1994-96 Advisory Council Report



In her charge to the Advisory Council, the Secretary of Health and Human Services, in consultation with the Commissioner of Social Security, asked the Council to look particularly at the long-term financing of Social Security. This is the first Advisory group since 1979 to give its major emphasis to long-term financing questions in its review of the system. While the 1982-83 National Commission on Social Security Reform did examine long-range financing, much of its time was spent on short-term financing and the pressing question of how best to finance the system from 1983 to 1989, because the Social Security Trust Fund resources faced imminent exhaustion. The 1984 Council focused on Medicare. The Council that would have been appointed in 1986 was replaced with a special advisory panel to evaluate the Social Security Disability Insurance program. The 1991 Council concentrated on health insurance in general, including Medicare.

While the Council has not found any short-term financing problems with the Old-Age, Survivors, and Disability (OASDI) program, there are serious problems in the long run. Because of the time required for workers to prepare for their retirement, and the greater fairness of gradual changes, even long-run problems require attention in the near term.

Except for the Consumer Price Index assumption (discussed below), the Council used the intermediate assumptions in the 1995 Trustees Report throughout its deliberations and recommendations, and the numbers that appear throughout the report are based on those assumptions. Although the 1996 report was issued while the Council was still in session, conforming the Council's report throughout would have required extensive additional work and time, for little purpose, because the differences between the findings of the 1995 and 1996 reports are minor in consequence.

The Council identified four major areas of concern.

Long-Term Balance

Under their intermediate assumptions, the Trustees of the Social Security Funds estimated that income (the sum of the revenue sources plus interest on accumulated funds) will exceed expenses each year until 20201 . The trust fund balances will then start to decline as investments are cashed in to meet the payments coming due. The Trustees estimated that although 75 percent of costs would continue to be met from current payroll and income taxes, in the absence of any changes full benefits could not be paid on time beginning in 2030.

The deficit over the traditional 75-year projection period was 2.17 percent of taxable payroll. This means that if payroll tax rates had been increased in 1995 by just over 1 percentage point each on employers and employees -- from their present level of 12.4 percent combined rate to 14.57 percent combined rate (excluding Medicare) -- the system would be in balance over this 75-year period. In the early decades of the projection period there would be surpluses, followed by deficits later in the projection period, but because of earnings on the trust funds, the 2.17percent payroll tax increase would eliminate the 75-year deficit.

Little support exists today for increasing payroll tax rates by 2.17 percentage points to provide long-term balance. But there are other ways to address the financing issue, including other ways of increasing income to the system and changes in benefits. The program can be brought into long-run balance without departing from its basic principles or undermining the economic well-being of future workers and program beneficiaries.

The Council's work, and the work of a task force of experts appointed by the Council to review the estimates of the Trustees, basically confirm the 1995 Trustees' estimates of the finances of the program. Consequently, one of the three major tasks the Council set for itself in the area of financing was to make recommendations that would eliminate the 2.17 percent of taxable payroll deficit. All members of the Council agree that this should be done, though there are differences of opinion on how the goal should be met.

Long-Term Balance Beyond the 75-Year Horizon

The second major problem with Social Security financing is the deterioration in the program's long-range balance that occurs solely because of the passage of time. Because of the aging of the U.S. population, whenever the program is brought into 75-year balance under a stable tax rate, it can be reasonably forecast that, without any changes in assumptions or experience, the simple passage of time will put the system into deficit. The reason is that expensive years previously beyond the forecasting horizon, with more beneficiaries getting higher real benefits, are then brought into the forecast period. There is no simple answer to the question of how much higher the long-term actuarial deficit is above the 2.17 percent to bring Social Security into balance beyond the 75-year horizon, but there could be a significant increase2 . All members of the Council agree that it is an unsatisfactory situation to have the passage of time alone put the system into long-run actuarial deficit, though there are again differences on how the problem should be corrected.

Contribution/Benefit Ratios

The third area of concern for the Council arises from the fact that from now on many young workers and workers of future generations under present law will be paying over their working lifetimes employee and employer taxes that add to considerably more than the present value of their anticipated benefits. This is the inevitable result of a pay-as-you-go system such as the United States has had, and an aging population. Although the money's worth that workers get from Social Security is only one of many criteria for judging the value of the Social Security system, the Council believes that the system should meet a test of providing a reasonable money's worth return on the contributions of younger workers and future generations, while taking account of the redistributive nature of the Social Security system.

The Council is breaking new ground by dealing so explicitly with money's worth issues. It does so because of concerns about equity from one generation to another. The Council feels that equity among generations is a serious issue and that it is important to improve the return on retirement saving for young people.

All members of the Council favor the objective of improving the money's worth given by Social Security to younger generations. There are again differences on how this objective should be achieved.

Public Confidence

The final issue involves public confidence in the system. Polling data suggest that younger people have unprecedentedly low levels of confidence that Social Security benefits "will be there" for them when they retire. Polling data also suggest some erosion in public confidence in Social Security over time. While some of this skepticism runs well beyond issuesthe Council was dealing with, the Council does want to reassure people about the future viability and fairness of Social Security.

* * *
In the past, efforts to deal with Social Security's financial difficulties have generally featured cutting benefits and raising tax rates on a pay-as-you-go basis. All Council members agree that the pay-as-you-go approach should be changed. But despite its best efforts, the Council was not able to agree on one single plan for dealing with Social Security's financial difficulties. Rather, Council members expressed interest in three different approaches to restoring financial solvency and improving money's worth returns. One group of members favors an approach, labeled the Maintenance of Benefits (MB) plan, that involves an increase in income taxes on Social Security benefits, a redirection to the OASDI funds beginning in 2010 of the part of the revenue from taxes on OASDI benefits now going to the Hospital Insurance (HI) Trust Fund, coverage of newly hired State and local government workers not currently covered by Social Security, a payroll tax increase in 2045, and serious consideration of a plan allowing the Government to begin investing a portion of trust fund assets directly in common stocks indexed to the broad market. Historically, returns on equities have exceeded those on Government bonds (where all Social Security funds are now invested). If this equity premium persists, it would be possible to maintain Social Security benefits for all income groups of workers, greatly improving the money's worth for younger workers, without incurring the risks that could accompany individual investment3 .

Another group of members supports an approach, labeled the Individual Accounts (IA) plan, that creates individual accounts alongside the Social Security system. This plan involves an increase in the income taxation of benefits (though not the redirection of HI funds), State and local coverage, an acceleration of the already-scheduled increase in the age of eligibility for full benefits up to year 2011 and then an automatic increase in that age tied to longevity, a reduction in the growth of future Social Security benefits is structured to affect middle- and high-wage workers the most, and an increase in employees' mandatory contribution to Social Security of 1.6 percent of covered payroll, which would be allocated to individual defined contribution accounts. These individual accounts would be held by the Government but with constrained investment choices available to individuals. If individuals were to devote the same share of their IA funds to equities as they now do for their 401(k) private pension funds, the combination of the annuity income attributable to their individual accounts and their scaled-back Social Security benefits would on average yield essentially the same benefits as promised under the current system for all income groups.

A third group of members favors an approach, labeled the Personal Security Accounts (PSA) plan, that creates even larger, fully-funded individual accounts which would replace a portion of Social Security. Under this plan, workers would direct 5 percentage points of the current payroll tax into a PSA, which would be managed privately and could be invested in a range of financial instruments. The balance of the payroll tax would go to fund a modified retirement program and modified disability and survivor benefits. When fully phased in, the modified retirement program would offer all full-career workers a flat dollar benefit (the equivalent of $410 monthly in 1996, the amount being automatically increased to reflect increases in national average wages prior to retirement) plus the proceeds of their PSAs. This plan also would involve a change in benefit taxation, State and local coverage, an acceleration of the already-scheduled increase from 65 to 67 in the age of eligibility for full retirement benefits, with the age increased in future years to reflect increases in longevity, a gradual increase from 62 to 65 in the age of eligibility for early retirement benefits (although workers could begin withdrawing the proceeds of their PSAs at 62), a reduction in future benefits for disabled workers, a reduction in benefits for women who never worked outside the home, and an increase in benefits for many elderly widows.

