Testimony of Kenneth Apfel Commissioner for Social Security Hearing before Subcommittee on Social Security The House of Representatives Committee on Ways and Means
February 15, 2000

Good morning, Mr. Chairman and Members of the Subcommittee. Thank you for inviting me to appear this morning to discuss issues related to the Social Security retirement earnings test (RET). First, I would like to discuss the President's framework for Social Security reform. Then, I will discuss another major concern of the Administration; the importance of Social Security to the economic well-being of elderly women and the need to improve their protection under the program. Finally, I will explain how the Social Security retirement earnings test has changed over the years and discuss the implications of eliminating the test.

The President believes that it is important to modernize the system by eliminating the outdated retirement earnings test at normal retirement age (NRA). The President would also like to work together to use the benefits of debt reduction to extend the solvency of Social Security to about 2050 and improve the effectiveness of Social Security in combating poverty among elderly women. He remains committed to working together with Congress on a bipartisan basis to enact reforms that make Social Security solvent for at least 75 years.

President's Budget Framework

Let me begin today by discussing the President's budget proposal as it pertains to Social Security reform. I strongly support the President's proposal. Maintaining fiscal discipline and paying down the debt gives us an historic opportunity to meet the challenges of the future. The President proposes to devote the Social Security surpluses to improving the balance sheet of the Federal government, and to transfer the resulting interest savings to the Social Security trust fund. In addition, his plan calls for investing a limited amount of the trust fund in equities as a prudent and workable solution for extending the life of the trust fund yet further into the future. This budget proposal is an important first step towards crafting a bipartisan agreement between the President and Congress that will keep faith with future generations of Americans.

As you know, the Social Security program faces a long-range deficit of 2.07 percent of taxable payroll under the intermediate assumptions of the 1999 Trustees Report. Because Social Security is fundamental to the economic well being of our aged population, ensuring the long-range solvency of the Social Security program must be one of our highest priorities. That is why the President's budget framework to preserve and strengthen Social Security is so very important. The President has proposed the following specific actions:

First, the President's framework provides for locking away the entire Social Security surplus each and every year. All Social Security surpluses, an estimated $1.7 trillion over the next 10 years under Social Security Trustees assumptions, would be devoted solely to improving the balance sheet of the Federal government and strengthening Social Security (under Administration assumptions, the Office of Management and Budget projects a $2.2 trillion Social Security surplus over 10 years). This framework will ensure that we achieve substantial public debt reduction, helping to prepare the government and the Nation for the retirement of the baby boomers.

The framework provides for the transfer of interest savings based on the cumulative amount of Social Security surpluses we actually experience over the next 15 years. The President's framework provides for transfers in years 2011 through 2050 to Social Security. The Social Security actuaries project that the interest savings that would result from setting aside the Social Security surpluses (assuming all the new resources are invested in government securities) would total $99 billion in 2011 and grow to $205 billion by 2016. Total transfers between 2011 and 2015 would be $690 billion. The transfers would extend the solvency of the trust funds until 2050.

Also, beginning in 2011, the framework calls for investing a sensible and measured proportion of the transfers in the equity market to achieve higher returns for Social Security. The equity investment would be limited to 50 percent of the cumulative transfer amounts, to the degree that these did not exceed 15 percent of the trust funds. The Social Security actuaries project that the Trust Fund's equity holdings would represent, on average, about 3 percent of the stock market over the 30-year period 2011-2040. Funds would be invested in broad market indexes by private managers, not the government. This would further extend the solvency of the trust funds to 2054, compared to the current projected exhaustion date of 2034.

Now is the time for action. If we act now, before there is a Social Security financing crisis, and while we enjoy the first budget surpluses in a generation, we can prevent a financing crisis from ever occurring. If we delay action for a generation, the size of the financing problem will grow. We have an historic window of opportunity to meet the challenge facing Social Security. . . and we must not let this opportunity slip away.

