and discourage others. For example, workers who think their life expectancy falls well short of the average might view the 20 percent reduction as a good deal and apply for benefits at age 62. Conversely, workers who expect to live into their 80s might regard the actuarial reduction as unacceptably high and defer receipt of benefits until age 65.

Over two-thirds of the workers awarded Social Security retirement benefits during the past decade implicitly decided that the reduction in their monthly check was a price worth paying to begin collecting benefits before age 65. Indeed, the majority of those awards were made to workers at age 62.⁽⁷⁾

Because of the reduction rules, raising the earliest eligibility age to 65 by itself would have very little positive effect on the long-term financial outlook of Social Security, according to the SSA's Office of the Actuary. The reason is that the number of beneficiaries would be smaller, but benefits per recipient would be larger. If workers responded to their inability to receive benefits by delaying retirement, the labor force would expand, national income would grow, and the government would receive additional tax revenues. That effect, however, would be partly offset by the higher benefits most of those workers would subsequently receive because their benefits would be based on higher average earnings.
 

ISSUES IN RAISING THE EARLIEST ELIGIBILITY AGE

Proponents of raising the earliest eligibility age for Social Security retirement benefits argue that the federal government should no longer be helping people retire at age 62, for several reasons. First, with the coming shift in the age distribution of the population, it makes little sense to give up the productive capacity and revenues that would result from more people working longer.⁽⁸⁾ Second, as lifespans have increased and the average job has become less physically demanding, most people can work longer. Third, by enabling workers to trade lower future benefits for early access to benefits, the current early-retirement rules contribute to the higher poverty rates experienced by people who live to very old ages.

Proposals to raise the eligibility age are usually part of plans that would reduce the overall growth in Social Security benefits to help close the gap between the program's projected costs and revenues. Raising the eligibility age is one way to lessen the impact of those plans on the monthly benefit levels of beneficiaries in their later years. As noted earlier, when the normal retirement age rises from 65 to 67, the permanent reduction for retired workers who start receiving monthly benefits at 62 will be 30 percent rather than 20 percent (under current law). If the NRA was increased to 70, as some plans have called for, that reduction would be almost 45 percent. Raising the earliest eligibility age would be one way to prevent workers from incurring such a large reduction in their monthly benefits.

Opponents of raising the eligibility age contend that it would be especially harmful to people who have little or no choice about when they stop working and who have few resources other than Social Security. Those opponents argue that many low-earning workers are in physically demanding or unpleasant jobs and that by age 62, if not earlier, they have worked long enough. Moreover, by that age, opportunities for those workers are not very plentiful if they lose their job, particularly if there is a weak labor market. Studies of the experiences of displaced workers demonstrate that older workers with limited education have an especially difficult time finding a new job, and the job they do find is apt to pay substantially less than the one they lost.⁽⁹⁾

Another argument made against raising the earliest eligibility age is that doing so would be unfair to workers with below-average life expectancies, especially if they left no survivors eligible for benefits. Although few people can accurately predict when they will die, systematic differences exist between demographic and economic groups. For example, on average, men can expect to live fewer years than women, blacks fewer years than whites, people who have certain diseases fewer years than people who are healthy, and so forth.⁽¹⁰⁾

A related set of issues pertains to the ages at which people can become eligible for age-based survivor and spousal benefits. As noted earlier, widows and widowers are eligible for reduced benefits at age 60 (or younger if they were caring for a dependent child), and spouses are eligible for reduced benefits at 62. Some proposals to increase the eligibility age for retirement benefits would raise the ages for survivor and spousal benefits as well.

Yet another set of issues involves the eligibility criteria for related federal programs. If people ages 62 to 64 were no longer eligible for early retirement benefits, some would be likely to apply for and be awarded Social Security Disability Insurance (DI) benefits. The hardships caused by raising the eligibility age for retirement benefits could be further ameliorated by lowering the eligibility age for Supplemental Security Income (SSI) old-age benefits or relaxing the criteria under which workers qualify for DI benefits at age 62.⁽¹¹⁾ The cost of expanding SSI or DI could be substantial--not only because of the increased payments but also because of the links between those programs and federal health programs. Under current law, most SSI recipients immediately qualify for Medicaid, and DI beneficiaries qualify for Medicare with a two-year waiting period. Moreover, DI beneficiaries do not incur the permanent reduction in monthly benefits that early-retirement beneficiaries do, so their lifetime Social Security payments would actually be higher by switching to DI.
 