If individuals allocated the assets in their PSAs in the same proportion as they do for their 401(k) private pension plans, the combination of the flat benefit payment and the income from their PSAs would, on average, exceed the benefits promised under the current system for all income groups. There would be a cost associated with the transition to this new system equivalent to 1.52 percent of payroll for 72 years. This transition cost would be met through a combination of increased tax revenues and additional borrowing from the public.

All of these approaches have in common that they seek to achieve more advance funding of Social Security's long-term obligations. They would also result in a higher level of national saving for retirement, although the impact on the nation's overall retirement saving would differ under the plans. The two individual account plans would raise overall retirement and national saving much more than the MB plan in the early years of the forecast horizon through the mandatory contributions of the IA plan or the transition tax of the PSA plan. These two plans are then likely to generate higher national income in the 21st century. While each of the proposals would increase investment in the stock market, one approach invests new Social Security funds directly into equities to realize a higher rate of return; another approach adds additional, mandatory saving on top of a scaled-back version of the existing benefit system; and the third approach moves from the current pay-as-you-go, largely unfunded system to one in which future benefits are more than 50 percent funded through PSAs. Each of these plans has different potential to create real wealth for retirement and provides for different ownership of that wealth. And each involves a very different vision for the future evolution of the U.S. retirement system.

This report first outlines those principles on which there was broad agreement within the Council. It covers recommendations for the general Social Security system, recommendations regarding research and data, and recommendations regarding private pensions. It then describes the three different proposals in more detail. There are then separate statements by Council members. There are three appendices, one giving an analysis of changes in the long-range actuarial balance since the 1983 Social Security reforms, one giving detailed actuarial evaluations of the proposals under alternative assumptions, and one providing cost information on some other proposals that have been made to resolve Social Security's financial problems. Also included are summary statements of two technical panels of experts commissioned by the Council -- one involving the assumptions and methods used by the Trustees, and one involving trends and issues in retirement saving. This volume is followed by a second volume that gives the complete reports of the two technical panels, along with the write-ups of many of the presentations given to the Council over the twoyears in which it met4 .


The Social Security system is part of a four-tier system of retirement income arrangements in which each tier is important and complementary to the others: Social Security, employer-sponsored pensions, individual savings, and a safety-net program, called Supplemental Security Income (SSI).

Social Security is the universal system that forms a base for the other three tiers, but all are important and need to be improved. Social Security and SSI are programs of the Federal Government and are directly administered by it. Employer-sponsored pensions and certain forms of individual saving are encouraged by Federal tax policy. All have an important role in providing for retirement income.

Social Security's replacement rates -- that is, the ratio of retirement benefits to final wages -- are relatively high for low-wage earners because it is expected that they will have less help from private pensions and individual saving. Replacement rates are lower for higher paid workers who are more likely to be covered by employer pensions and who are better able to save on their own. SSI is designed to make sure that the most impoverished of the elderly, blind, and disabled are brought up to a minimum standard of living. If the Social Security system were to try to perform this function, it would lose its character as a wage-related contributory system. Thus, the Social Security system is designed to form part of a four-tier system for retirement income, and the Council believes it should continue to be structured in this way.

The Council believes that Social Security should continue to take responsibility for paying benefits to low-wage, full-time, regular workers that are sufficient to keep them from having to turn to means-tested assistance for support in old age. But it is SSI's role to support those who, for whatever reason -- part-time work, irregular employment histories -- have incomes that fall below a standard of minimal adequacy when all sources of income are taken into account.

It is of great importance to the nation that this four-tier retirement system, based on a compulsory Social Security system, be continued.

Later in this report the Council will suggest alternative ways in which the long-range cost of Social Security can be brought into balance with its long-range income. The Council rejects the proposition that Social Security will "not be there" for future generations. Indeed, the Council proposes steps to improve the money's worth ratios for these younger workers.

Social Security is a defined-benefit plan, under which the government agrees to provide specific benefits based on past earnings. One goal of the system has long been universal coverage, but it has taken time to accomplish that goal. Since the program is now just about universal, workers are guaranteed that their protection will follow them from job-to-job. Today 96percent of workers in paid employment are covered. About 141 million persons will make contributions in 1996. They are buying more than retirement protection alone. Ninety-eight percent of children under age 18 can count on the protection of monthly cash benefits if a working parent dies, and about 80 percent of men and women workers ages 21 to 64 can count on monthly cash benefits in the event of a severe and prolonged disability.

Social Security requires that all workers -- provident and improvident alike -- contribute to their future security. A compulsory program ensures that these contributions take place. Compulsion also makes it possible to provide for redistribution of protection from the higher to the lower paid and to avoid the problem of adverse selection that could occur if individuals were allowed to decide when and to what extent they wished to participate. Compulsion reduces the need for public assistance. In contrast, a voluntary plan would allow the improvident to escape their share of paying for their own future retirement needs -- leaving the community as a whole to pay for them through some safety net program like SSI.

The Council favors partial advance funding for Social Security.

Historically, Social Security has been financed on a current pay-as-you-go basis, with tax inflows barely covering benefit outflows. As a result of the 1977 and 1983 Social Security Amendments, the Social Security Trust Funds began to accumulate some reserves, now about $500 billion (140 percent of next year's outflow), in advance of the sharply rising retirement costs of the baby boom generation.

All three plans favored by Council members endorse the practice of partial advance funding. The MB approach would increase the taxation of Social Security benefits, extend coverage to more State and local government workers, cut benefits by extending the computation period (or alternatively, slightly increase the contribution rate), slow down future benefit increases from changes in the CPI, and redirect some revenues now going to the HI Trust Fund to build up OASDI reserves. The two approaches involving individual accounts would build up reserves both in the central system and in the individual accounts5 . Whatever the approach, the Council feels it is desirable to build up these reserves soon.

Early action should be taken to reform Social Security.

While OASDI revenues are projected to exceed expenditures for another two decades, the Council urges that early action be taken to reduce the actuarial imbalance in the system. Expenditures are expected to rise rapidly once the baby boom generation begins retiring in little more than a decade. The difference between expected expenditures and expected revenues will grow in the future. Actions taken now to trim the growth of benefits, raise taxes, or raise income in other ways could generate additional assets immediately. These additional assets could be invested (whether directly or through individual accounts), and these investments could raise money's worth ratios for younger workers.

The more notice provided to individuals, the more time they would have to replace any reduction in benefits. If the nation delays making any changes in Social Security for another 10 years, it would require a 25 percent reduction in benefits to rebalance the system without tax increases. By that time the baby boom generation will be on the cusp of its own retirement. The sooner the nation adopts changes and puts people on notice, the less disruptive the changes will be.

Moreover, early action also ensures that the widest possible array of options is available to policymakers. For example, any increase in retirement ages must be phased in gradually over time. Prompt action makes it more likely that raising the retirement age will affect a large majority of the baby boom generation and thus such a policy change could contribute significantly to solving the long-term deficit. Finally, because of the demographic bulge created by the baby boom, the longer the delay, the more likely it is that a solution will involve cuts in benefits for retirees or near-retirees, who are less able to change their work and saving plans to provide for their own retirement.

The new Advisory Board should work with the Social Security Administration (SSA) and Congress to manage Disability Insurance (DI) costs.

As is common practice, the Council considered forecasts for the combined OASDI program. All three approaches for balancing the system over the long run contain implicit or explicit changes in Disability Insurance policies. But beyond that, the Council did not study the DI program carefully, and urges that SSA and Congress be concerned about its increasing cost.

The long-term actuarial balance of Social Security should not be adversely affected solely by the passage of time.

