Saving Social Security

Robert D. Reischauer
Robert Reischauer Headshot
Robert D. Reischauer Distinguished Institute Fellow; President Emeritus - Urban Institute

November 19, 1998

Mr. Chairman and Members of the Committee, I appreciate this opportunity to discuss the future of the Social Security program with you. My statement deals with three issues:

  • The surprisingly fortuitous environment that exists today for reforming or restructuring the nation’s mandatory pension system,
  • The lessons that can be drawn from past efforts to reform Social Security, and
  • The broad options available to strengthen or restructure the program.

The policy and political environment

Restructuring the largest and most popular federal program, one which touches the lives of virtually all Americans, is a monumental challenge. Such an effort can only hope to succeed if policy makers and the public understand the nature of the problem and the need for change, if Republicans and Democrats are willing to forgo short-run political advantage and agree to cooperate for the long-run national good, if the economy remains strong and the budget in surplus, and if there is strong leadership. On all of these dimensions, conditions are unusually favorable as the nation approaches the new millennium.

Every member of Congress understands the nature of the problem. Each knows that while Social Security currently is recording substantial surpluses which add to the Trust Funds’ reserves, those surpluses will turn into deficits around 2021, and by 2032 reserves will be depleted (see Figure 1). When the Trust Funds are exhausted, substantial benefit reductions or increases in payroll taxes will be needed to sustain the program. The widespread appreciation by lawmakers of the problem has generated a broad bipartisan commitment, at least at the rhetorical level, to address the problem sooner rather than later.

The public is also well aware of the long-run difficulties that face Social Security. For at least a decade, the media, pundits, and politicians have bombarded Americans with the consequences of the retirement of the baby boom generation for Social Security and Medicare. Polls indicate that most people realize that the program, as currently structured, is not sustainable. Polls also show that the public would like its political leaders to address the problem. Because the economy is strong, incomes are rising, and inflation is low, the majority puts “fixing Social Security” high on the list of public sector priorities—higher than expanding various social programs or cutting taxes.

For over a decade, the unified budget deficit problem has cast a pall over virtually all policy initiatives. With the focus on deficit reduction, any initiative to strengthen Social Security’s financial position could have been characterized as an attempt to balance the unified budget on the backs of Social Security beneficiaries or taxpayers. Three multi-year deficit reduction packages, enlightened monetary policy, a strong economy, and a hefty dollop of good luck have combined to banish the deficit scourge, at least for the next few years. According to CBO’s baseline projections, the unified budget should be in surplus until around the middle of the second decade of the next century. If policies are not changed, aggregate surpluses over the next decade will amount to roughly $1.5 trillion; if these surpluses are used to pay down the national debt, the ratio of debt to GDP by 2015 will be lower than at any time since before the Great Depression. Of course, over the next few years, Social Security will account for all of the projected unified budget surpluses; non-Social Security taxes are expected to fall short of covering the costs of the government’s non-Social Security activities until after 2004 (see Figure 2). Nevertheless, the projected unified budget surpluses should provide a bit of fiscal flexibility that policy makers may use to ease the transition to a reformed Social Security system.

In the past, few politicians have been eager to lead Social Security reform efforts because such undertakings are fraught with political risks. Over the course of the 1990s, however, many Members of Congress have put forward proposals that would fundamentally restructure Social Security in an effort to strengthen the program for the long term. Their courage has stimulated a lively debate and reduced the political risks of a frank discussion of this issue. Constructive as this has been, someone of national stature must provide leadership, structure the debate, communicate with the public, and push the effort forward when, as it inevitably will, movement stalls. Realistically speaking, this role can only be filled by a president. Barring an immediate crisis as there existed in the early 1980s, however, few presidents will volunteer for this assignment. President Clinton, a president with a keen sense of history, a desire to leave a significant legacy, and the freedom that comes with a second term, has stepped forward to provide such leadership. It may be some time before another chief executive is willing to fill that role.

As we inch closer to beginning a national dialogue and debate on how best to ensure that future generations have adequate incomes in retirement, both policy makers and the public should understand that a very unusual confluence of circumstances has created the current environment in which it is possible to consider addressing Social Security’s long-run fiscal problems before they reach crisis proportions. But this environment is as fragile as it is rare. It far from guarantees that the effort will succeed or even get off the launching pad without exploding. Rather, it means that reform has become a long shot rather than an impossibility.