WHEN DO WORKERS USUALLY RETIRE?

Although 65 is still commonly considered the ordinary retirement age and is when workers become eligible for full Social Security retirement benefits, most people in the United States actually stop working before then. At age 55, five out of six men were in the labor force in 1997; at age 62, only half were; and at age 65, the labor force participation rate was down to one-third (see Figure 1).⁽¹²⁾ Likewise, although two-thirds of women were in the labor force at age 55, only 40 percent were at age 62.
 

FIGURE 1. LABOR FORCE PARTICIPATION RATES OF MEN AND WOMEN, BY SINGLE YEAR OF AGE, 1997

SOURCE: Congressional Budget Office based on data from the Bureau of Labor Statistics.

People who stop working in their early 60s can generally anticipate many years of retirement. As mentioned above, the average 62-year-old man today can expect to live another 17 years, and the average woman another 21 years, according to projections by the SSA's Office of the Actuary. The office projects that life expectancy will be about two years longer by the middle of the next century. Moreover, many demographers assert that the SSA's lifespan projections will turn out to be too low because of continuing advances in the medical field and healthier lifestyles.⁽¹³⁾

Such lengthy periods of postemployment leisure for men are a relatively recent phenomenon. At the turn of the century, men generally worked as long as they were healthy enough to do so. Even as late as 1950, nearly half of all men 65 or older were still in the labor force, compared with only one in six today (see Figure 2). By the mid-1980s, however, the employment rates of older men had fallen to near the current levels.⁽¹⁴⁾
 

FIGURE 2. LABOR FORCE PARTICIPATION RATES OF MEN AND WOMEN 55 OR OLDER, 1948-1997

SOURCE: Congressional Budget Office based on data from the Bureau of Labor Statistics.

The story is more complicated for women because of their rapid movement into the paid labor force. In 1950, only about 40 percent of women in their 30s and 40s, and even fewer older women, were in the labor force. As more women developed careers, their participation rates at all ages increased. But as with men, participation drops sharply well before age 65.

Individuals decide when to retire on the basis of a large number of factors, only some of which are directly affected by government policies. The age at which they become eligible for Social Security is obviously one factor determined by law. Their personal savings, job opportunities, and health, and the health and retirement decisions of other members of their family, are examples of factors not determined by the government.

Researchers have linked the early-retirement phenomenon to the growth in the nation's overall affluence.⁽¹⁵⁾ Social Security, pensions, and private savings have enabled workers to look forward to a time when they could afford to live without working and without being financially dependent on their children. Whole industries have developed to cater to the needs and desires of a retired population.

Researchers have also tried to determine the degree to which the structure of pensions and the Social Security program encourages workers to retire and the extent to which workers respond to such encouragement. In particular, a number of studies have shown that defined benefit pension plans--in which workers become eligible for a pension when they reach a certain age and have been with their employer a specified number of years--often provide an incentive for workers to retire as soon as they become eligible. For example, under the Civil Service Retirement System (the pension plan for most federal workers who joined the civil service before 1983), a 55-year-old with 30 years of federal service becomes eligible for an inflation-indexed monthly pension equal to 56 percent of his or her recent monthly earnings. Once they are eligible, workers might view their new situation as one in which their net pay is cut in half: they could continue working for, say, $50,000 or not work and get $28,000. Even if they recognize that continuing to work will increase their subsequent pension, they may well conclude that the increase would not be enough to make it worthwhile.⁽¹⁶⁾ Alternatively, they could leave federal employment, collect the pension, and work elsewhere.