The Council believes that all factors known at the time of the 75-year projections should be considered and reported on by the Trustees. The fact that the 75-year projection period changes as the years pass creates a problem in this regard. Because of the financing scheduled in present law, even as the Trustees are reporting that the system is in balance over the next 75 years, they know that if everything else stays the same, they will not be able to report actuarial balance in, say, another 10 years because demographic trends make the additional 10years more expensive.

The Council therefore urges that any legislative action be designed to assure financial solvency for the 75-year horizon, but that instead of just arriving at actuarial balance for 75 years, in addition assure that the ratio of fund assets to annual expenditures, known as the trust fund ratio, be stable over the final years of the forecast horizon. All the plans discussed in this report try to satisfy this requirement.

Social Security should provide benefits to each generation of workers that bear a reasonable relationship to total taxes paid, plus interest.

Many important values served by a Social Security system are not fully captured by looking solely at money's worth or rates of return. Nevertheless, the Council believes that it is important that young workers perceive that the system is fair. This perception in turn suggests that the younger generation should be well treated in terms of the issue of money's worth, taking into account the fact that within each generation there will be a redistribution toward the lower paid.

In the Council's opinion, this goal cannot be reached unless Social Security, in one way or another, is allowed to receive a greater return on accumulated funds than low-yielding Government bonds. Thus, plans put forth by the Council would, in one way or another, involve investment in private equities of part of the retirement funds. At the same time, the plans differ in their impact on the nation's saving rate, with the two individual accounts options raising national saving, and subsequent national income, more than in the MB plan.

Ways to achieve the higher money's worth ratios differ greatly from one plan to another. In fact, these differences are so great that the Council found no basis for compromise, and indeed these differences, in some cases, are so great as to be much more important than the single similarity of investing in equities.

Any sacrifices in bringing the system into balance should be widely shared and not borne entirely by current and future workers and their employers.

Although reductions in the growth of future benefits and/or increased payroll and income taxes may be involved in bringing the system into long-range balance, the Council believes that present beneficiaries should also share in any sacrifices that might be required. While the details differ, the Council believes that the fairest way to bring these sacrifices about is to apply appropriate income taxation to Social Security benefits. The incomes of low-wage earners would still be protected -- about 30 percent of Social Security beneficiaries would not pay any taxes under any of the Council's proposals -- and the degree of sacrifice involved would depend upon standard income tax principles.

Maintaining full cost-of-living-adjustments (COLAs) throughout the period of benefit receipt is one of Social Security's most important contributions to individual security.

Inflation-proof benefits are generally available only in Federal retirement plans. Most State and local plans offer only partial protection against inflation, and private pension plans usually do not offer automatic adjustments, although they may provide ad hoc benefit increases to recipients from time to time.

In the case of Social Security, the goal of keeping up with wages and prices has been part of the program for decades. Before an explicit provision was added to the law in 1972, the financing of the program allowed for such periodic updating, and this was accomplished, although with some lag. Since the mid- 1970s, benefits have been updated on an annual basis.

Inflation protection makes Social Security benefits increasingly valuable to people as they grow older. The longer people live, the more likely they are to have depleted whatever personal savings they may have accumulated, but expenses continue and, particularly in the case of health care and other services, are likely to increase with age.

The Consumer Price Index (CPI)is the basis for calculating inflation adjustment rates. Because the CPI is so important to Social Security, the Council urges that the market basket on which the CPI depends be updated much more frequently than at present. The Council also supports current efforts to remove bias from the CPI and believes that the Social Security cost-of-living adjustment (COLA) should follow the changes in the CPI made by the Bureau of Labor Statistics (BLS) wherever they may lead. The Council, however, does not support changes in the COLA motivated by political considerations. No matter how well-intentioned or even how accurate such changes might later prove to be, the Council believes it would be a bad precedent. Changes should be made only as a result of careful expert consideration by the government agency charged with that task.

To evaluate the financial effects of different policy approaches favored by various members, the Council has assumed that there will be a downward adjustment in the inflation assumption due to a change in the way the CPI is calculated. For these comparative purposes, all of the plans were analyzed assuming a downward adjustment in inflation rates of 0.21 percent per year, announced by the BLS in March, 1996. In addition, it was assumed that all real variables in the forecast (such as real wage growth and real interest rates) are 0.21 percent per year higher than was assumed by the Trustees. Were these same assumptions to have been adopted by the Trustees, the long-term actuarial deficit would be lowered by 0.31percent of taxable payroll, or by 14percent.

Conventional means-testing of Social Security is unwise.

The amount of Social Security benefits depends on lifetime earnings -- a key component of lifetime income. Whether one is eligible at all for Social Security depends on having had earnings in covered employment, or on being a dependent family member of someone who has. The payment of benefits is to cover part of the likely drop in family income when a wage earner is no longer working due to total and long-continued disability or the death of the wage earner. Thus, in determining a person's benefit, Social Security already takes into account the "means" people have.

The Council rejects the further proposition that Social Security should also condition benefits on assets or other income at retirement -- conventionally known as "means-testing." The fact that benefits are paid without regard to a beneficiary's current income and assets is the crucial principle that allows -- in fact encourages -- people to add savings to their Social Security benefits and makes it feasible for employers and employees to establish supplementary pension plans.

Moreover, means-testing would send the wrong signal to young people and wage earners generally. The message would be: "If you are a saver and build up income to supplement Social Security, you will be penalized by having your Social Security benefits reduced." This message is both unfair to those who work and save and creates the wrong incentives.

Social Security should be financed by taxes on worker's earnings, along with taxes paid by employers, earmarked taxes on benefits, and interest earnings on accumulated reserves, without other payments from the general revenue of the Treasury.

Many foreign systems have contributions from general revenues to their social security systems, either to pay for administrative costs or for part of the benefits. The Council recommends against that procedure6 .

The method of financing Social Security entirely by dedicated taxes has given the system considerable protection from having to compete against other programs in the general budget. The fiscal discipline in Social Security arises from the need to ensure that income earmarked for Social Security is sufficient to meet the entire cost of the program, both in the short run and long run, rather than from competition with other programs in the general budget.

Unless a program is especially protected as, for instance, by dedicated taxes, the Federal budget results are inevitably determined by competition in allocating spending during the budget cycle, depending on the revenue generally available. Social Security, on the other hand, is a very long-range program-- people pay dedicated taxes today toward benefits that may not be received for 30 or 40 years -- and should not be part of an annual budgetary allocation process. There would be less security in a retirement system that changed benefits -- those being paid now or those payable many years hence -- because of short-term budgetary considerations. Social Security's obligations and contributions can certainly be changed and have been. But the process of doing so requires a long lead time and consistency with internal principles of the program.

Benefits for low-wage retirees should be protected.

The Council is mindful of numerous statistical data that report very high rates of poverty among certain groups of older beneficiaries -- rates that are much higher than in other developed countries. To the extent that projected future benefit growth needs to be reduced under any of the proposed plans, the Council prefers to protect benefits for low-wage workers. Council members differ on the protection that ought to be afforded higher wage workers from the OASDI Trust Fund itself.

Any reductions in the growth in benefits should be made in a way that avoids real reductions between one cohort and the next, or "notches."

Previous reductions in the growth of Social Security benefits have sometimes involved situations where younger retirees with the same relative earnings records as older retirees received lower real benefits. These real benefit reductions, or notches, have been perceived to be inequitable. The Council urges that benefit changes be made in a way that avoids real reductions in monthly benefits for similarly situated workers who retire at later dates.

The structure of family benefits in Social Security should be revised.