Some will argue that there is no need to act preemptively. They will point out that, under current projections, the Trust Funds will not be depleted until 2032 and that long-run projections are fraught with uncertainty. Small increases in trend economic growth, the fertility rate, immigration, or real interest rates could push the date of insolvency off for a decade or more. The effect that small changes in current economic conditions and those assumed for the future can have on the long-run outlook was illustrated dramatically when the 1998 Trustees report estimated that the date at which the Trust Funds would be depleted was 2032, not 2029 as estimated in the 1997 report. Some analysts expect the 1999 Trustees report to contain another small reprieve.

The fact that current projections do not point to an immediate crisis should not be used as an excuse to put off action. Uncertainty is a two edged sword. Current projections of the Trust Funds’ balances could prove to be too optimistic if medical advances and improved personal health add more to average life expectancy than the actuaries have assumed or if economic growth falls short of expectations. Moreover, Social Security will begin to put pressure on the budget long before the Trusts Funds are exhausted. The position of the unified budget is improved by Social Security only as long as the program’s primary surplus—its non-interest income less its expenditures—grows.

This primary surplus, which was about $52 billion in fiscal 1998, will begin to decline by a few billion dollars a year after 2003 (see Figure 1). When this occurs, the unified budget’s position will begin to deteriorate if taxes are not raised or spending is not cut. If the unified budget is in surplus, this deterioration may be viewed as acceptable because it will be attributable to the shrinking primary surplus of Social Security. The situation will become more severe around 2021, however, when Social Security’s income, including interest receipts, falls short of covering benefit payments and administrative expenses, forcing the program to redeem some of the Treasury securities held by the Trust Funds. Unless the on-budget accounts are in substantial surplus at that time, the nation will be faced with the unpleasant choice of increasing borrowing from the public, hiking taxes, or slashing expenditures.

While there may be no fiscal imperative to act this year to reform Social Security, the sooner we begin to strengthen the fiscal position of the nation’s mandatory pension system the easier decisions will be, the more gradual the process can be, the less wrenching the adjustments need be, and the more options policy makers will be able to consider.

Retirees and those approaching retirement often have little or no ability to compensate for or adjust to benefit reductions or tax increases. Both political reality and considerations of equity, therefore, require that whatever shape reforms take they not change significantly the rules of the game for such individuals. This judgement is reflected in the proposals that would partially privatize Social Security, virtually all of which leave those age 55 and older under the existing system. This means that if decisions concerning how best to reform Social Security are postponed until action is unavoidable two decades from now, the bulk of the large babyboom generation will not contribute to solving the problem and the burden on younger generations will be all the larger. Even a decade from now the politics of reform will be much tougher than they are now. Starting in 2002, the fraction of the adult population age 55 and over will begin to rise more rapidly than it has during the demographic holiday the nation has enjoyed over the past decade (see Figure 3). By 2021, when Social Security’s surpluses are projected to turn into deficits, 39 percent of the adult population will be age 55 or older, up from 29 percent this year. The growth in the importance of retirees and near retirees among likely voters is even more dramatic because older citizens tend to vote at higher rates than do younger Americans. While less than one third of voters in the 1996 presidential election were age 55 and over, 45 percent of voters in the elections of 2020 will be in this age group if current age specific voting patterns persist. In such an environment, increased taxation of benefits and delays or reductions in COLAs will probably be ruled out. The longer decisions about reform are put off, the greater the role payroll tax hikes are likely to play in the solution. Because delay constrains the policy options that realistically can be considered to strengthen the system, it would be wise to act now even if the program changes agreed to are not phased in for a decade or two.

Favorable economic, budget and political conditions have opened a window of opportunity for reform. This window may close abruptly and not reopen before today’s problem has been transformed into a full blown crisis which demands precipitous action, as was the case in 1983.

Lessons from Past Efforts

Policy makers have addressed Social Security’s long-run fiscal problems a number of times over the course of the last quarter century. The two most substantial initiatives—those of 1977 and 1983—were prompted by the impending exhaustion of Trust Fund reserves. Unlike the current situation, some immediate corrective legislation was unavoidable. Despite this key difference, the experiences of the past offer some simple and obvious lessons for the current situation.