Similarly, researchers have found that various aspects of the Social Security program may have contributed to the decline in labor force participation among older men (beyond simply making retirement more affordable).⁽¹⁷⁾ Many workers may view the gains in their future monthly Social Security benefits from working longer to be inadequate compensation for delaying those benefits. For some workers--such as those with below-average life expectancy--that view may be correct. For others, that view may be incorrect. But they may not understand the implications of their decision to take early benefits on their future well-being. The rules about actuarial reductions before the normal retirement age, delayed retirement credits after the NRA, and the amounts that beneficiaries may earn without losing their benefits are complicated and have changed many times.⁽¹⁸⁾

There is little doubt that the earliest age of eligibility for Social Security retirement benefits plays an important role in many people's decisions about when to withdraw from the labor force. The marked decline in the labor force participation rates of men, and to a lesser extent of women, at age 62 (as shown in Figure 1) suggests that many people do leave the labor force as soon as they become eligible for benefits. In 1997, for example, about 64 percent of 61-year-old men were in the labor force, but just 53 percent of 62-year-old men were. Similarly, for women the participation rates at 61 and 62 were 46 percent and 40 percent, respectively. Those drops are larger than the declines in participation at earlier ages.

Researchers have shown that the likelihood of retiring at age 62 was not especially high before benefits became available at that age. Moreover, they have found similar patterns in other countries that have different eligibility ages, providing further evidence that the age of earliest eligibility for benefits is an important factor in many people's retirement decisions.⁽¹⁹⁾

But it is not clear how much raising the earliest eligibility age for Social Security would increase total employment. Certainly, if people who would stop working and begin collecting benefits at age 62 under current law were denied eligibility until age 65, some of them would delay retirement--especially those whose only source of retirement income was Social Security. Moreover, such a major change in the Social Security program could have ripple effects on the work decisions of younger people as well. However, the majority of the workers who take early Social Security benefits also receive other pensions, and indeed many of them stop working before their Social Security benefits begin. They might well choose to withdraw from the labor force by age 62 whether or not they could collect Social Security.
 

WHO TAKES EARLY BENEFITS?

With over 1 million workers awarded early Social Security retirement benefits each year, anecdotes to fit almost any story can be found. Some of those retirees are in good health, would probably have no trouble working for many more years, and have large pensions and considerable financial assets. For them, delaying the start of Social Security benefits might not present any real problems; they could either continue to work or live comfortably off their pensions and other wealth. Other early retirees are in poor health, unattractive to potential employers, and have no other source of support. For them, delaying the start of benefits could cause a severe hardship unless they could qualify for Disability Insurance.

Data from the Census Bureau's Survey of Income and Program Participation (SIPP) provide a basis for systematically examining the economic circumstances, health, and other characteristics of 62- and 63-year-olds earlier in this decade. The Census Bureau selects respondents for the survey to be a nationally representative sample of the U.S. population. In the early 1990s, about 3,000 people who were age 62 or 63 took part in the SIPP. Nearly 2,000 of them were receiving early Social Security benefits. This paper's analysis excludes respondents who said they were 64 in order to reduce the chances of erroneously classifying someone as an early beneficiary who was actually 65.⁽²⁰⁾ (According to data from the Social Security Administration, most early beneficiaries in recent years were awarded benefits at age 62 or 63.)

Only one other analysis is available that provides directly relevant information about the characteristics and circumstances of recent early beneficiaries.⁽²¹⁾ That study--by Burkhauser, Couch, and Phillips--used a sample of about 1,200 people who were 62 in 1993 or 1994, drawn from the new Health and Retirement Study. Fewer than 400 of those 62-year-olds were receiving Social Security benefits. Although the Health and Retirement Study is likely to become the single best source of data about the retirement decisions of people born in the 1930s, for now the number of people in its sample who have reached Social Security retirement age is quite small, which limits the analyses that can be done.⁽²²⁾ The main finding of the study by Burkhauser, Couch, and Phillips was that fewer than one in 10 of the men who claimed Social Security retirement benefits at age 62 were both in poor health and in a household with no income from pensions. Estimates from the SIPP, reported below, confirm that result.

Characteristics of Early Retirees

The majority of the men and two-thirds of the women age 62 or 63 in the SIPP data were receiving Social Security benefits when first interviewed (see Table 1). Nearly all of the male beneficiaries were receiving retired-worker or Disability Insurance benefits, whereas about one-third of the female beneficiaries were receiving benefits as wives or widows.⁽²³⁾ The DI beneficiaries are excluded from the remainder of this analysis.
 