Statistical studies suggest that it costs retired survivors about three-fourths as much to live as it takes retired couples. On the basis of this information and because of the extent of absolute poverty among elderly widows, many members of the Council believe that the benefit structure for spouses and widow(er)s should be revised. Dependent spouse benefits (benefits payable to a wife or husband when both spouses are alive) should be gradually reduced, so that they are either a smaller fraction of the benefit paid to the principal earning spouse or a flat dollar amount. Survivors' (widow or widower) benefits should be increased as a fraction of the benefits the couple received before one spouse died. The Council recommends gradually increasing Social Security's survivor benefits to the higher of the decedent's benefit (that is, the present-law survivor's benefit), or 75 percent of the combined benefit that the survivor and decedent spouse were receiving when both were alive. These changes would be gradually phased in beginning around the year 2000. Reducing spouse's benefits from 50 percent to 33 percent of the worker's basic benefit and guaranteeing aged survivors 75 percent of the couple's combined benefit would cost 0.15percent of taxable payroll, adding 7 percent to the long-term actuarial deficit7 .

The long-standing goal of universal coverage under Social Security can be obtained by extending mandatory coverage to all State and local government employees hired after 1997.

To the extent feasible, everyone who works for pay should be covered by the Social Security program. Every occupational group contains substantial numbers of people who at one time or another will need the protection of the program. Over the course of a lifetime, it is impossible to foresee who will and who will not need this coverage. Moreover, all Americans have an obligation to participate, since an effective Social Security program helps to reduce public costs for relief and assistance which in turn, means lower general taxes. There is an element of unfairness in a situation where practically all contribute to Social Security, while a few benefit both directly and indirectly, but are excused from contributing to the program.

Coverage has been expanded during much of the 60-year life of the program. Those easiest to cover administratively -- workers in business and industry -- were the only persons covered in the initial Social Security Act of 1935. Over time, the program's coverage has grown to include the self-employed, nonprofit groups, agricultural and household workers, the Armed Services, the Congress, and all other newly-hired Federal employees hired after 1983.

A high proportion of State and local government employees are already covered as a result of Federal/State compacts containing complicated voluntary group provisions. But there is still a lack of complete State and local coverage. It results from an earlier belief that Federal compulsion of the States to participate might not be Constitutional. In the light of several Supreme Court decisions dealing with Federal/State relationships in the area of labor law, it is now generally thought that there is no Constitutional barrier to compulsory coverage.

Basically the issue of covering the last sizable group of workers not under Social Security is an issue of fairness. Newly-hired State and local government employees, like everyone else, should be part of this important national program. The workers would generally gain from being under Social Security's inflation-proof provisions and from coverage that could be combined with that from work outside of their State and local government employment. Moreover, spouses would benefit from Social Security's survivors protection which is not always provided under State government pension plans.

For Social Security, this extension of coverage saves money over both the short- and long-term. Partly this is because contributions will be received for many years before the benefit payments must be made, partly because under present law a high proportion of these State and local government employees will get Social Security benefits anyway. Including all State and local government workers hired after 1997 saves 0.22percent of taxable payroll, 10 percent of the present long-term actuarial deficit8 .

There should be improved incentives for people to extend their working careers.

One natural solution to the future problem of increasing numbers of retirees per worker is to improve the incentives for workers to extend their working careers. Present law already includes measures to bring Social Security's delayed retirement credit to the full actuarial equivalent so that the same expected amount of benefits will be paid no matter when the worker retires between ages 62 and 70. This change eliminates a disincentive that has until now discouraged workers from extending their careers. Congress and the Administration have recently passed a measure to raise the exempt amount under the retirement earnings test for beneficiaries who have attained the age of eligibility for full benefits, another increased incentive to extend working careers. The Council favors extending the benefit computation period so that those who work steadily through age 67 get higher benefits than those who do not (see below). Those plans featuring individual accounts also increase retirement work incentives as compared to the present system, because contributions to the individual accounts continue as long as workers work.

The period over which the indexed average wage is computed should be extended to 38 years.

Average indexed earnings, on which Social Security retirement benefits are now based, are computed over an averaging period of the 35 years of highest earnings. With the age of eligibility for full retirement benefits rising to 67 under present law, and as a means to increase retirement work incentives, it seems reasonable and consistent to increase the working lifetime period to 38 years for purposes of calculating the average earnings on which retirement benefits are based. This change would be phased in from 1997 to 19999 .

This change would result in a slight reduction in projected future benefits for those retiring after 1999; those with relatively short work histories being the most affected. But it tightens the relationship of lifetime contributions to benefits, increases incentives to extend working careers, and improves the overall equity of the program. The change reduces benefits an average of 3percent and saves 0.28 percent of taxable payroll, 13percent of the long-term deficit10 .

The income taxation of Social Security benefits should be revised.

Many Council members feel that the fairest way to ask present retirees to share in the cost of bringing Social Security into balance is by revising the taxation of Social Security benefits. In 1993, when the income tax provisions applying to Social Security beneficiaries were broadened, the goal was to tax Social Security recipients in a way similar to the recipients of other contributory defined benefit pension plans -- that is, including as taxable income recipients' benefits to the extent that they exceed what workers had paid in. However, Congress chose a proxy of 85 percent of benefits as representing amounts not attributable to employee contributions rather than having taxes on benefits computed individually. The taxation of 85 percent of the benefit was designed to protect workers from having to pay more than they would have under an individual computation, but in many cases it was recognized that individuals would be required to pay less than would be the case in an individual computation. Since the administrative difficulty of individual calculation is manageable, taxes can be computed on an individual basis as is the case with others who receive contributory defined benefit pensions.

There are three other aspects of the present taxation of Social Security benefits that should be recognized to understand the various benefit taxation options. One is that in 1983 the policy was established of having the revenue from the taxation of Social Security benefits dedicated to the OASDI Trust Funds. The second is that the additional revenue from the 1993 introduction of the 85 percent rule goes to the Hospital Insurance (HI) program under Medicare rather than to OASDI. The third is that the tax on Social Security benefits does not apply to those who have adjusted gross incomes below $25,000 if single and $32,000 for couples.

All three plans proposed by various Council members would drop this last provision by phasing out the low-income thresholds by year 2007. Because of the general protections the personal income tax law provide for low-income households, 30percent of Social Security beneficiaries still would not have to pay taxes on their benefits. This change would save 0.16 percent of taxable payroll in the present Social Security system, 7percent of the long-term actuarial deficit.

Both the MB plan and the IA plan would, in addition, replace the 85percent rule with a scheme whereby benefit taxes are computed individual-by-individual. This change would save another 0.15 percent of taxable payroll, another 7percent of the long-term actuarial deficit. Then the MB plan would, in addition, shift revenues now going to HI to OASDI bwtween 2010 - 2020, saving another 0.31percent of taxable payroll, another 14 percent of the long-term actuarial deficit. The PSA plan, on the other hand, would move to a system of full taxation of employer-paid benefits (paid for with tax deductible contributions) and no taxation of employee-paid benefits (paid for with already taxable contributions), implying an immediate elimination of the taxes that now go to the HI program.

The change in the age of eligibility for full retirement benefits from age66 to age67 should be accelerated, and this age should later rise in line with overall longevity.

Under current law, the age of eligibility for full retirement benefits (currently age65) will gradually increase to age66 for workers who attain age62 in 2005. It will then remain at age 66 for 11 more years and then gradually increase to 67 for workers who attain age62 in 2022 or later. A majority of Council members favor speeding up this schedule, so that the age of eligibility for full retirement benefits increases to 67 by the year2011. This change would save 0.10percent of taxable payroll, 5percent of the long-term actuarial deficit. After 2011 these same members favor a more gradual rise in the age of eligibility in line with longevity trends (at the rate of about 1 month every 2years). Making this additional change would save another 0.40percent of taxable payroll, 18percent of the long-term actuarial deficit11 .


The Social Security Administration (SSA) should enhance its research and analysis capabilities with greater in-house resources and outside consultant services.

The Council has become aware that the resources available within SSA to provide research and analysis are not sufficient to address adequately the problems facing the program over both the short- and long-term. Sophisticated research and analysis must be more readily available to policymakers to develop sensible policy changes.

