December 15, 1998

The Feldstein Social Security Plan
by Robert Greenstein, Wendell Primus, and Kilolo Kijakazi

Table of Contents

I.  Introduction and Overview
II.  How the Feldstein Proposal Works
III. Weaknesses of the Plan
IV. Conclusion

Introduction and Overview

Increasing attention is being paid to a Social Security proposal developed by Martin Feldstein, a Harvard economist and former chairman of the Council of Economic Advisers under President Reagan. The Feldstein plan is generally presented as a painless way to restore Social Security solvency without any reductions in the level of benefits the current benefit structure promises, with no tax increases, and, in fact, with almost all retirees receiving greater retirement income than under current law. When something appears too good to be true, it usually is. This adage holds for the Feldstein proposal.

While proponents of the plan claim it would restore long-term Social Security solvency with no pain, analyses of the plan — including analyses by the Congressional Budget Office and the Office of the Chief Actuary at the Social Security Administration — show otherwise. The Social Security system faces insolvency because the system has promised more in benefits than it will collect in revenues. The Feldstein plan promises more in retirement benefits than current law would provide but collects no more revenue to cover the added costs. The Feldstein plan would devote much of the surplus projected in the unified budget to finance the increases the plan would provide in retirement benefits. These surpluses, however, are uncertain in magnitude, are not expected to last for more than about two decades, and have other claims on them. When the surpluses wane, major new financing would have to be found. That would entail cutting other parts of the federal government heavily, imposing large tax increases, or borrowing substantial sums and swelling the already-large budget deficits that CBO forecasts will return when the baby boom generation has retired. The Social Security actuaries project that in the absence of budget cuts or tax increases to pay for the plan, the Feldstein proposal would reduce budget surpluses and then increase deficits by $6.4 trillion over the next 25 years.(1)

The Feldstein plan essentially poses as a "free lunch" entailing no tough choices. This appearance, however, is deceptive. The plan leaves a gaping hole in financing that future Congresses would have to address. The plan is reminiscent in this respect of the feature of the first Reagan budget that David Stockman called a "magic asterisk." And because large reductions in other programs ultimately would be needed unless taxes were raised substantially or deficits allowed to mount, the plan would place programs funded through general revenues at a significant disadvantage and likely lead to some sacrifice of the needs of younger generations to finance increased benefits for the elderly.

Furthermore, by soaking up much of the surpluses and likely leading to large budget cuts when surpluses dissipate, the plan also would make it harder to rescue Medicare. Medicare's financing hole is so large (much larger than Social Security's) that it is highly unlikely Medicare solvency can be restored without a combination of major, and controversial, program reforms and additional resources. The Feldstein plan would make it more difficult to find such resources.

Moreover, the increased retirement benefits the plan would provide would accrue disproportionately to the affluent. Retirement income would rise much more, both in dollar terms and in percentage terms, for the elderly who are most well-off than for those with modest incomes. Cutting other government benefits or raising taxes to finance increases in retirement benefits that go disproportionately to the affluent elderly is a formula for exacerbating disparities in income in the United States, which already are at their widest point since the end of World War II and exceed the disparities in any other western industrialized nation.

Of particular concern, the Social Security component of the Feldstein plan is likely to prove politically unsustainable over time. The plan provides federal financing for the establishment of individual accounts and reduces Social Security benefits $3 for each $4 in income that a retiree receives from his or her account. As a result, when the plan is fully mature and workers entering the labor force today are reaching retirement age, many middle- and upper-income retirees would receive most or all of their retirement income from their private accounts and little or none from Social Security benefits. Yet they would have paid substantial payroll taxes to the Social Security trust fund.

A Social Security system that collects substantial payroll taxes from everyone but provides the bulk of its Social Security benefits to retirees who earned below-average wages and their spouses and survivors is not likely to endure. The private accounts would appear to a large segment of the population to be a much better deal than Social Security, and pressures to shift more of workers' contributions from Social Security to their individual accounts would inevitably mount and could prove irresistible. The Feldstein plan thus contains the seeds of its own transformation to a much fuller privatization of Social Security.

While posing these problems, the plan appears to fall short of achieving the basic goal of restoring long-term solvency to the Social Security system. In their analysis of the Feldstein plan, the Social Security Administration actuaries found it unlikely the proposal would fully restore Social Security solvency.(2) (This matter is discussed in the box below.)

Finally, although the plan's proponents claim it would boost national saving, the plan would be at least as likely to reduce national saving as to enlarge it. In its analysis of the plan, the Congressional Budget Office noted that compared to current law, "The Feldstein proposal would increase future budgetary pressures and most likely reduce national saving...."(3) The Social Security actuaries also concluded that the plan's "effects on national saving and investment are unclear."(4)

These issues are discussed in more detail below.


How the Feldstein Proposal Works

Under the Feldstein proposal, workers would contribute two percent of their earnings, up to the maximum amount of earnings subject to the payroll tax (currently $68,400), into private retirement accounts managed by financial investment companies. These contributions would be in addition to the payroll tax that workers currently pay to the Social Security trust fund. Workers' contributions to these accounts would be reimbursed dollar-for-dollar by the federal government through a refundable income tax credit. As a result, the government would pay all of the cost of these deposits in private accounts; workers would get these accounts without paying additional amounts themselves.