The first of these is that policy makers should refrain from portraying the reforms they are proposing as the solution to the problems of the nation’s mandatory pension system for all times. Undoubtedly, economic, demographic, and social developments over the next half century will not follow the paths that seem most likely today. Further adjustments may be required quite soon after any major reforms are adopted, as was the case after the 1977 legislation was enacted.

A second obvious lesson is that significant changes are more acceptable if the workers and taxpayers are given a considerable amount of time in which to prepare. For example, the increase in the age at which unreduced benefits are available that was enacted in 1983 will first affect those turning age 62 in 2000. The seventeen year delay between enactment and implementation has given those affected fair warning and an opportunity to change their personal saving behavior and work effort to compensate for the reduction in benefits that this change represents.

A third lesson that can be garnered from the experiences of the past is that successful reform efforts usually involve gradual, rather than abrupt, change. A giant leap is often unacceptable whereas a series of small steps that end in the same place is viable. For example, the two year increase—from age 65 to 67—in the age at which unreduced benefits will be paid will take place two annual month steps stretched over a 23 year period. Similarly, the bite imposed by subjecting a portion of benefits to the income tax will grow gradually each year because the income thresholds above which this policy applies are not indexed.

Finally, the experience of the past suggests that bipartisan cooperation is indispensable if the reform effort is to succeed. Debates over Social Security offer unlimited opportunities to engage in demagoguery for political advantage. The issues are complex, the consequences of some reform proposals are unknown, many people do not understand how the program now works, and the program’s benefits are vitally important to the well being of most older Americans. Such conditions create a highly combustible environment in which to hold a national debate about restructuring. It will take immense self restraint and rhetorical moderation on the part of lawmakers, policy experts, and interest groups to move forward without touching off a conflagration that could scar the political landscape so badly that few would risk raising the issue again until the problem has become a full blown crisis. Divided government—Republican majorities in Congress and Democratic control of the White House—may improve the outlook because the responsibility for governing the nation and for policy development is shared by both political parties.

Evaluating the Broad Options for Reform

The debate over how best to provide basic income for future retirees has been about two broad options. Advocates of the first of these, which goes under the generic label of privatization, believe that the nation’s interests would best be served if Social Security were scaled back and a defined-contribution pension plan consisting of individual accounts were established to supplement Social Security. Proponents of the other approach believe that it is important that the nation’s mandatory pension system remain exclusively a defined-benefit plan. They therefore seek ways to strengthen the long-run financial position of the current system through measures that would increase the program’s income and reduce its expenditures while preserving Social Security’s underlying principles. Each broad approach has advantages and disadvantages. In addition, as is so often the case, the details of each specific proposal make a great deal of difference.

When evaluating the broad options for reform, it is important to keep the fundamental purpose of the nation’s mandatory pension system clearly in focus. That purpose is to provide workers with a secure and predictable basic retirement pension, one that will last as long as the worker and the worker’s spouse are alive and one whose purchasing power will not be eroded by inflation. This pension should be viewed as the foundation upon which other sources of retirement income are built.

It is also useful to lay out explicit criteria for comparing the broad options and specific plans. My colleague Henry J. Aaron and I have spelled out four such dimensions in a book (Countdown to Reform: The Great Social Security Debate) that will be released by The Century Foundation Press on December 1. They are the adequacy and equity of the benefits provided by the system, the degree to which the system protects participants against risks which they are ill equipped to bear, the system’s administrative costs and complexity, and its impact on national saving.


Closing the Projected Long-term Social Security Deficit

  Deficit or Change in Deficit As Percent of Payroll Proportion of Adjusted Long-term Deficit Closed
Projected long-term deficit—1998 Trustees Report 2.19 N/A
Effects of correcting the Consumer Price Index -0.45 N/A
Adjusted long-term deficit 1.74 N/A
1. Gradually reduce spouse’s benefits from one-half to one-third of the worker’s benefits and raise benefits for surviving spouses to three-quarters of the couple’s combined benefit. + 0.15 – 9
2. Cut benefits by increasing the unreduced benefit age—raise the age to 67 by 2011 rather than by 2022 and thereafter raise the age at which unreduced benefits are paid to keep the fraction of adult life spent in retirement constant. – 0.49 28
3. Increase the initial age of eligibility from 62 to 64 by 2011 and thereafter raise the age of initial eligibility at the same pace as the unreduced benefit age. – 0.23 13
4. Increase the period over which earnings are averaged from 35 to 38 years. – 0.25 14
5. Cover all newly hired state and local employees. – 0.21 12
6. Tax Social Security benefits the same as private pension income. – 0.36 21
7. Gradually invest the trust fund’s balances that exceed 150 percent of yearly benefits in common stocks and corporate bonds. – 1.20 69
Source: Estimates from the Office of the Actuary, Social Security Administration.