The Council believes that the research and analytical capacity of the agency, particularly the Office of the Actuary and the Office of Research, Evaluation, and Statistics, should be upgraded. This upgrading could be accomplished through additional in-house resources, enhancements in staffing, and with greater use of outside consultant and contractual research services.

Stochastic simulation modeling should be used as a tool for recognizing explicitly the uncertainty surrounding the Trustees' demographic and economic assumptions.

In addition to the projections now being made by the Office of the Actuary, the Council believes that the projections for the 75-year horizon should also be made through the use of stochastic simulations. This technique employs probability distributions rather than specific point estimates in forming future scenarios and assessing their likely probability. It permits policy analyses to be conducted in a way that more realistically incorporates uncertainty into its measures of long-term financial viability.

Private researchers should be granted greater access to agency data.

The Council believes that the value of research and analysis of Social Security data will be enhanced by providing private researchers with greater access to these data as well as to the economic and actuarial models used in forecasting and analysis. Access will facilitate collaboration of outside research with SSA research for an exchange of ideas, analysis, and findings.

A technical panel of experts should be convened periodically (say every five years) to conduct a comprehensive review of the assumptions and methodology used in evaluating the financial status of the OASDI Trust Funds.

The Advisory Council appointed a technical panel of experts to review the assumptions and methods used in evaluating the financing of the OASDI Trust Funds. The Council believes that it is highly desirable to continue this periodic appointment of outside actuaries, demographers, economists, and other experts to a panel for such a review.


The Council supports the creation of inflation-indexed bonds by the Treasury.

The Council supports the Treasury in its recently expressed intention of offering inflation-indexed bonds. The Council believes these bonds would make it easier to plan for retirement by translating lump sum accumulations into inflation-protected retirement income flows. Indexed bonds would also make it easier for private pension plans and private insurers to offer indexed life annuities.

Government and employers should provide workers with adequate and realistic retirement income information.

The Council supports the initiative of the Departments of Labor and Treasury to raise each individual's awareness of how much needs to be saved for an adequate retirement living standard (beyond what is available from Social Security). SSA should also continue to provide workers with realistic information for their retirement planning through its "Personal Earnings and Benefit Estimate Statements." The Council supports proposals to enable employers to provide more retirement planning assistance, including information on investment allocation and returns.

Information is also needed on how to save and how to allocate retirement savings distributions. Enhanced education on the advantages of maintaining retirement savings distributions in tax-free accounts would help counter the tendency of individuals to spend too much out of their retirement savings.

Retirement plans should be encouraged to cover a larger portion of the labor force and to meet more of the need for increased retirement savings.

The Council believes that the elaborate retirement plan rules that have been added since the early 1980s should be streamlined and revised to encourage the growth of private pension plans. The Council supports recently enacted legislation that permits small businesses to adopt simplified pension plans and creates "safe-harbor" nondiscrimination rules for all businesses maintaining a salary reduction plan. Redundant and overlapping rules should be simplified or eliminated. Policy should be aimed at providing consistent contribution and benefit limits for similar types of retirement savings plans.

"Full funding" of private defined benefit pension plans should be permitted.

The Council feels that changes in law since 1987 that have unduly limited the funding of the obligations of defined benefit plans are unwise. The Council urges that employers be allowed to fund fully their ongoing qualified benefit obligations calculated with reasonable economic assumptions and some margin for risk. Projected benefit obligations should include anticipated pay increase effects in both salary-related and flat-dollar defined benefits.

Rules that would encourage or require workers to maintain their retirement saving in tax-free accounts until retirement should be promoted.

The Council supports efforts to maintain retirement savings in tax-free accounts after employment ceases, especially prior to early retirement age, and expresses concern about the growing trend of converting lifetime benefits to lump sum distributions. The Council supports changes that would eliminate access to account balances attributable to employer contributions during employment with that same employer ("in-service" hardship withdrawals and loans of amounts attributable to employer contributions). While the Council recognizes that increased restrictions on withdrawal or loans of employee-made contributions could discourage participation in some plans, it favors such restrictions.

Certain provisions of the Employee Retirement Income Security Act (ERISA) of 1994 should apply to State and local government pension plans.

The Council believes that the provisions of Title I of ERISA governing disclosure, fiduciary standards, and spousal protection (pre- and post-retirement survivor annuities and qualified domestic relation orders) should be applied to State and local government plans to protect the benefits of the workers and their survivors. The Council does not recommend extending the minimum funding standards to State and local governments at this time, particularly in light of the effects of its recommendations to cover all new State and local government hires under Social Security.


One group of Council members favors a maintenance of benefits (MB) plan12 . It would maintain the present Social Security benefit and tax structure essentially as is, though with an extension of the benefit computation period, or alternatively a small increase in the contribution rate, and coverage of newly hired State and local government employees. The plan would gain revenue by more complete Federal income taxation of current Social Security benefits and by a redirection of some taxes on OASDI benefits now going to the HI Trust Fund phased in between 2010 and 2019. The goal of eliminating "drifting out of balance because of the passage of time" would be reached in this plan by a 1.6 percent combined employer-employee payroll tax increase in 2045. It envisions, after a period of study and evaluation, the possibility of a large-scale investment of OASDI Trust Fund monies in the equity market in order to help bring the program into balance and to greatly improve the money's worth ratios for young workers and future generations. The analyses in the report assumes that this change would be made within a few years.

The important specific provisions are as follows:

All Social Security benefits in excess of already taxed employee contributions would be included in Federal taxable income and the proceeds deposited in the OASDI Trust Funds. This includes revenue now going to the HI Trust Fund, which would be redirected to OASDI phased in between 2010 and 2019 as Medicare is refinanced for the long run.

All State and local government employees hired after 1997 would be covered under Social Security.13

The benefit computation period would be extended from 35 to 38years phased in over the 1997-1999 period, reducing benefits an average of 3 percent, or alternatively, contribution rates would be increased in 1998 by 0.15 percent of covered wages for employees, matched by the employer.

Except for the possible alternative of a small 1998 increase, deductions from workers' earnings and the matching contributions of employers would not be increased for the next 50 years at which time, in order to meet the new actuarial test of a stable trust fund ratio at the end of the 75-year estimating period, the combined employer-employee payroll tax rate would be increased by 1.6percentage points.

The supporters of the MB plan have urged the adoption within the next year or two of changes that would reduce the 2.17percent of payroll deficit to 0.80 percent of payroll, thus postponing the estimated trust fund exhaustion date from 2030 to 2050. And they have urged for further study and examination a plan that would eliminate the remaining 0.80 percent of payroll deficit by investing a portion of future trust fund accumulations in stocks of private companies indexed to the broad market.

As stated in the section on agreed-upon "findings, principles, and recommendations," the Council favors the movement from pay-as-you-go financing to partial advance funding of Social Security. When the size of the fund was limited to a contingency reserve -- defined as a reserve equal to 100percent to 150percent of the next year's outgo -- earnings on investments made little difference to long-range financing. Now with the trust fund ratio soon moving beyond 150 percent (the fund is currently at about $500billion, 140 percent of next year's outgo), the return on investment can make a major difference in long-range financing. The plan, therefore, proposes to change investment policy from exclusive use of special Government issues with a yield equal to the average on all outstanding long-term debt of the United States (projected to average 2.3percent in real terms over the next 75 years). Under the plan, Social Security, like other public and private defined benefit pension plans, would invest a sizable portion of the growing fund in private equities.

Historically, the differential between the real return on Government bonds and the real return on stocks has been about 4.7percentage points per year. Under this proposal the major part of the OASDI Trust Funds would still be in Government bonds drawing a projected 2.3percent real return, and up to 40percent of the funds would be invested in stocks projected to draw 7 percent real return. This means that under the plan the total real return on the total fund would eventually build up to 4.2 percent. Consideration should also be given to investments in corporate bonds and instrumentalities of the United States such as Fannie Mae, which would raise the return somewhat further.