When a worker retired, the worker's account would apparently be converted to an annuity paying a monthly benefit for as long as the beneficiary remained alive. As noted, for each $4 dollars the retiree received in income from his or her account, the Social Security benefits to which the retiree otherwise would be entitled would be reduced $3.

Under this approach, all retirees would appear to come out ahead. If due to some catastrophe a retired worker received no income at all from his or her account, he or she still would receive his or her full Social Security benefit, without any benefit reductions. And as long as the worker received any income from the individual account, the combined amount the retiree would get from the private account and Social Security would exceed the Social Security benefit the retiree would receive under current law.

Faced with a Social Security shortfall that results from retirement income promises that exceed the revenue the government will collect to make good on them, the Feldstein plan would make the promises of government-funded retirement income more generous and do so without increasing the revenue the federal government would collect. This raises an obvious question. Where does the money come from to cover the large additional cost of financing the refundable tax credit that would reimburse the workers for their deposits into private accounts?

According to Feldstein and his colleague Andrew Samwick, the refundable tax credit would be paid for through about 2015 by using government budget surpluses. They acknowledge that between 2015 and 2030, as the surpluses dissipate, the refundable tax credit would have to be financed through other means — cuts in other parts of the budget, tax increases, or deficit spending. Feldstein and Samwick contend that starting in 2030, the plan would pay for itself; they assert that the private accounts would increase national saving and investment to such a degree that corporate profits would rise to substantially higher levels than would otherwise be the case and the increased corporate tax revenue collected on these increased profits would provide all of the additional funding needed. The amount of additional funding needed would be significantly reduced because Social Security would not be paying out as much in benefits due to the provision of the plan specifying that Social Security benefits be lowered when a beneficiary receives income from a private account.


Weaknesses of the Plan

1.  Increasing Retirement Income at the Expense of Other Needs

As noted, retirees would receive retirement income from their private accounts as well as a Social Security benefit initially calculated under current benefit rules but then reduced $3 for each $4 the beneficiary is receiving from his or her private account. This means workers would be guaranteed a level of retirement income equal to the Social Security benefit they would receive under current law plus at least 25 percent of the income from their individual accounts. As Brookings economists Henry Aaron and Robert Reischauer note in their new book, Countdown to Reform: The Great Social Security Debate, the Feldstein plan would provide larger total retirement benefits than those offered under any other major reform plan under consideration or under the current Social Security system, despite the fact that the current system is short of revenue to honor its promises.(5)

To address the costs of increasing government-funded retirement income, the plan would first consume much of the unified budget surplus, 98 percent of which over the next decade is contributed by the Social Security system. When the surpluses wane, a budget crunch would hit. Taxes would have to be raised, other programs cut, or large deficits incurred.

The plan thus would render it more difficult to provide increased resources to meet unaddressed needs in other areas such as Medicare — a program with a long-term financing shortfall much larger than Social Security's — and areas ranging from education to expanding health care coverage to promoting basic research that can help boost productivity. It also would make it harder to pass tax cuts. In fact, once the surpluses dissipate, the overall resources available for other needs would have to be reduced substantially; the plan would create a necessity at that time for major new rounds of budget-cutting or tax increases, unless budget deficits were allowed to rise to unprecedented levels. Aaron and Reischauer note that the plan "... would generate severe budget pressures, particularly after currently projected budget surpluses end. The fiscal duress would affect all government spending and taxes."(6)

In short, the plan would increase retirement income for the elderly — especially the more affluent elderly — at the expense of other priorities. When the fiscal impact of the plan is considered in combination with the impact of whatever steps ultimately are taken to shore up Medicare, the consequences for the rest of the budget could be grim.

2.  Boosting Retirement Income Without Paying for It

There is substantial risk that offsetting budget cuts or tax increases of sufficient magnitude to finance the plan would not be approved. (There also is a risk that budget surpluses of the magnitude currently forecast might not materialize.) If this occurred, budget deficits — which CBO already forecasts will return after 2020 and climb to record levels for periods other than war or recession when the baby boomers retire in increasingly large numbers — would climb to still-higher levels.

CBO has sounded a warning about this aspect of the plan. In an August 4, 1998 analysis of the Feldstein proposal prepared for House Ways and Means Committee chairman Bill Archer, CBO warned that because the plan contained no mechanism to finance its added costs, it would add to long-term deficits. CBO also stated that while most other Social Security reform plans seek to reduce Social Security's large unfunded liabilities, "the Feldstein proposal guarantees the current retirement benefits and does not reduce the government's overall liabilities..."(7) CBO added that because the plan's costs "must be financed one way or another, the plan would implicitly increase the tax burden on future workers if no further adjustments are made on the spending side of the budget."