Taking into consideration both the fundamental purpose of the mandatory retirement system and the four criteria enumerated above, I have concluded that it would be best to work to strengthen the long-run fiscal position of the existing defined-benefit program. There are many ways in which this might be accomplished. The specific measures Henry Aaron and I have endorsed in our book are laid out in Table 1. Together, they are more than sufficient to close the program’s long-term deficit. This means that some recommended program changes could be phased in more gradually than we have proposed or even dropped from the package of reforms without compromising the objective of closing the long-term deficit.

The particular changes included in our reform package were chosen not simply to close the deficit. They were also intended to modernize Social Security in ways that reflect the economic and societal developments that have occurred over the last half century. Most of these changes are familiar to those who have followed the Social Security discussion during the past few years. However, the largest of the proposed program changes—investing a portion of the Trust Fund balances in private equities—is not and, therefore, needs some further explanation.

Shifting Trust Fund investments from government to private securities would not have a direct or immediate effect on national saving, investment, the capital stock, or production. The Trust Funds, however, would earn higher returns because they would hold assets other than relatively low-yielding government bonds. This would reduce the size of the benefit cuts and payroll tax increases needed to close the program’s long-run deficit. While this is attractive, the question which has troubled many since the inception of Social Security concerns the possibility that Trust Fund investments in private securities might lead to inappropriate government influence over private companies. For example, Social Security trustees might be subject to political pressures that would force them to sell shares in companies that produce products some people regard as noxious (for example, cigarettes or napalm) or that pursue business practices some people regard as objectionable (such as hiring children or paying very low wages in other countries, polluting, or not providing health insurance for their workers) or to use their stockholder voting power to try to exercise control over private companies.

If institutional safeguards were not available to reduce the risk of such interference to a de minimis level, I would oppose such investments. But such safeguards can be established. The core of this protection would be provided by a new institution, the Social Security Reserve Board (SSRB), which would be modeled after the Federal Reserve Board. The governors of this independent entity would be appointed by the president and confirmed by the Senate for staggered fourteen year terms. They could not be removed for political reasons. The SSRB would be empowered only to select on the basis of competitive bids a number of fund managers. The fund managers would be authorized only to make passive investments in securities—bonds or stocks—of companies chosen to represent the broadest of market indexes. These investments would have to be merged with funds managed on behalf of private account holders. To prevent the SSRB from exercising any voice in the management of private companies, the fund managers could be required to vote the Trust Funds’ shares solely in the economic interest of future beneficiaries.

Such a system would triply insulate fund management from political control by elected officials. Long-term appointments and security of tenure would protect the SSRB from political interference. Limitation of investments to passively managed funds and pooling with private accounts would prevent the SSRB from exercising power by selecting shares. The diffusion of voting rights among independent fund mangers would prevent the SSRB from using voting power to influence company management and would protect voting rights of private shareholders from dilution. Congress and the president would have no effective way to influence private companies through the Trust Fund unless they revamped the SSRB structure. While nothing, other than a constitutional amendment, can prevent Congress from repealing a previously enacted law, the political costs of doing so would be high.

While privatization proposals are attractive on some dimensions, I think that individual accounts should not be part of the nation’s mandatory pension system. Such accounts introduce added risk and unpredictability into retirees’ basic pensions. With individual accounts, benefits will vary depending on when during their lives individuals worked and contributed to their accounts, what assets they invested their account balances in, and market values when workers retire. Unless retirees are required to convert their individual account balances into inflation-protected annuities upon retirement, there would be no guarantee that they would have adequate income through their final years.

Furthermore, it is difficult to maintain adequate social assistance in a system in which individual accounts play an important role. The social assistance provided through Social Security has helped workers with low lifetime earnings, spouses with limited or no participation in the workforce, survivors, and divorcees. It has made Social Security the nation’s most important and least controversial anti-poverty program.