Distinguishing between retirement and receipt of Social Security benefits is important, as underscored by the fact that nearly one-quarter of the men who were receiving benefits were still working, and one-fifth of the men not receiving benefits were not working (see Table 2). One-third of the male beneficiaries (and a much smaller fraction of the nonbeneficiaries) had not worked since before turning 60.
 

The male beneficiaries were much less likely to have graduated from college or to have been in white-collar occupations than the nonbeneficiaries. That finding is consistent with the argument that workers in physically demanding jobs are more likely to take advantage of the option to retire early.

The comparisons between female Social Security beneficiaries and non-beneficiaries are generally similar to those for men. As with male beneficiaries, female beneficiaries were much less likely to be college graduates and were somewhat less likely to have been in white-collar jobs than women who were not receiving benefits. Likewise, female beneficiaries were much more likely to have stopped working before age 60 (or to have never worked) than other women. Female beneficiaries were also somewhat more likely than nonbeneficiaries to be married. (Reflecting the longer lifespans of women and the tendency of wives to be younger than their husbands, both female beneficiaries and nonbeneficiaries were much more likely than men to be widowed.)

Early beneficiaries of both sexes were generally well off economically, although they had lower median income than nonbeneficiaries (see Table 3). The median family income of male beneficiaries was about $29,000. A substantial portion of that came from sources other than Social Security, such as pensions and financial assets.
 

Net worth provides a better measure than income for comparing the economic status of beneficiaries and nonbeneficiaries because of the difference in their current employment. The typical male beneficiary in the SIPP was in a household with a net worth of $125,000.⁽²⁴⁾ Excluding the equity in its home, the household's median net worth was about $46,000. The typical female beneficiary was in a family with lower income and a smaller net worth.

Those numbers, of course, obscure important differences in the economic circumstances of individual beneficiaries. As other studies have shown, many beneficiaries do not have pensions or significant financial resources besides Social Security. Without Social Security benefits, those retirees would be poor unless they or members of their family worked more or saved more.

Among the early beneficiaries examined here, one-quarter of the men and nearly one-third of the women had non-Social Security income below the poverty threshold. (The poverty thresholds for individuals and couples in the early 1990s were cash incomes of about $7,000 and $9,000, respectively.) The men whose non-Social Security income was below the poverty threshold were much less likely than other men who took early benefits to have a pension or someone in the family who was still working (see Table 4). Moreover, among the men with low income who did have pensions or family earnings, the average amount of income they received from those sources was quite small.⁽²⁵⁾ Likewise, the beneficiaries with low non-Social Security income had very little income from property such as savings accounts and stocks.⁽²⁶⁾ Similar patterns were found for women (see Table 5).
 

Had early benefits not been available to them, undoubtedly many of the workers or their spouses would have either delayed retirement, worked more hours, or saved more. Thus, these estimates overstate the antipoverty effectiveness of Social Security for those beneficiaries. As the next section illustrates, information about the beneficiaries can be used to indicate how many of them might have had particular difficulty delaying retirement or working more hours if early benefits had not been available.

Who Might Be Dependent on Early Benefits?

Dependency is a subjective concept. In the context of the debate over raising the earliest eligibility age for Social Security, it includes two elements: the extent to which Social Security beneficiaries (and their families) have other sources of income, such as pensions and savings, and the extent to which beneficiaries who would not have an adequate income could delay retirement. Adequacy of income might be gauged either on an absolute scale (such as a poverty measure) or in comparison with the level of income while working. The presence of a pension other than Social Security can serve as at least a rough indication of whether a retired worker has a source of income that is related to his or her past earnings, in lieu of information about how much the worker previously earned.

A problem with any income criterion is that it does not take into account the effects that the absence of Social Security benefits would have on the individual's other potential sources of income. In particular, raising the earliest eligibility age for Social Security would cause some people (or their spouses) to delay retirement. Moreover, if workers knew that they would not be eligible for Social Security benefits until a later age, some of them who wanted to stop working by 62 would increase their saving rate so they could still do so. Others who stopped working at 62 anyway might qualify for Disability Insurance, Food Stamps, or other government transfer payments.