These problems are not limited to a transition period, as Feldstein contends. CBO rejected Feldstein's claim that the plan would eventually pay for itself because it would boost national saving substantially, which in turn would lead to large increases in corporate investment, corporate profits, and corporate income tax collections. CBO finds these claims to rest on highly unrealistic assumptions.(8)

The analysis of the Feldstein plan that the Office of the Chief Actuary at the Social Security Administration issued December 3 reveals the magnitude of the fiscal problem the plan poses. The Social Security actuaries examined three types of costs (or savings) that would result from the plan: the cost of the tax credit the plan establishes to reimburse workers for deposits into private accounts; the savings in Social Security expenditures that would result from the plan's provision that Social Security benefits be lowered when a beneficiary receives income from a private account; and the increased interest payments on the national debt the government would have to make. (Because the plan increases government costs, it would cause the national debt to be larger than would otherwise be the case, thereby increasing the cost of interest payments on the debt.) The actuaries found that when these costs and savings are all taken into account, the plan would entail an added cost of $6.4 trillion over the next 25 years. (The $6.4 trillion consists of $3.4 trillion in costs for the plan's tax credit, minus $500 billion in Social Security savings, plus $3.5 trillion in increased interest payments on the debt. These costs could be offset by $2.9 trillion in program cuts or tax increases — rather than $6.4 trillion — since program cuts or tax increases of that magnitude would keep the national debt from growing larger and would thereby avert increases in the cost of interest payments on the debt.)

The actuaries found the plan would have a negative effect on the federal budget in each of the next 75 years, failing to pay for itself in even a single year. The actuaries reported that while the plan would lessen the size of the funding shortfall in the Social Security system, it would do so by requiring "substantial contributions from the General Fund of the Treasury indefinitely into the future." The actuaries also determined that "...the effect on annual unified budget balances would continue to be substantially negative throughout the long-range period [i.e., for the next 75 years]..."(9)

Moreover, the plan would not only punch gaping holes in the budget but might fall short of fully restoring Social Security solvency. The actuaries' analysis concludes that "it is unlikely that this proposal alone could completely eliminate the currently projected actuarial deficit." (See box on pages 8-9.)

In summary, the Feldstein plan promises future retirees higher benefits than under current law without making any tough choices. It does so by relying upon unrealistic assumptions that overstate the plan's effects on government revenue collections and leaving it up to future Congresses to fill the plan's yawning financing gaps. The Feldstein plan would lead to shrinkage of other parts of government, sizeable tax increases, and/or much larger deficits and also may fail to restore long-term balance fully to the Social Security system.

3.  Inequitable distribution of benefits

Under the Feldstein plan, the affluent would receive much larger increases in retirement income than people with low or moderate incomes. This means that unless budget deficits are allowed to swell, the plan ultimately entails cutting other government benefits or services that may be of primary value to the lower half of the population, or increasing taxes in a manner that cannot be predicted at this time, to finance increases in retirement income primarily for the more affluent half of the population. Ultimately, the net effect is likely to be a transfer of income from those of more modest means to those already living comfortably. This would be done even though high earners have less need for added retirement income since they are more likely to be covered by employer-sponsored pension plans and to have significant personal savings they can use in retirement.(10)

Aaron and Reischauer provide an example which illustrates that gains in retirement income would be much greater for those at higher income levels than for those who earn modest wages. They examine both a low earner whose average earnings over his or her career equal $12,000 a year and a more highly paid worker with average earnings of about $67,000. Under current law, the low-wage worker would receive a Social Security benefit of $560 a month. Aaron and Reischauer estimate that under the Feldstein plan, this worker could be expected to receive $240 a month from his or her private account. Since each $4 in private-account income would result in a loss of $3 in Social Security benefits, the receipt of $240 a month in income from the worker's private account would cause the worker's Social Security benefits to be reduced $180. The worker's net gain would be $60 a month.

The Actuaries' Analysis of the Feldstein Plan

A memorandum the Office of the Chief Actuary of the Social Security Administration issued December 3 examines the impact of the Feldstein plan both on the solvency of the Social Security system and on the overall federal budget. The memorandum essentially summarizes the actuaries' principal findings in the following two sentences: "While it does not appear likely that this provision alone [i.e., the Feldstein plan] could completely eliminate the currently projected OASDI long-range actuarial deficit of 2.19 percent of taxable payroll, it could eliminate a substantial portion of the deficit. While the proposal could improve the solvency of the OASDI program, without reducing the total expected level of retirement income, it would require substantial contributions of revenue from the General Fund of the Treasury indefinitely into the future."(11)a

The memorandum's most significant finding concerns the impact of the Feldstein proposal on the federal budget. The actuaries project that the plan would lower unified budget surpluses and/or increase deficits by $6.4 trillion over the 25-year period from 2000 to 2024.b

These amounts equal the costs both of the tax credit the plan establishes and of the increased interest payments the Treasury would have to make on the national debt (since the plan's increased costs would swell the national debt to higher levels), minus the savings in Social Security expenditures that would result from the provision of the plan that reduces Social Security benefits when a beneficiary receives income from a private account.

Using modestly different economic assumptions, CBO has projected unified budget surpluses totaling $3.4 trillion over the next 25 years. By adding $6.4 trillion in cost over this period, the Feldstein plan would turn an aggregate $3.4 trillion surplus over this period into approximately a $3 trillion deficit.

The plan would wipe out the surplus that CBO projects will emerge in the non-Social Security budget starting in 2006; the non-Social Security budget would instead be in deficit every year. The plan also would accelerate the year in which deficits return in the unified budget. In its August 1998 long-term projections, CBO forecast that surpluses would continue in the unified budget until about 2021. The actuaries' estimates of the budgetary effects of the Feldstein plan suggest that unified budget surpluses would end — and deficits return — in about 2014.