Importance of Social Security for Women

I now want to talk about the importance of Social Security in the economic security of women and why any comprehensive reform of Social Security must address the high incidence of poverty among elderly women.

For 60 years, Social Security has provided a solid floor of financial protection in the event of a worker's retirement, death, or disability. It has allowed the great majority of Americans to retire with the dignity that comes from financial independence, without fear of poverty or reliance on others.

No government program has had a more positive impact on the lives of older women than Social Security. There can be no doubt - Social Security is a vitally important element in the retirement income security of our sisters, our mothers, our grandmothers, and our great grandmothers.

The President is committed to helping elderly women, who typically have higher poverty rates than other elderly. On numerous occasions, the President has made it clear that he wants to address their situation as part of the effort to close the long-range deficit in Social Security. He has stated that "We should reduce poverty among elderly women who are nearly twice as likely to be poor as our other seniors."

Even though Social Security does a good job of keeping most elderly families above the poverty threshold, poverty rates vary greatly between different groups. For example, poverty rates are higher among nonmarried women than married women beneficiaries.

Only 5 percent of aged married women are poor; in contrast, 22 percent of divorced, 20 percent of never-married, and 18 percent of widowed women age 65 and older are poor.

Widows account for the largest proportion (66 percent) of poor aged beneficiary women. There are 1.2 million aged widows who receive Social Security benefits and have incomes below the poverty line ($7,818 for an aged individual in 1998).

Social Security Reform and Women

We need to be aware that Social Security reform proposals can have very different effects on the benefits received by women compared to those received by men. These differences stem from the fact that, although Social Security program rules are gender neutral, individuals are affected differently because of their lifetime earnings patterns and life expectancies differ.

Women tend to have lower lifetime earnings and work fewer years in covered employment than men and, because of their longer life expectancies, will spend more time than men in retirement. Therefore, the possible differential effects of any proposed program changes on women need to be closely reviewed as we discuss the options and trade-offs of ways to ensure the solvency of Social Security.

Income security remains an elusive goal for many elderly women. This is why a comprehensive Social Security reform package must not only achieve solvency but include provisions to protect elderly women. Elderly women often are more dependent on Social Security because they are less likely to have pensions and sometimes outlive their assets. Almost three-quarters of Social Security beneficiaries over age 85 are women. It is essential that we work together in a bipartisan effort to ensure that they have the best protection that society can provide.

The Retirement Earnings Test

Now, I will discuss the issue of the retirement earnings test (RET). Let me begin by briefly reviewing the philosophy behind the earnings test and how that philosophy, and the test itself, have changed over the years.

Social Security was designed as a social insurance program under which workers and their dependents were to be insured against the loss of earnings as a result of retirement, disability, or death of the worker. Benefits are intended to partially replace the earnings that are actually lost due to these events. In that context, the retirement earnings test was designed as an objective measure of the extent to which earnings are lost due to retirement.

The Social Security program has always had an earnings test. However, the "all-or-nothing" test in the original 1935 Social Security Act has been modified numerous times to allow retirees to supplement benefits with earnings up to a specified level. Even before the first benefits were paid in 1940, the test of retirement was modified so that a beneficiary could earn up to $14.99 in covered earnings before losing benefits for that month.

Since 1940, many other changes to the retirement earnings test have been made. The Social Security Amendments of 1950, for example, exempted people age 75 and over from the earnings test. In 1954, the retirement test was broadened to include non-covered wages, and the age at which the test no longer applied was lowered from 75 to 72. The concept of reducing benefits by $1 for each $2 of earnings above the exempt amount was introduced in the Social Security Amendments of 1960, and the 1972 Amendments provided for the earnings test exempt amount to be increased automatically with increases in average wage levels. In 1983, the age at which the test no longer applies was lowered to 70. In 1990, the withholding rate of $1 of benefits for each $2 of earnings was changed to $1 for $3 for beneficiaries aged 65 to 69.