The analysis presented here uses three sets of information from the SIPP to identify people who might have particular difficulty holding a job after age 62. The first set is based on self-reported disability and is similar to the approach used in the study by Burkhauser, Couch, and Phillips (using a different data set).⁽²⁷⁾ SIPP respondents were asked whether they had any health condition that limited the kind or amount of work they could do. The second set is based on their response to a question about the main reason they had stopped working. People who said they had lost their last job or had left it because of health reasons are probably more likely to have trouble holding a job than those who said they stopped working because of retirement, old age, or dissatisfaction with the job. The third set of information is based on educational attainment. People who did not go beyond high school might have more difficulty finding new employment if they lost their job than people who went on to college.

As noted earlier, one-quarter of the men and nearly one-third of the women in the SIPP who were receiving benefits at age 62 or 63 had annual non-Social Security income below the poverty threshold. And those beneficiaries were much less likely than others to have any income from pensions or to have significant financial assets. If those beneficiaries had not been eligible for Social Security benefits, would many of them have had trouble working? Perhaps. One-third of the men said they had a physical, mental, or other health condition or disability that limited the kind or amount of work they could do (see Table 6). Nearly half reported not graduating from high school. And almost one-quarter said they had either lost their last job or left it because of a health problem. The percentage of women with non-Social Security income below the poverty threshold who said they had a work-limiting disability or had not graduated from high school was similar to that of men (see Table 7). A smaller fraction of those women, however, had lost their last job.
 

Neither self-reported disabilities nor the lack of a high school diploma necessarily translates into inability to work. Many people in those situations continue to work beyond age 62. But certainly, having a disability or not having a diploma is likely to limit job opportunities. For example, men in their late 50s and early 60s who were still working year-round on full-time schedules in 1997 and had not graduated from high school were earning less than half the amount that college graduates were earning.⁽²⁸⁾

That information on the income and employability-related characteristics of early beneficiaries in the SIPP can be used in various ways to provide alternative answers to the question, Were a substantial portion of the men and women who took early Social Security benefits dependent on them? As illustrated by the range of estimates in Table 8, the answer depends on whether it is based on a single characteristic or a combination of characteristics.
 

On the basis of either a simple poverty measure alone or the absence of a pension alone, roughly one-quarter to one-third of the early beneficiaries were dependent. Basing dependency on the presence of a work-limiting disability or the absence of a high school education provides a similar range of estimates. But if dependency is determined on the basis of being poor and having a work-limiting disability, its incidence falls to about one in 10. Other combinations of income-related and work-related problems produce similar estimates.
 

CONCLUSION

One way to resolve the projected shortfall between Social Security revenues and costs in the next century would be to slow the growth in spending for the program. Several plans to do that call for raising the earliest eligibility age for Social Security retirement benefits from 62 to 65. If that happened, workers who would otherwise have begun collecting benefits at age 62 either would have to get by on other sources of income until they reached the new eligibility age or would have to work longer. Proponents of raising the earliest eligibility age contend that most workers could do one or the other without serious problems. Opponents argue that by age 62 many people are dependent on Social Security benefits and do not really have the option of working longer.

The findings from this analysis--which examined the income and circumstances of workers who claimed early retirement benefits in the early 1990s--provide empirical evidence that bolsters both positions (though care must be taken in using those results to predict what would happen if the Social Security program was changed in the future). On the one hand, the majority of workers who took early-retirement benefits had pensions and other sources of income sufficient to keep them well above the poverty line even if they had not received Social Security. Moreover, many of the beneficiaries who did not have sufficient resources probably could have continued to work if necessary.

On the other hand, a sizable minority of early Social Security beneficiaries had non-Social Security income below the poverty threshold and might well have had serious difficulty finding a job. Their problems could be addressed by lowering the eligibility age for SSI or loosening the eligibility criteria for older workers applying for DI benefits. But either of those measures could be expensive.

1. See National Commission on Retirement Policy, The 21st Century Retirement Security Plan (Washington, D.C.: Center for Strategic and International Studies, May 19, 1998), available at http:/www.csis.org/retire/retireplan.html.

2. The main difference between raising the normal retirement age and imposing an across-the-board cut in benefits is that the latter approach would also reduce benefits for Social Security Disability Insurance. For a further examination of those and other options for slowing the growth of Social Security spending, see Congressional Budget Office, Long-Term Budgetary Pressures and Policy Options (May 1998), Chapter 3.