The actuaries' other major finding is that the long-term shortfall in the Social Security program would be reduced but probably not eliminated. That the plan would improve Social

Security solvency is to be expected given that the plan pours trillions of dollars into private accounts and requires that Social Security benefits be reduced $3 for each $4 paid in retirement income from a private account.

Under what the actuaries term the "basic plan" — their best assessment of how the Feldstein plan would work — the plan would reduce the long-term shortfall in the Social Security system by close to two-thirds. But it would delay the point at which Social Security becomes insolvent by only four years, from 2032 to 2036.

The actuaries also make a series of alternative assumptions about how the plan might work. Under the set of assumptions most favorable to Feldstein, in which an optimistic assumption is used in nearly every area in which an assumption must be made, the plan would eliminate 97 percent of the long-term Social Security shortfall, although the program still would experience a temporary period of insolvency starting in 2039. This set of assumptions assumes more aggressive and risky investment strategies by beneficiaries than may actually occur; the actuaries note that the level of stock market investment reflected in this assumption "is well above the average 401(k) experience, and thus likely represents more risk than the average investor desires." This set of assumptions also assumes there will be no requirement for workers to convert their accounts to annuities when they retire and that workers will "self annuitize" instead — i.e., will withdraw amounts from their accounts on a monthly basis without purchasing an annuity. The actuaries note that such an approach carries "substantial risk" to beneficiaries and that "the retiree who attempts to self annuitize has a very good chance of outliving the assets." (Moreover, Feldstein has said his plan includes an annuitization requirement, so the highly optimistic assumptions would not appear fully applicable to the plan.) Only when the optimistic assumptions are employed simultaneously in each area does the plan come close to restoring Social Security solvency.

Overall, the actuaries analyze eight different possible scenarios (i.e., eight different sets of assumptions) for the plan. The portion of the long-term Social Security shortfall that would be eliminated ranges from less than half of the shortfall to nearly all of it. In no case would the year of insolvency be delayed more than seven years, from 2032 to 2039, and Social Security would still be insolvent in 2073 under all but the most optimistic set of assumptions. One of the striking findings of the actuaries' analysis is that despite the large new amounts of federal resources that would be poured into retirement pensions — and the very large costs that would result (as well as the increases in federal deficits and the national debt that would ensue unless major budget cuts or tax increases were enacted) — Social Security solvency probably would not be fully restored.

Finally, all eight scenarios make a crucial assumption in an area where the Feldstein plan is vague. The plan fails to specify what would happen to the funds in an individual account when the account-holder dies. The actuaries assumed that the government would repossess three-fourths of all funds in such accounts and deposit the proceeds in the Social Security trust funds. It is unlikely, however, that such a provision — which would likely be viewed as a confiscatory 75 percent death tax on private accounts and opposed strongly by account-holders and their families — could survive politically over time even if it could be enacted in the first place. Without the assumption that the government would take back 75 percent of an account at death, the Feldstein plan falls well short of restoring Social Security solvency no matter how rosy the other assumptions regarding the plan are.
a Memorandum from Stephen C. Goss, December 3, 1998, p. 1.
b This amount is the actuaries' projection of the fiscal impact of what they term the "basic plan," which is their best assessment of how the Feldstein plan would work. The actuaries also examined alternative assumptions regarding the Feldstein plan, some more pessimistic and others more optimistic than the assumptions reflected in the "basic plan." The plan's negative fiscal impact over the first 25 years would be about the same even under the most optimistic assumptions — $6.2 trillion rather than $6.4 trillion.

The more highly paid worker would receive a $1,375 monthly Social Security benefit under current law. Under the Feldstein plan, he or she would receive an estimated $1,340 a month from his or her private account. The $1,340 monthly payment from the worker's account would result in a reduction of $1,005 in the worker's monthly Social Security benefit. This worker's net gain would be $335.

Although the federal government was covering the cost of all deposits into the individual accounts (through the refundable tax credit), the more highly paid worker would receive a net gain five to six times larger in dollar terms than the low-paid worker — $335 a month (about $4,000 a year) compared to $60 a month (or $720 a year). In percentage terms, the highly paid worker's gain would be more than two times the gain the low-paid worker would receive; the low-paid worker's pension income would rise 11 percent, while the more highly paid worker's pension income would climb 24 percent. (See Table 1.)

Moreover, these figures probably understate the degree to which the plan would disproportionately benefit the more affluent. Highly paid individuals would be more likely to place the funds in their accounts in investments carrying more risk but providing higher yields, both because these individuals would have access to (and the ability to afford) better investment advice and because their financial position would enable them to bear more risk. Low earners would be likely to invest more conservatively and receive below-average rates of return as a result. Because high earners would tend to receive higher rates of return on their accounts than low earners, the degree to which these plans would disproportionately benefit the affluent would probably be greater than the above example suggests.