The eventual 40percent stock investment share is arbitrary. It is conservative by private pension standards -- these private pensions typically invest over 40 and up to 60 percent of their funds in equities. It is also a rather conservative attempt to capture directly for Social Security some of the economic benefit to the nation from investing the Social Security surpluses in Government bonds. When an additional dollar of Social Security reserves is invested in Government obligations (and assuming no change in taxes or spending in the rest of the budget), it absorbs an additional dollar of Federal borrowing, thereby releasing an additional dollar of private saving for private investment. Thus, it could be argued that the proper credit for the economic contribution of building the Social Security fund would be the return on all physical capital, about 6percent net of corporate taxes, rather than the 4.2percent sought by this plan.

Yet it seems wise to adopt a policy of seeking a conservative return for Social Security in this plan since it would be a considerable departure from current practice. In any case the 40 percent allocation to equities is not a magic number -- the share could be 35percent or 45percent.

In evaluating this proposal, the Council had outside help in stochastic modeling of the degree of financial risk involved as the amount put in private equities is increased. (See "Presentations to the Council," Volume II of the Council's report, pp. 341-345.) When considered over a long period of time, the model showed only a slight increase in financial risk at this level of equity investment. As a matter of financial theory, the diversification achieved by investing in both stocks and government bonds should also reduce portfolio risk for the OASDI Trust Fund. Under the MB plan as presented, investment in equities would begin in the year2000 and reach 40percent of the accumulations by 2014.

The trust fund investment in equities would be overseen by an investment policy board nominated by the President and confirmed by the Senate. The investment policy board would be subject to legislated fiduciary standards mandating that trust fund investment policy is intended solely for the economic benefit of Social Security participants, and not for other economic, social, or political objectives. The investment policy board would have two major responsibilities: (1) selecting among alternative passive market indexes, and (2) conducting a competitive bidding process to select the equity index portfolio managers. The investment policy board would be an expert panel charged with selecting among passive market indexes such as the Standard and Poor's Indexes, the Wilshire Indexes, the Russell Indexes, and various other U.S. and global market passive indexes that may be developed in the future. Selection of the portfolio manager or managers would be competitively bid among leading equity index managers serving large institutional accounts, with the objective of securing the highest level of technical expertise and relevant portfolio management experience at the lowest possible cost. The investment policy board would monitor portfolio and investment manager performance, and consider changes from time to time in the passive index or portfolio managers as appropriate.

It would also be important to determine the best way to neutralize the effect of Social Security holdings on stockholder voting on company policy. Perhaps just barring the voting of Social Security-held stocks by law would be enough. Or it might be desirable for voting of Social Security stocks to automatically be scored in the same proportion as other stockholder votes. Or if some votes on important policy such as changing management requires more than a majority for approval, perhaps the computation of the base to which the proportional vote is applied should be computed without counting stock held by Social Security. In one way or another, the neutrality of Social Security's voting rights should be established.

It is important not to confuse this proposal to invest part of Social Security funds in stocks indexed to the market with the various individual account proposals from other Council members. The sole purpose of Social Security's investment in equities is to secure a higher return than can be obtained from the present practice of investing all funds in Government bonds. Otherwise, the Social Security structure would be little changed. Social Security would remain a defined benefit plan with the amount of benefits and the conditions under which they are paid, and the definition of who pays how much, continuing to be a matter of Federal law. The program would continue to be administered by the Federal Government. The individual account proposals, by definition, establish compulsory savings plans with the individual investing the savings and in retirement getting back whatever the investment yields. In other words, individual accounts represent a partial shift from a defined benefit plan to a defined contribution plan like a 401(k) or an Individual Retirement Account (IRA). The MB Plan remains a defined-benefit plan.


Another group of Council members favors an individual accounts (IA) plan. 14 The goal is to preserve the social adequacy protections in the present Social Security benefit system while still raising overall national retirement saving. The social adequacy protections are largely preserved, though the growth of Social Security benefits would be gradually reduced. Part of this reduction in the growth of benefits applies to workers at all earnings levels resulting from an acceleration in the increase in the age of eligibility for full retirement benefits up to year 2011, followed by a slower increase in line with overall longevity, and by lengthening the computation period from 35 to 38 years. Part of the reduction in growth is focused primarily on middle- and high-wage workers because it comes from an adjustment in the benefit schedule, though with some reductions in spouse's benefits and improvements in survivor's protection. There would be a mandatory additional contribution of 1.6 percent of covered payroll that would be held by the Government as defined-contribution individual accounts. Individuals would have constrained investment choices on how these funds were to be invested -- ranging from a portfolio consisting entirely of bond index funds to equity index funds. The accumulations from the individual accounts would be converted by the Government to single or joint minimum guarantee indexed annuities when individuals elect retirement, which could happen any time after the present age of earliest eligibility for Social Security retirement benefits (age 62). The combination of the reduced growth in benefits, the increased age of eligibility for full retirement benefits, and the proceeds of the individual accounts would leave total benefits on average at about the levels of present law for all income groups.15

The important specific provisions are as follows:

Mandatory defined contribution individual accounts in the amount of 1.6percent of covered payroll would be created.

Defined contribution individual accounts in the amount of 1.6 percent of covered payroll would be created and funded by employee contributions. Individuals would have constrained choices on how the funds were to be invested. These individual accounts would be clearly excluded from the Federal budget, though the budgetary treatment of the rest of Social Security would remain unchanged.16

The accumulated funds would be converted to single or joint minimum guarantee indexed annuities when the individual elects retirement, any time after the age of earliest eligibility for benefits (now age62). 17 The minimum guarantee provision would assure that some portion of the purchase price of the annuity--say, an amount equal to 1 year's worth of the annuity--would be payable in all cases. Thus, even if a worker who had elected a single annuity died after receiving only one annuity payment, an additional sum would be paid to the survivors. As is the case with other pension plans, a married worker would have a choice (with the consent of the spouse) on whether a single or "joint and survivor" annuity was chosen. (The "joint and survivor" option would provide a lower basic annuity while the worker was alive, but would continue to pay a portion of the annuity to the survivor after the worker's death.)

The gradual increase in the age of eligibility for full retirement benefits would be accelerated and extended.

The increase in the age of eligibility for full retirement benefits from 65 to 67 would be accelerated. After year 2011, when this age of eligibility reaches 67, it would rise slowly along with overall longevity. At present rates, this age would rise about 1month every 2years, meaning that it would reach age70 by year 2083.18

The growth of basic benefits would be slowed, mainly for middle- and high-wage workers.

The 32 and 15percent conversion factors in the present benefit schedule would be gradually lowered over time to 22.4 and 10.5percent, respectively. 19 Benefit changes would be phased in to avoid notches, or drops in real benefit levels, while keeping the basic OASDI tax rate at 12.4 percent of taxable payroll. The combination of all changes in benefits and the individual accounts would on average keep full benefits (including the proceeds of the individual accounts) at roughly present levels for all income classes of workers.

Survivors protection for two-earner couples would be increased and dependent spouse benefits would be lowered.

A surviving spouse's benefit would be determined as the highest of (1) his or her own basic benefit, (2) the deceased spouse's basic benefit, or (3) 75 percent of the couple's combined basic benefits. This change would increase benefits for survivors of two-earner couples. The dependent spouse benefit would be gradually phased down from 50percent to 33 percent of the worker's benefit. If the worker died before becoming eligible for retirement benefits, the survivor would have to wait to receive 75percent of the combined benefit until the deceased would have become eligible.

Regular Social Security benefits would be taxed under income tax principles and deposited in the OASDI Trust Funds.

All Social Security benefits in excess of already taxed employee contributions would be included in Federal taxable income. The income thresholds would be gradually phased out. Both provisions are the same as those in the MB reform plan. But unlike that plan, there would be no redirection of taxes on Social Security benefits from the HI Trust Fund to the OASDI Trust Funds.

All State and local government employees hired after 1997 would be covered under Social Security.

The benefit computation period would be extended from 35 to 38years, phased in over the 1997-1999 period.