The most recent change to the retirement earnings test occurred in 1996. With the strong support and leadership from the President, the annual exempt amounts for beneficiaries aged 65 to 69 was legislated to rise annually. This year the annual exempt amount is $17,000; in 2001, it will be $25,000, and by 2002, it will reach $30,000. This increase gives many older Americans the opportunity to supplement their Social Security benefits while remaining productive members of the workforce.

The Current Retirement Test

Under current law, for beneficiaries age 65-69 in 2000, benefits are reduced $1 for every $3 of earnings above $17,000 per year. For beneficiaries between the ages of 62 and 65, benefits are reduced $1 for every $2 of earnings above $10,080 per year. Unearned income, such as interest income, dividend payments, private pensions and the like, is not counted for purposes of the retirement earnings test.

In addition, workers are exempt from the test when they reach age 70. For a worker below age 70, his or her earnings above the exempt amount affect not only his or her own benefits, but also the benefits of family members receiving benefits on the worker's earnings record. However, if a dependent or survivor beneficiary has earnings above the exempt amount, those earnings can affect only that individual's payments.

Delayed retirement credits (DRCs) are provided to compensate workers age 65-69 whose benefits are withheld under the retirement earnings test. The DRC increases the worker's retirement benefit for each month that benefits are fully withheld after the full benefit retirement age, now age 65 but scheduled to rise to 67 by 2022. The DRC is currently 6-percent per year for workers age 65 in 2000. The DRC percentage will increase 0.5 percentage point every two years until it reaches 8 percent per year for workers reaching age 65 in 2008 and later. When the DRC is 8 percent per year, benefits lost due to the retirement earnings test and/or delayed retirement generally will be offset in an actuarially fair manner by the increase in benefits resulting from DRCs.

The present-law actuarial reduction provisions, in conjunction with the earnings test for workers aged 62-to 65, are designed to provide the early benefit claimant who works, on average, with the same total lifetime benefits as would be received if benefits had started at age 65. A person who files a Social Security claim before reaching the full benefit retirement age receives a reduced benefit. However, once the person reaches age 65, the benefit payment is adjusted upward to account for any benefit amounts withheld due to earnings prior to age 65. The effect of this procedure is that whenever a monthly check, otherwise payable to someone between 62 and 65, is partially or totally withheld under the earnings test, the amount "lost" is repaid, on average, over the course of the beneficiary's remaining lifetime beginning at 65.

In 1999, an estimated 1.2 million beneficiaries had some or all of their benefits withheld for some portion of the year under the earnings test due to work at age 62 or above. About 800,000 beneficiaries lost some or all of their benefits under the test as a result of their work at ages 65-69. The benefits of 150,000 auxiliary beneficiaries are also limited or withheld due to the earnings of the primary beneficiary. With respect to beneficiaries aged 62-64, about 230,000 working beneficiaries had all or part of their benefits withheld, and 25,000 auxiliary beneficiaries are affected.

Issues Associated with Eliminating the Retirement Earnings Test

As I indicated earlier, the President has said that we should eliminate the retirement earnings test. However, an important issue is whether the RET should be eliminated at age 62 or at the normal retirement age, currently age 65, scheduled to gradually rise to 67. (For example, workers born in 1938 and eligible for early retirement at age 62 this year, have a normal retirement age of 65 and 2 months.) This issue involves important trade-offs. On one hand, the RET is confusing to beneficiaries, and probably reduces their work effort to some degree. It is also difficult to administer. Eliminating the test would end these problems. On the other hand, eliminating the RET at age 62, by itself with no other changes, would likely increase poverty for many older beneficiaries--particularly elderly women--and would increase the long-range program deficit by a small, but measurable amount. Eliminating the test at the normal retirement age would have a negligible impact on poverty for older beneficiaries and would not result in a long-range program cost. Let me elaborate on these points.