Table 1

Wage Earner

Average Monthly Earnings Social Security Benefit under current benefit structure Monthly Income from Private Account Social Security Benefit After Offset Total Pension Income under Feldstein Plan (Sum of income from private accounts and Social Security) Overall Change in Pension Income
Low Earner $1,000 $560 $240 $380 $620 +$60/ +11%
High Earner $5,600 $1,375 $1,340 $370 $1,710 +$335/ +24%
Source: Henry Aaron and Robert Reischauer, Countdown to Reform, p. 127.

Low-wage workers also might face other difficulties. Some low-income workers who do not normally file an income tax return (very poor workers are not required to file) might continue not to file and lose the refundable tax credit designed to reimburse them for their deposits into their private accounts. Today, some low-income workers entitled to the Earned Income Tax Credit lose it because they do not file a return; the same might hold true here.

4.  Potential for the Plan to Undermine Long-term Political Viability of Social Security

The Feldstein plan's political allure relates to the fact that it seems to let policymakers "have their cake and eat it too." For those who support the guaranteed benefit that Social Security provides and its special protections for low-wage workers, spouses and divorced women without substantial earnings records, widows, the disabled, and the children of disabled and deceased workers, the Social Security benefit structure would be maintained. For proponents of privatization, individual accounts would be created. In addition, retirees would receive more retirement income than under current law without having to pay more out of their incomes to secure it.

Even if it could somehow secure adequate financing, however, the plan is unlikely to be sustainable over the long term. Over time, it could lead to the dissolution of Social Security.

Aaron and Reischauer's example of how low-wage and high-wage workers would fare under the plan helps to illuminate this point. Under current law, the low-wage worker depicted in Table 1 would receive a $560-a-month benefit from Social Security, while the high-wage worker would receive $1,375. Under the Feldstein plan, the low-wage worker would receive a Social Security benefit of $380 a month when his or her benefit was reduced by $3 for each $4 the worker received from a private account. Meanwhile, the high-wage worker would get a Social Security benefit of $370 a month — a slightly smaller amount — when his or her benefit was reduced in this manner. The high-wage worker would receive a lower Social Security benefit despite having contributed tens of thousands of dollars more into the Social Security trust fund than the low-wage worker.

To be sure, the Feldstein plan would give much larger tax credits to high-wage workers than low-wage workers and, as explained above, would raise overall income much more for highly paid workers than for others as a result of the income that the private accounts would generate. But the Social Security program would have the appearance of being unfair to middle- and upper-income individuals.

Indeed, when the system the Feldstein plan establishes was fully mature, it would result in workers with above-average wages and salaries getting only small — and in some cases, no — benefits from Social Security when they retired, because of the reduction of $3 in Social Security benefits for each $4 received from private accounts. Yet these workers would have paid hefty payroll taxes to the Social Security trust fund.

Because people would seem to be paying substantial payroll taxes to Social Security and getting little back from it, Social Security would likely appear to much of the middle class and more affluent segments of the population to be a bad deal. It would seem to provide them a very poor rate of return compared to what their private accounts were paying. These disparate rates of return would partly reflect the fact that the Social Security trust funds would bear all of the burden of financing the benefits of workers who had already retired or worked for many years before the individual accounts were established. The trust funds also would bear all of the burden of providing more adequate benefits to low-income retirees, low-earning spouses and divorced women, and covering widows, the disabled and the children of disabled and deceased workers. Although not obvious to many workers, a sizeable portion of the Social Security payroll tax is essentially an insurance premium for the disability and life insurance protection that Social Security provides. The private accounts, by contrast, would bear none of these burdens, which would enable them to appear a better deal to the average worker.

Moreover, workers would receive periodic statements from the Social Security Administration, as is required under current law, as well as statements from the investment company managing their private accounts. Workers comparing these statements and not appreciating the burdens Social Security had to bear would conclude that Social Security compared poorly to the private accounts. The fact that individual accounts could be invested in equities while Social Security resources would continue being invested solely in low-yielding Treasury securities would exacerbate this problem.

In short, under such a system, Social Security is not likely to be politically sustainable over time. Aaron and Reischauer observe that the Feldstein plan "is likely to undermine political support for a defined-benefit guarantee like Social Security among high and moderate earners because most of them would eventually receive pensions based predominantly on their private accounts."(12)

The plan might not be sustainable politically for another reason as well. Workers and retirees might sharply criticize the reduction of $3 in Social Security benefits for each $4 received in income from an individual account as a confiscatory 75-percent tax on their accounts. If the 75 percent offset rate were lowered, however, the Social Security financing shortfall would grow larger, ultimately requiring Social Security benefit cuts or payroll tax increases. These unattractive choices might add to pressures to replace Social Security more fully, or entirely, with private accounts.(13)

Aaron and Reischauer on the Political Sustainability of
Social Security in Combination with Private Accounts

In their new book on the Social Security debate, Henry Aaron and Robert Reischauer write: "...Plans that combine traditional Social Security with new private accounts contain an element of political risk. The personal accounts component of such hybrid systems would pay a return equal to the yield of whatever assets depositors chose for their portfolios. The traditional Social Security component would provide workers with smaller benefits per dollar of payroll taxes than the personal accounts because most of the taxes would be used to support benefits for previous retirees. In addition, some of the taxes paid by high earners would go to support the social functions of Social Security — the extra benefits for low earners and families with children. None of the taxes deposited into personal accounts would be siphoned off for these purposes. Because Social Security would continue to bear responsibility for supporting past retirees and providing social assistance, it would appear to generate lower returns, especially for middle and high earners, even if the trust funds earned returns as high as or higher than those of individual accounts. After comparing the returns on personal accounts with the apparent yield offered by Social Security, many workers would conclude that they could do better if they were permitted to shift payroll taxes from Social Security to their individual accounts. The conclusion would be false....But the threat to the political viability of Social Security would be real. It is doubtful whether such a system would be politically sustainable.