Proceeds from the defined contribution individual accounts would be taxed under consumption tax principles.

The individual accounts would not be included in the Federal budget. To conform their tax treatment with the tax treatment of other defined contribution pension saving, the individual accounts could be taxed in either of two ways:

1. They could be made tax-deductible when saved and taxable when the benefits were paid. This tax treatment could involve an immediate revenue loss, though it would be mitigated if the accounts reduce other tax-preferred saving.

2. Conversely, they could be made taxable when saved and deductible when received. Because the savings generating the individual accounts would still only be taxed once, this tax treatment would have about the same net effect in present value terms as the deferred tax treatment that is now received by most other defined contribution pension saving. This tax treatment could involve ultimate Federal revenue loss, if the individual accounts reduced other tax-preferred saving.


A third group of Council members favors reforming Social Security to move toward a system of relatively large individual accounts, making a substantial portion of the new system fully funded. 20 When fully phased in, a two-tiered system would take the place of the present Social Security system. The first tier would provide a flat retirement benefit for full-career workers and the second tier would provide fully-funded, individually-owned, defined contribution retirement accounts, referred to as Personal Security Accounts (PSAs). In contrast to the IA plan, the funds in these accounts would not be held or managed by the Federal Government; the investment options would be less restricted; and workers would not be required to annuitize their accumulations at retirement. Also, in contrast to the IA plan, the PSAs would be funded with 5 percent of the current payroll tax. The new system would be implemented January1, 1998, and be fully effective for workers under the age of 25. Survivors and disability insurance benefits would be modified but continue to be financed by the OASDI Trust Funds, and administered through SSA. Under the basic actuarial forecasts, the combined benefits of the PSA plan would on average generally exceed those of the present system for most workers of different income classes.

To bring the current system back into financial balance while allowing for the new individual accounts, this plan changes the way basic (tier I) retirement benefits are calculated. These changes include covering newly hired State and local workers, raising the age of eligibility for full retirement benefits, and moving toward a system of flat benefits for full-career workers. These and other changes would be phased in very gradually. When fully effective, tier I retirement benefits, spouses benefits, and survivors and disability benefits would be financed by the 7.4percent portion of payroll tax not used to fund PSAs.

As with any reform that shifts from pay-as-you-go financing to more complete advance funding, there is a cost of transition to the new system. This is because, in moving toward a system in which each generation pays for its own benefits rather than those of its elders, some workers have to help pay for the benefits that come due under the old system to earlier generations as well to prefund their own retirement. While transition costs can be met in a number of ways, the plan assumes a 72-year payroll tax increase of 1.52percent, supplemented by added Federal borrowing.

The important specific provisions are as follows:

Tier II: Personal Security Accounts (PSAs) would be created.

This plan would create PSAs dedicated to retirement saving and financed by reallocating 5percentage points of the employee's share of the current OASDI tax rate (now 12.4 percent, employee and employer share combined, excluding the HI portion of the tax). This is approximately half of the tax now used to finance retirement benefits under Social Security. PSAs would be individually owned and privately managed, subject to necessary regulatory restrictions to make sure they were invested in financial instruments widely available in financial markets and that they were held for retirement purposes. Individuals could begin withdrawing funds from their accounts at age 62, and any funds remaining in their accounts at death could be included in their estates. Every worker under age55 in 1998 would participate in the 5percent payroll tax reallocation and would receive tier II benefits based on their accumulations plus interest.

The new benefit schedule would be phased-in.

Current retirees and workers age 55 and older in 1998 would be covered under the existing Social Security system, subject only to changes in retirement ages, disability benefits, and benefit taxation that apply generally under the proposal. Workers under age25 would receive benefits only under the new system, with the tierI benefit provided by the Government and the tier II benefit based on accumulations in their PSAs. Workers age25 to 54 in 1998 would receive a tier I benefit that was a combination of their accrued benefit under the existing system and the benefit they accrued under the new system (with the amounts depending on the proportion of their work life in the two systems), as well as their tierII benefit.

Tier I benefits would provide a floor of support.

The first tier benefit would be calculated as follows: for workers under age 25 in 1998, those with full careers under Social Security (those with 35 or more years of covered employment) would on, retirement, receive a flat dollar benefit equal to $410 monthly in 1996 (equivalent to 65 percent of the current poverty level for an elderly person living alone or 76percent of the benefit payable to a low-wage worker retiring in 1996). This benefit would be wage-indexed from 1998 until the year the worker becomes eligible to retire, and, as under the existing law, price-indexed thereafter. For individuals who do not work a full career, half of the flat benefit would be earned with 10 years of covered earnings, with a 2-percent increment for each additional year of work up to 25years.

Workers age 25 to 54 in 1998 would receive their accrued benefit under the existing system, plus a prorated share of the flat benefit provided under the new system. The accrued benefit would be calculated under present law as of January 1, 1998,21 and communicated to the worker as soon as practicable after that date; it would be wage-indexed until the year the worker becomes eligible to retire. The flat benefit would be prorated to reflect the proportion of potential workyears (out of 40yea rs between ages22 and 61) that occur under the new system. The composite benefit would be price-indexed after eligibility.

For all workers under age55 in 1998, tierI benefits would be supplemented by the proceeds of the personal security accounts.

The gradual increase in the age of eligibility for full retirement benefits would be accelerated and extended.

The increase in the age of eligibility for full retirement benefits from 65 to 67 would be accelerated (using the same schedule as the IA plan). After the year 2011 this age of eligibility would increase to keep pace with increases in life expectancy among the elderly. Under the assumptions used in the 1995 Trustees' Report, this age of eligibility for full retirement benefits would rise by about 1 month every 2years, attaining 68 for workers reaching age62 in the year2035, attaining 69 in the year2059, and 70 in the year2083. To facilitate individuals' retirement income planning, the schedule for increasing the age of eligibility for full retirement benefits would be established in the law as of the date of implementation; however this schedule would be subject to review every 10 years by the Social Security Trustees to ensure that the age increases were having their intended effect, which would be to stabilize the ratio of potential retirement years to potential work years in future decades.22

The age of eligibility for early retirement benefits would increase in tandem with the age of eligibility for full benefits, with the eligibility age for early retirement reaching 65 in the year 2035 and then staying at that level.23

The retirement earnings test would be eliminated.

Over the 1998-2002 period, the retirement earnings test, which reduces benefits on account of earnings above an exempt amount, would be eliminated at the age of eligibility for full retirement benefits, both for benefits computed under existing law and under the proposal (i.e., tierI benefit and past service credits). No earnings test would apply to withdrawals from tierII personal accounts.

OASDI benefits for spouses, survivors, and disabled workers would be altered.

Spouses would receive the higher of the retirement benefit they would be entitled to receive based on their own earnings' history, or one-half of the retired-worker benefit of their spouse, or one-half of the full flat benefit when the system is fully phased in. In addition, they would be entitled to any accumulations in their own PSA based on contributions during their working years.

As under the IA plan, aged surviving spouses would be entitled to 75percent of the OASDI benefit that would have been payable to the couple if both spouses were alive; this would be phased in over 40 years (1998-2037). In addition they would be eligible to inherit any balances in the PSA of their deceased spouses. For young survivors, benefits would be calculated as under the present law. A surviving spouse would be eligible to inherit any PSA balance of the deceased worker and would have access to this PSA balance when the surviving spouse reached age62.

For workers disabled after January1,1998, benefits would be calculated as under present law except that the percentage of the primary insurance amount payable (now 100 percent) would gradually be reduced as the age of eligibility for full retirement benefits increases, ensuring that disabled-worker benefits do not exceed benefits for age-65 retirees.24 The reduction in disability benefits would be limited to 30percent.

There would be new rules for benefit taxation.