3. Often, employers integrate their pensions with Social Security, reducing the amount of the pension when the retired worker becomes eligible for Social Security. Had the earliest eligibility age for Social Security been higher, some of those employers might have offered smaller monthly pensions.

4. The average unemployment rate in the 1990-1993 period (which included a recession) was 6.7 percent--about 1 percentage point more than the average for the past half century.

5. Women and men have been permitted to receive reduced retired-worker benefits since 1956 and 1961, respectively. According to Robert J. Myers, the Chief Actuary of the Social Security Administration from 1947 to 1970, the reduction in the earliest eligibility age for women in 1956 resulted from a "domino" effect: a desire to help widows under 65 and a desire to provide immediate spousal benefits to wives who were younger than their husbands resulted in pressure to reduce the eligibility ages for those types of benefits; then, it seemed unfair to require female workers to wait until a later age for benefits than female nonworkers. The pressure to reduce the eligibility age for men in 1961 resulted from the high unemployment of that period. See Robert J. Myers, Social Security, 4th ed. (Philadelphia: University of Pennsylvania Press, 1993), pp. 518-520.

6. Many women might not incur the full early-retirement reduction because they can switch from receiving reduced retired-worker benefits to full survivor benefits upon the death of their husband. If the widow is at least the normal retirement age when her husband dies, she becomes eligible for a full survivor benefit (equal to his benefit) if that benefit is higher than the one she had been receiving based on her own earnings record.

7. Social Security Administration, Annual Statistical Supplement, 1997, p. 265. In 1996, 1.1 million of the 1.4 million awards for retirement benefits were to people ages 62 through 64. Over 800,000 of those awards were to 62-year-olds. (Those estimates exclude the 200,000 Disability Insurance beneficiaries who automatically became retired-worker beneficiaries when they reached 65.) Moreover, the vast majority of women who claimed benefits as wives of retired workers did so before 65 as well.

8. One advocate notes that if the earliest eligibility age was raised to 65, retired workers would still receive many more years of benefits than Social Security beneficiaries did when the program began. He calculates that the average retired worker at age 65 in 2010 would receive about 18 years of benefits, compared with the 13 years that a retiree could expect in 1940. See "Proposals for Long-Term Reform of Social Security," statement of C. Eugene Steuerle, Senior Fellow, Urban Institute, before the Senate Committee on the Budget, January 29, 1997.

9. For example, see Congressional Budget Office, Displaced Workers: Trends in the 1980s and Implications for the Future (February 1993). That study, as well as numerous other studies of displaced workers, was based on data representing all adults from biannual supplements to the Current Population Survey. Workers age 55 or older were less likely than younger workers to be displaced from their job, but those who were incurred larger losses. An analysis based on more recent data (from the Health and Retirement Study of workers in their 50s in 1992) also found that workers who lost jobs that they had held for at least five years typically incurred sizable earnings losses; see Kenneth A. Couch, "Late Life Job Displacement," Gerontologist, vol. 38, no. 1 (1998), pp. 7-17.

10. A report by the National Center for Health Statistics estimates that in 1996 average life expectancy was about 14 years for a black male at age 65, 16 years for a white male, 17 years for a black female, and 19 years for a white female. See Stephanie J. Ventura and others, "Births and Deaths: United States, 1996," Monthly Vital Statistics Report, vol. 46, no. 1, supplement 2 (September 11, 1997), Table 16.

11. For further discussion of the link between eligibility ages for Social Security retirement benefits, DI, and SSI, see 1994-1996 Advisory Council on Social Security, "Report of the Technical Panel on Trends and Issues in Retirement Saving" in Report of the 1994-1996 Advisory Council on Social Security, vol. 2 (January 1997), pp. 93-96.

12. Labor force participation rates are the best measure of a group's activity, but they do not depict individual movements into and out of the labor force. Some of the people not working at age 62 might go back to work later, at least on a part-time basis, and some of the people who were still working might consider themselves retired because they had left a job they had held for many years. See Joseph F. Quinn, "New Paths to Retirement" (paper presented at the Pension Research Council conference "Forecasting Retirement Needs and Retirement Wealth" at the Wharton School, University of Pennsylvania, April 1998).