"....A central question about such whether they would protect low earners in practice as well as they appear to do on paper. Those who support Social Security in part because it has created a stable and mutually supportive coalition on behalf of pensions for everyone and social assistance for low earners have reason to be concerned that various individual account proposals would put social assistance in jeopardy..."a
a Henry J. Aaron and Robert D. Reischauer, Countdown to Reform, pp. 157-158.

5.  Would these plans boost national saving?

Much of the math underlying the Feldstein plan rests on the assumption it would substantially boost national saving. This assumption is dubious.

The Congressional Budget Office has found that compared to current law, the plan would likely reduce national saving. The tax credit the plan provides to cover the costs of the deposits in private accounts would reduce the budget surplus and, in later years, increase the deficit. Government surpluses add to national saving, while government deficits reduce it. By reducing near-term surpluses and swelling long-term deficits, the plan would shrink national saving.(14)

What if Social Security Retirement Benefits Were Eventually
Replaced Fully by Private Accounts?

If Social Security retirement benefits ultimately faded away and were replaced by private accounts, the retirement income of the nation's elderly population would be subject to market risk to a greater degree. The amount of income an individual had to live on in his or her declining years would depend heavily on how the markets performed and how lucky or wise the individual was in his or her investments, as well as on what portion of the individual's account was consumed by administrative and management fees.

Work by Gary Burtless of Brookings is of note in this regard. Burtless examined what would have happened to average male wage-earners if they had entered the labor market at age 22 and established individual accounts, contributed the same percentage of their earnings to these accounts, and made identical investments in indexed funds. He found that an individual reaching retirement age and converting his or her account to an annuity at age 62 in 1975, a year the market was down, would have received a monthly annuity check for the rest of his or her life less than two-fifths as large as the monthly check an individual who reached retirement age and retired seven years earlier, in 1969, would have received. Although steps can be taken to lessen the effects of stock market swings on private accounts, those who retire and convert their accounts to annuities during "bear markets" still would have substantially less to live on than those fortunate enough to do so during "bull markets." (As another example, a retired worker who feared the market would keep falling and converted a retirement account invested in equities to an annuity or a savings account on August 31, 1998 would have lost 21 percent of the income the account would generate, compared to what the worker would have received if he or she had converted the account to an annuity six weeks earlier.)

The replacement of Social Security with private accounts would pose particular risks for women. For example, under Social Security, a divorced woman can receive Social Security spousal benefits (described below) if she was married to her ex-spouse at least 10 years, is 62 or over, and has not remarried. After her ex-spouse dies, she qualifies for survivor benefits. These benefits reflect the fact that many women sacrifice earnings to raise children or to perform unpaid work in a school, church, hospital, or other setting and would not have adequate retirement income if they receive benefits based solely on their own earnings record. The point here is that the spousal and survivor benefits a divorced woman receives do not reduce either the Social Security benefits her ex-spouse himself receives or the spousal and survivor benefits for which his current spouse may qualify if he has remarried. A private account, by contrast, does not enlarge to cover more beneficiaries in the event of divorce or death; there is no more money in the account to go around. Any amounts taken from a private account — for instance, for a divorced spouse — reduce the amount remaining in the account for other beneficiaries. Divorced women would be likely to fare less well under individual accounts than under Social Security.

So would married women who spent a number of years out of the labor force raising children. Social Security pays the lower-earning spouse a spousal benefit that equals 50 percent of the higher-earning spouse's benefit if the benefit the lower-earning spouse otherwise would receive, based on her own earnings record, would be smaller than that. An individual account does not enlarge to pay spousal benefits. Among those for whom this would pose problems are two-parent families with stay-at-home mothers.

Low-wage workers would be in particular jeopardy. The benefits that Social Security provides to a low-wage worker equal a much higher percentage of such a worker's pre-retirement earnings that the benefits Social Security pays to a high-wage worker equal of that worker's earnings. Individual accounts generally contain no mechanism to provide more adequate benefits to low-wage workers (or, for that matter, to widows and other dependents of deceased workers, divorced women, or low-earning spouses).

Furthermore, substantial portions of the funds in private accounts could be consumed by administrative and management costs and hence not be available to provide retirement income. Annual fees on investments in stock mutual funds now average 1.2 percent; a one percent annual charge, the amount the Social Security Advisory Council estimated would be charged on privately managed accounts, would consume approximately 20 percent of the funds in an account over a worker's 40-year work career. (Some approaches that would use the federal government to manage the accounts would reduce these administrative costs significantly, but the costs would still be substantial.)

In addition, when amounts in retirement accounts are converted to annuities, the company selling the annuity typically takes about another 15 percent to 20 percent of the value of the accounts to cover its costs, risks, and profits. In combination, administrative costs and annuitization costs can consume a quite-substantial share of the resources in private accounts.