Eliminating the RET could have an effect on the choice of older Americans of whether and how much to work. Although the benefit witholdings under the RET are roughly offset by higher benefits later on, many people perceive the RET as a tax on their labor income. Eliminating this perceived disincentive would have two effects: (1) some people would choose to remain in the workforce or to continue to work full-time because they would not face the same reduction in their current Social Security benefits; and (2) some people would choose to work less, making up for lower earnings with higher current Social Security benefits.

There is only a limited amount of evidence on the overall effect of the RET on labor supply and it is impossible to form a definitive conclusion. It does, however, seem plausible that eliminating the RET would lead to a modest increase in work activity, an assumption that is reflected in the Social Security actuaries' estimates of the impact of eliminating the RET on Social Security solvency.

Eliminating the RET below the normal retirement age might also change the risks that older Americans face down the road. Eliminating the earnings test would result in more workers electing to receive benefits as soon as they become eligible (about 60 percent of workers already claim benefits at age 62). The decision to claim benefits earlier reduces these individuals' monthly Social Security benefits. The lower benefit is intended to be actuarially fair so that, over their lifetimes, beneficiaries receive, on average, the same total benefits. However, many people do not perceive that these benefits are paid back to them over their lifetime. Once beneficiaries who claimed reduced benefits stop working, they may not have sufficient outside resources to offset this reduction in Social Security benefits.

And I want to emphasize this point: the widow(er)s of workers who retire before full benefit retirement age also get permanently reduced benefits due to the worker"s decision to retire early. Thus, elimination of the RET at 62 could also have a negative impact on the number of elderly women living in poverty in the future. The poverty impact of eliminating the RET between age 62 and the NRA is dependent on changes in filing behavior and changes in work/retirement decisions, and cannot be predicted exactly.

At present, most women who have ever been married ultimately receive benefits based on their deceased husband's earnings record. Age reductions in the deceased worker's benefit are generally passed on to the widow(er)'s benefit. Thus the worker's decision to take benefits at 62 would result in lower, perhaps inadequate, benefits for his survivor many years later. A very likely consequence of eliminating the earnings test for those below NRA, if no other changes were made, would be an increase in the number of elderly widows who are poor.

The Administration believes that the best policy is to eliminate the earnings test at NRA. Eliminating the RET at age 62 raises serious concerns about increasing poverty among elderly women, and we would not want to consider it without at least making sizeable changes to the program to mitigate these deleterious effects.

Next Steps

The Administration believes that we can make a substantial down payment on Social Security reform with two simple, clear bipartisan steps:

The first step is to pass a straightforward bill to repeal the RET. If Congress sends the President a clean bill to repeal the RET at NRA, with no extraneous, non-Social Security matters whatsoever, he will sign it.

The second simple step is for Congress to pass and send a bill to the President that would extend the solvency of Social Security to about 2050 and include significant measures to reduce poverty among elderly women. The President has given Congress straightforward legislation that would simply ensure that we devote the interest savings earned by paying down the publicly-held debt to making Social Security stronger. By agreeing to this simple step, we can extend the life of Social Security to the middle of the next century.


The President has shown strong leadership in providing a framework for preserving the financial well-being of the Social Security program. If adopted, this framework would take us a long way towards closing the long-range actuarial gap of 2.07 percent of taxable payroll. It gives us a solid foundation on which to preserve our social insurance program throughout this century. And it does much more: President Clinton's approach would pay down the publicly held debt, thereby increasing national savings and promoting economic growth, which will reduce burdens on future generations.

Just as the Administration and the Congress worked together successfully to tackle the economic challenges facing this nation and put our fiscal house in order, we can also work together to eliminate the RET in the right way. And I believe that we can work together to resolve the long-term Social Security solvency issues as well. We have an opportunity that we could not have imagined just a few years ago. We can begin to deal with the future, and address long-term generational challenges. We must seize this moment and focus on strengthening and protecting the Social Security system for future generations of Americans.

I will be happy to answer any questions the Members may have.