In addition, individual accounts, unavoidably, will increase the complexity and administrative costs of the nation’s basic pension system. Added burdens will be imposed on employers, workers, and the government. Under some privatization plans, the administrative load could be sufficiently onerous as to make the proposed system unworkable. Under other privatization plans, the impact of added administrative costs would largely be felt through reduced returns on individual account balances. Even this effect could be significant. A one percent annual charge—which is close to the average mutual fund fee—imposed over a forty year career would reduce the size of a retiree’s pension by about 20 percent.


Strengthening the long-run fiscal position of the existing defined-benefit Social Security program or restructuring that program and supplementing it with a system of individual accounts is going to involve some sacrifice by taxpayers, beneficiaries, or, most likely, both. In recent months, some analysts have suggested that this need not be the case—that there exist ways to “save Social Security” that are painless. Some even go so far as to promise that “current law” benefits need not be reduced nor payroll taxes raised. By and large, these plans utilize the projected unified budget surpluses to fulfill their promise of a free lunch.

I would like to conclude with several cautions about these tempting mirages.

First, the projected unified budget surpluses are not manna from heaven that have no other uses. If they are used to seed individual accounts in a privatized system rather than to pay down the national debt, debt service costs will loom larger in the baseline budget projections and national saving may be reduced. Furthermore, if the surpluses are devoted to “saving Social Security,” they will not be available to sustain Medicare, fund tax relief measures, or support expanded spending on defense, medical research, education, or other areas that many members of both parties regard as high priorities.

Second, as the experience of the past few years amply demonstrates, projections of the unified budget deficit/surplus are very uncertain. Congress and the president can enact policies that reduce surpluses, as was the case in October. But even abstracting from policy changes, estimates of the future position of the budget jump around from year to year because economic assumptions and technical factors change. Figure 4 provides my estimates of CBO’s ten-year baseline budget projections excluding the effects of policy changes. Between the projection released in March 1995 and that issued in July of 1998, the projected fiscal 2005 budget situation improved by about $500 billion. While the budget projections over the last few years have shown a steadily improving bottom line, the experience of the late 1980s and early 1990s serves as a warning that the budget outlook can deteriorate just as rapidly and as significantly as it can improve. Given this situation, there is a risk to linking the future of the nation’s mandatory pension system to surpluses that may or may not materialize. If the expected surpluses are not realized, a contentious debate could develop over whether the federal government should incur deficits and borrow more money from the public so that workers can deposit these funds in their personal retirement accounts.

Finally, proposals that promise future beneficiaries all of the benefits called for by current law with no increase in payroll taxes are distributionally inequitable and represent a misdirection of scarce resources. One such plan would establish a refundable 2 percent income tax credit for deposits workers make into personal retirement accounts. The cost of this tax credit—some $3.4 trillion over the next 25 years—would be financed largely out of the projected unified budget surpluses. Upon retirement, the account holder’s Social Security benefit would be reduced by 75 cents for every dollar of pension provided by the personal retirement account. In other words, no one would be worse off and all those who had a positive account balance would be better off.

Benefits under this approach would rise proportionately more for high earners than for low earners. The contribution to individual accounts and, hence, the size of account balances would be a constant fraction of income. Social Security benefits are proportionately larger for low earners than for high earners. Since the plan would reduce Social Security benefits by three-quarters of any benefits derived from individual accounts, pensions for high earners would rise proportionately more than would pensions of low earners as the following simple numerical example illustrates.

Monthly Amounts
  average earnings Social Security individual account total pension change in pension
Low Earner $1,000 $560 $240 $620 +11%
High Earner $5,600 $1,375 $1,340 $1,720 +25%

Considering that high earners are more likely than low earners to be covered by employer sponsored pension plans and are more likely to have significant personal savings that they can use in retirement, one may question the equity of a proposal that would use budget surpluses to boost pensions disproportionately for those who are well off. One may also question whether the best way to “save” the nation’s mandatory pension system is to increase benefits for all future retirees. It may be that the public sector is forced to increase, above currently promised levels, the resources it devotes to future retirees. But certainly the most likely area for that pressure to manifest itself is in medical care which may absorb amounts much larger than the surpluses projected for the next decade and a half.