13. For a critique of the techniques that the SSA uses to project life expectancy, see 1994-1996 Advisory Council on Social Security, "Report of the Technical Panel on Assumptions and Methods," in Report of the 1994-1996 Advisory Council on Social Security, vol. 2 (January 1997), pp. 148-153.

14. From 1993 to 1997, the participation rate of men ages 65 to 69 increased by 3 percentage points to 28.4 percent, and those of men ages 55 to 59 and 60 to 64 increased by smaller amounts. It is too soon to tell whether those increases resulted from the recent overall strength of the labor market or from more fundamental factors.

15. See Dora L. Costa, The Evolution of Retirement (Chicago: University of Chicago Press, 1998).

16. The civil service pension is based on a worker's average earnings in the highest three years, typically the most recent three-year period. Workers with 30 years of service are eligible for 56 percent of that average pay. Each additional year of service will add 2 percentage points. If workers expect that their earnings growth will match the inflation rate, then an additional year of service will add only 3.6 percent (2 ÷ 56) to their monthly pension. That amount might seem too little to warrant giving up one year of the pension.

17. That literature is summarized in 1994-1996 Advisory Council on Social Security, "Report of the Technical Panel on Trends and Issues in Retirement Saving," pp. 12-27. In an earlier article, Franco Peracchi and Finis Welch demonstrate the importance of considering the decline in the labor force participation rates of older men in a broader labor market context that encompasses, for example, the reductions in the participation rates of younger men as well. See Peracchi and Welch, "Trends in Labor Force Transitions of Older Men and Women," Journal of Labor Economics, vol. 12, no. 2 (1994), pp. 210-242.

18. For example, many people seem to be unaware that the NRA will increase from 65 to 67. Also, it is doubtful that very many people understand how amounts earned while receiving benefits can increase their subsequent benefits.

19. That evidence is summarized in Jonathan Gruber and David Wise, Social Security Programs and Retirement Around the World, Working Paper No. 6134 (Cambridge, Mass.: National Bureau of Economic Research, August 1997).

20. The appendix provides further information about the SIPP and compares the information about Social Security beneficiaries in that survey with information from the Social Security Administration.

21. Richard V. Burkhauser, Kenneth A. Couch, and John W. Phillips, "Who Takes Early Social Security Benefits? The Economic and Health Characteristics of Early Beneficiaries," Gerontologist, vol. 36, no. 6 (1996), pp. 789-799.

22. More than 12,000 people in 7,700 households were interviewed in 1992, and most of them were interviewed again two years later. The study's plan calls for conducting at least four more interviews, scheduled at two-year intervals, and for linking the information provided by the respondents with information obtained (with their permission) from their Social Security and Medicare records and their employers' health and pension plans.

23. The appendix compares the estimates of beneficiaries in Table 1 with estimates from the Social Security Administration's administrative records. The only notable difference is that a higher percentage of the female beneficiaries in the SIPP reported that they received their benefits as retired workers rather than as spouses or survivors. The SIPP designers paid particular attention to developing and testing procedures to accurately identify why someone was receiving Social Security benefits. See Denton R. Vaughan, "Development and Evaluation of a Survey-Based Type of Benefit Classification for the Social Security Program," Social Security Bulletin, vol. 52, no. 1 (January 1989), pp. 12-26.

24. Estimated net worth includes reported financial assets minus liabilities and net equity in homes. It does not include the value of future Social Security benefits or pensions, if any.

25. Social Security beneficiaries under age 70 who work are subject to an earnings test. For beneficiaries ages 62 to 64, each dollar of earnings above a specified threshold reduces their benefit by 50 cents. The annual earnings threshold is indexed to increase at the same rate as the growth of wages. Between 1990 and 1993, it rose from $6,840 to $7,680. A monthly threshold applies during the first year of retirement.

26. Spending down their assets would be one way that retired workers could try to maintain living standards if the eligibility age for Social Security benefits increased. However, beneficiaries with low non-Social Security income generally would have few financial assets to liquidate. Their median household net worth, excluding home equity, was less than $12,000 for men and less than $8,000 for women.

27. Burkhauser, Couch, and Phillips, "Who Takes Early Social Security Benefits?" pp. 789-799.

28. Bureau of the Census, Money Income in the United States: 1997, p. 35.

---