Finally, there is a question of the administrative feasibility of individual accounts. A new report by the Employee Benefit Research Institute warns the administrative complexities are daunting; EBRI reports that establishing individual accounts for 148 million American workers "could be the largest undertaking in the history of the U.S. financial market, and no system to date has the capacity to administer such a system." EBRI also cautions that individual accounts cannot be administered like 401(k) plans, with contributions made each pay period through payroll deduction, without adding significant employer burdens, especially on small businesses.a
a Kelly A. Olsen and Dallas L. Salisbury, "Individual Social Security Accounts: Issues in Assessing Administrative Feasibility and Costs," Employee Benefit Research Institute, November 1998.

CBO found other weaknesses as well in Feldstein's arguments regarding the plan's effects on national saving. As noted earlier, Feldstein assumes each dollar placed in a private account would be an additional dollar the account-holder saved. CBO disagrees, noting that "other saving most likely would fall because personal retirement accounts would increase workers' resources and would reduce their need to save in other forms for their own retirement."(15) In other words, some of the funds in private accounts probably would substitute for funds workers would have saved anyway.(16)

Another problem with Feldstein's arguments about national saving is that he assumes all funds in private accounts will remain there until retirement and not be withdrawn until then. A requirement barring early withdrawals could, and probably would, be part of legislation initially establishing private accounts. But such a requirement would not be likely to last long. Since these accounts would be marketed and thought of as an individual's private property, individuals would eventually insist on access to their accounts for such purposes as medical emergencies, purchase of a home, college tuition, and the like. The rules barring access to the accounts would likely weaken over time, resulting in less saving.

The history of Individual Retirement Accounts is illustrative. When IRAs were originally established, early withdrawals from them were not permitted. Over time, as Aaron and Reischauer point out, "pressures have mounted to permit individuals to gain access to IRA balances before retirement. Congress has succumbed to those pressures, giving individuals access to these accounts under a lengthening set of conditions."(17) The most recent expansion of early access to funds in IRAs came just last year, on a bipartisan basis, in the Taxpayer Relief Act of 1997. Congress also has allowed individuals in the Thrift Savings Plan, the plan for federal employees, to borrow against their accounts.

The Gramm Variant of the Feldstein Plan

Senator Phil Gramm (R-Texas) has developed a modified version of the Feldstein plan. The Gramm version maintains the core elements of the Feldstein design: 1) workers would make deposits into private accounts; 2) when they drew retirement income from their accounts, their Social Security benefits would be reduced (in this case, by $4 for each $5 in income from their accounts); and 3) no changes would be made in the Social Security benefit structure, nor would any new government revenues be raised. Under both the Feldstein design and the Gramm variation, all workers would receive as much or more in retirement income as under current law, and long-term Social Security solvency would supposedly be restored without any benefit reductions or tax increases.

The Gramm plan differs from the Feldstein design primarily in how it finances the deposits into private accounts. Instead of providing workers with a refundable tax credit to pay for the deposits into their private accounts, the plan would divert three percentage points of each worker's Social Security payroll tax contributions from the Social Security trust funds to the worker's private account (and transfer some general revenue surpluses into the Social Security trust fund to help finance Social Security benefits for current retirees). After 2037, the portion of payroll tax contributions diverted from the Social Security trust fund to individual accounts would rise further, reaching eight percentage points or each investor's Social Security payroll tax contributions by 2055.

Like Feldstein, Gramm acknowledges his plan is not fully financed during a transition period. Senator Gramm says that after budget surpluses begin to wane, there would be about a 20-year period when additional resources would be needed, with the result that other programs would have to be cut, revenues raised, or deficits incurred. Gramm contends that his plan would pay for itself after 2037. Like Feldstein, Gramm assumes his plan would generate a large increase in national saving and investment, attendant increases in corporate profits, and sizable increases in corporate tax revenues that would help it pay for the added government resources the plan would devote to retirement income.

Analyses of the Gramm proposal comparable to those CBO and the actuaries have prepared on the Feldstein plan are not available. The Social Security actuaries did, however, conduct an analysis of the Gramm proposal in April.a (The current version of the Gramm proposal is similar, although not identical, to the version the actuaries examined.)

The actuaries' found the Gramm proposal lacked sufficient information for them to estimate its effect on the solvency of the Social Security system. They raised serious questions, however, about Senator Gramm's contention that the plan would be fully financed after 2037, warning that the plan could be out of balance indefinitely. The actuaries also noted that while the plan claims to guarantee the Social Security benefits that the current system provides and to ensure that no one would receive less than these benefit amounts, the plan may fall short of providing the necessary financing for the Social Security trust funds to honor this promise.

The actuaries raised particular concern about the ability of the Gramm plan to finance disability benefits and survivor benefits for the dependents of workers who die before retirement. The amount the Gramm plan allocates for these Social Security benefits is only about two-thirds of the amount the actuaries estimate will be needed to finance these benefit under the plans. The plan consequently would appear to place disability and survivors benefits at some risk. Unless a future Congress were willing to redirect some of the resources going into individual accounts back to the Social Security trust funds or to raise Social Security payroll taxes, the Social Security disability insurance program would face a serious financial crunch that could lead to major cutbacks in it.
a Memorandum from Stephen C. Goss, Deputy Chief Actuary, Social Security Administration, "Considerations in evaluating the 'Social Security Preservation Act' Proposed by Senator Phil Gramm," April 27, 1998.