Benefits of workers age55 and older, benefits for work before 1998 for workers between the ages of 25 and 54, benefits for the disabled, and benefits for current retirees: The share of benefits financed by employer contributions, which were tax deductible when made, would be included in taxable income when received. This means that 50 percent of benefits would be taxable, and that the HI Trust Fund would immediately lose any future benefit tax revenues credited to it.

Tier I: Because tierI benefits would be financed by employer contributions, which were deductible when made, all tier I benefits would be taxable. The revenues from benefit taxation would be credited to the OASDI Trust Funds.

Tier II: Because tierII benefits would be financed with (already taxed) employee contributions, the proceeds (including the inside build up) of the PSAs would not be taxable.

All State and local government employees hired after 1997 would be covered.

The transition would be financed.

Transition costs arise because, under the present system, there are large unfunded accrued obligations -- that is, benefits scheduled to be paid to current retirees and to workers who have already paid taxes in excess of assets on hand. Under the plan, these obligations would be met as they mature. At the same time, the new fully-funded component of the system would be implemented. During the phase-in of the new system, the cost of meeting obligations under the existing system is sometimes referred to as the "transition cost."

Transition costs would be met with a combination of added taxes and added Federal borrowing. The SSA actuaries project that a 1.52 percent supplement to the payroll tax would cover average long-range transition costs over the next 72years. However, because the unfunded accrued obligations under the existing system are highest in the next couple of decades and taper off in later decades, there is a shortfall of revenues between about 2000 and 2034 and an excess of revenues thereafter. It is assumed that the shortfall would be met by issuing bonds to the public for the next 40years (totaling an estimated $1.9trillion in 2034, in 1995 dollars), and that these bonds would be fully repaid by the excess of tax revenues in the later period.

The proponents of this plan would prefer not to use a payroll tax to finance the transition and would prefer to use a broad-based consumption tax. However, setting up the administrative apparatus to collect such a relatively small tax, while leaving the current income tax in place, may not make economic sense. Should the U.S. tax system move toward a consumption base, these Council members would regard the consumption tax as a highly preferable means of helping to finance the transition. One of the advantages of using a consumption tax to finance the transition is that current retirees (who have received relatively generous returns on their contributions) would share in the transition cost. This would not be the case if the transition is financed solely by workers through a payroll tax.

Despite the apparent similarity between this 1.52percent transition tax and the 1.6 percent increase in the payroll tax in the IA plan, it is important to note that these two taxes are qualitatively different. In the IA plan, a payroll contribution is mandated in the law and it determines the full contribution to defined contribution individual accounts managed by the Federal Government. Under the PSA funding arrangement described here, the "supplemental tax" is the projected cost rate of transiting to a system in which half of the retirement program is fully funded through privately-held individual accounts. The actual cost rate would depend on national income in future years (which is assumed to be unaffected by the various reform options).

1 Beginning in 2012, the trust funds will pay out more in annual benefits than they collec in payroll and income taxes. Interest on the trust fund investment in government securities will keep the balance positive until 2020.

2 Appendix I makes this point in another way. It gives a complete analysis of how the actuarial balance has changed since the 1983 Social Security reforms, including the impact of adding the years from 2060 to 2070 to the forecast horizon.

3 There would, of course, still be stock market risk borne by the whole OASDI system.

4All members of the Council support the Findings and Recommendations section of this report except where explicitly stated otherwise, but six members take exception to starting the report this way.

Members Robert M. Ball, Edith U. Fierst, Gloria T. Johnson, Thomas W. Jones, George Kourpias, and Gerald M. Shea: "We would begin this report with a statement of the great successes of our Social Security system and the importance of continuing it without fundamental change. We recognize, of course, that such a beginning is not compatible with the proposals of the other two groupings of Council members, but the reader should know that we have included it as the Introduction to our statement titled "Social Security for the 21st Century."

We also find this Overview almost deceptively agreeable. The reader should know right at the beginning that we do not consider the plans described in the report to be the three best plans for obtaining the financing objectives agreed to. We are completely opposed to any plan which substantially reduces current Social Security protection as defined by law and substitutes in part a compulsory savings plan with that part of the protection dependent on earnings from individual investment. Thus, we strongly oppose the other two plans in this report and would not select them as a second choice, or a third or fourth choice.

We describe later why we are just about totally opposed to the statement titled "Comparison of Plans" agreed to by the other Council members."

5 In the case of the PSA plan, central system reserves are lower, but demands on these reserves are also greatly reduced.

6While all Council members support the principle of dedicated revenue sources, three members do not agree that other tax sources or general revenue financing should be categorically excluded.

7 While all Council members approve of this change in principle, a minority of the Council does not believe it is desirable to increase the net cost of the program in this way at this time.

8Three members oppose the inclusion of currently uncovered State and local employees because of the financial burden that would be placed on workers and employers who are already contributing to other public pension systems.

9The number of computation years is determined by counting five less than the number of years elapsing after 1950 (or age 21, if later) and up to age 62. Subtracting 5 years allows for periods of low or no earnings. Several members do not favor this extension.

10This change is not relevant in the PSA plan, which switches to a different means of computing benefits.

11 Several members disagree with this set of changes. See endnote number 7 to the statement titled "Social Security for the 21st Century" endorsed by Robert M. Ball, Gloria T. Johnson, Thomas W. Jones, George Kourpias, and Gerald M. Shea.

12The members favoring this plan are Robert M. Ball, Gloria T. Johnson, Thomas W. Jones, George Kourpias, and Gerald M. Shea. Edith U. Fierst supports most, but not all, provisions of the plan.

13Although three members of the Advisory Council who generally support the Maintain Benefits (MB) plan oppose this proposal, it is supported by the other ten members of the Council.

14 The members supporting this plan are Edward M. Gramlich and Marc M. Twinney.

15 To avoid the budgetary and management complications of central fund equity investment, there would be no such direct equity investment in this plan. See statement by Edward Gramlich and Marc Twinney.

16 While technically Social Security is now off-budget, most budget resolutions and general discussions of the budget amalgamate the Social Security account with that of the rest of the Government, hence effectively including Social Security in the Federal budget.

17 If a worker dies before reaching retirement age, the accumulated funds would be held for the surviving spouse and would be available (in the form of an annuity) when the surviving spouse became eligible for widow(er)'s benefits at age 60. If the worker did not leave a surviving spouse, the funds would go to the worker's estate.

18 Benefits would continue to be available on an actuarially reduced basis for retired workers and spouses at age 62, and for aged surviving spouses at age 60.

19 Under the current benefits formula, three conversion factors are used to compute benefits from average indexed monthly earnings (AIME). For workers becoming eligible for benefits in 1996, benefits equal 90 percent of the first $437 of AIME, plus 32 percent of the AIME from $437 to $2,635, plus 15 percent of the AIME in excess of $2,635.

20 The members favoring this plan are Joan T. Bok, Ann L. Combs, Sylvester J. Schieber, Fidel A. Vargas, and Carolyn L. Weaver.

21 The worker's primary insurance amount would be calculated as if he or she became disabled on January 1, 1998, and would be prorated to reflect the age of the worker in 1998.

22 The age of eligibility for full retirement benefits would be adjusted to keep the ratio of potential retirement years (life expectancy at this age) to potential work years (this age minus age 20) constant.

23 The earliest age of eligibility for aged surviving spouses' benefits would also rise in tandem; benefits would continue to be available for surviving spouses two years earlier than for retired workers.

Some members supporting this plan would have preferred to leave the age for receipt of early retirement benefits at 62 in light of the fact that there are only small long-term gains to the Trust Funds of delaying the age to 65. Those who support the increase believe that it is desirable for workers to delay retirement because of improvements in health and longevity and the potential gains to the economy; also the disability program is specifically geared to meeting the needs of older workers who cannot continue to work due to a disabling condition.

24 Workers would convert from disability benefits to retirement benefits at age 65 (as is now the case). At this time, benefits would be recomputed under the new retired-worker benefit rules. Disabled workers would be eligible to draw on their personal accounts at the time of conversion to retirement benefits.