In short, the claim that the Feldstein plan will substantially boost national saving is highly uncertain. That claim provides a slender reed on which to base such radical changes in the nation's six-decade old retirement security system.



At a time when we have not fully funded the promises we have made to the elderly under the current Social Security program, and when we face large financing gaps in Medicare and unmet needs in other areas, the Feldstein plan would make new promises to the elderly and direct substantial new resources to retirement pensions without increasing government revenues to defray these added costs. Eventually, the rest of government would have to shrink, other taxes have to be increased, or deficits allowed to rise to higher levels than the levels they already are projected to reach when the baby boom generation has retired. The plan poses as a "free lunch" entailing no pain or tough choices. In reality, the plan would be likely to put programs funded through general revenues at a substantial disadvantage and to sacrifice the needs of younger generations to increase benefits directed to the elderly, especially the more affluent elderly.

This raises the question of what the goal of Social Security reform is supposed to be. Is it to restore solvency to the Social Security system or to make a significant shift in resources toward retirement pensions (especially for those most well-off) and away from other needs?

The plan also would weaken the progressive nature of the current benefit structure, widening the nation's already-large income disparities. In addition, it would establish a hybrid private account/Social Security benefit structure not likely to be politically sustainable over time. The plan would set in motion a dynamic that could lead eventually to the dismantling of much or all of Social Security as we know it today.

The gap between the political allure plans such as this may hold for elected officials today and the plans' likely long-term impact on millions of Americans is great. In this area of policymaking, a long-term perspective is essential. It would be unfortunate if policymakers were to make decisions on Social Security to maximize short-term political advantages without due regard to their decisions' long-term effects.

End Notes:

1. Memorandum from Stephen C. Goss, Deputy Chief Actuary, Social Security Administration, "Long-Range OASDI Financial Effects of Clawback Proposal for Privatized Individual Accounts - INFORMATION," December 3, 1998.

2. Memorandum from Stephen C. Goss; December 3, 1998, p. 9.

3. Congressional Budget Office, "Analysis of a Proposal by Professor Martin Feldstein to Set Up Personal Retirement Accounts Financed by Tax Credits," August 4, 1998, p.12.

4. Memorandum from Stephen C. Goss, p. 9.

5. Henry J. Aaron and Robert D. Reischauer, Countdown to Reform: The Great Social Security Debate, The Century Foundation Press, 1998, p. 126.

6. Henry J. Aaron and Robert D. Reischauer, p. 127.

7. CBO, p. 2.

8. For example, Feldstein assumes every dollar placed in a private account would be a new dollar in saving — i.e., a dollar that otherwise would have been spent on consumption items rather than saved. This is an assumption most economists would regard as exaggerated, since some individuals will feel less of a need to amass other savings for their retirement if they have an individual account and they consequently may reduce the amounts they otherwise would save. Feldstein also assumes that every dollar in new saving the private accounts generate will be invested in the corporate sector, with none invested in housing or non-corporate businesses; this is another improbable assumption that CBO questions. In addition, Feldstein assumes that after retiring, individuals will continue to receive rates of return that can be secured only by investing in risky portfolios. Portfolios of this nature are inconsistent with the purchase of annuities that provide a fixed, guaranteed monthly benefit payment to people who have retired. The CBO analysis examines each of these Feldstein assumptions and is sharply critical of the Feldstein analysis in this area.

9. CBO, pp. 8-9.

10. Statement of Robert D. Reischauer, p. 14.

11. Memorandum from Stephen C. Goss, December 3, 1998, p. 1.

12. Henry J. Aaron and Robert D. Reischauer, p. 127.

13. In their analysis of the plan, the actuaries note that the 75 percent offset "would likely be unpopular once workers began to retire with substantial distributions from their large account balances and observe large offsetting reductions in their OASI retirement or survivor benefits." See Memorandum from Stephen C. Goss, December 3, 1998, p. 9.

14. CBO notes that if one drops use of the CBO baseline and instead assumes as a starting point that all of the budget surpluses will be used for spending increases and/or tax cuts and that all of these spending increases and tax cuts would increase consumption and not add to national saving, then the Feldstein plan would increase saving compared to this starting point. Such a starting point, however, is one that assumes budget deficits will return sooner and climb to even higher levels than the CBO baseline projects. With such a starting point, it remains highly questionable whether increased amounts of government-funded retirement income should be promised without a way to finance them fully.

15. CBO, p. 2.

16. Feldstein has argued in other papers that Social Security discourages private savings because workers anticipate receiving Social Security benefits when they retire and hence do not feel a need to save as much. These Feldstein arguments are controversial, but if Feldstein is correct on this score, then the plan he has proposed would discourage other saving to a greater degree than the current Social Security system since his plan promises workers they will receive more retirement income than the current system provides.

17. Henry J. Aaron and Robert D. Reischauer, pp. 156-157.

Additional related reports: