S uddenly, Social Security is no longer sacrosanct. Critics say the program is going bankrupt, about to be overwhelmed by the graying of the baby-boom generation. Many young workers have little confidence that Social Security will be there when they retire. This, in turn, has invited an unprecedented discussion of radical structural changes. For the first time, the official Social Security Advisory Council is offering three wide-ranging proposals for restructuring. All three envision investing some Social Security funds in private financial markets. Two of the three would create, for the first time, individual investment accounts financed with Social Security payroll charges.
This is a remarkable turnabout, substantively and politically. In 1983, the last time Social Security's finances were shored up, so strong was the support for the existing system that advocates of privatization on the Reagan White House staff were kept far from the deliberations. In the 1992 presidential election, all candidates repeated the quadrennial ritual of swearing fealty to America's most expensive and popular social program. But suddenly, the idea that a portion of one's Social Security taxes could fund an individual retirement account seems to combine political appeal for baby boomers with the prospect of stabilizing the system's finances and perhaps raising America's low rate of private saving as well.
In what follows, we set the stage for understanding why Social Security is suddenly in play, the dimensions of its financial problems, and the logic of the three reform proposals. What, after all, are the goals of the Social Security program? How well has it met them in the past and can it do so in the future? Is privatization desirable or necessary?
SOCIAL SECURITY IN PERSPECTIVE
Social Security, by design, is unlike any private insurance or retirement plan. It combines a nearly universal pension program with a highly redistributive income transfer program as well as an insurance policy offering disability and survivorship benefits. In contrast to nearly all private pension plans, it is also totally portable and fully indexed for inflation. Because of its inclusiveness, Social Security has been immensely popular since its inception.
In 1995, Social Security benefits exceeding $330 billion were distributed to more than 43 million recipients and their families under the old-age, survivors and disability programs (OASDI), accounting for over one-fifth of all federal government spending. The average benefit was about $7,700 per year. Payroll taxes totaling $365 billion were collected from 141 million workers, averaging about $2,600 per covered worker. Besides payroll taxes, 1995 Social Security revenues included $6 billion in receipts from part of the income taxation of Social Security benefits (the other part supports Medicare), and another $35 billion in interest earned on accumulated reserves. Currently, the agency's revenues exceed expenditures by about $60 billion per year.
Although these reserves are large and growing (about $500 billion at the beginning of 1996), they currently represent less than 18 months of projected payouts. The system, in other words, is not pre-funded like a private annuity or pension plan, but rather relies on pay-as-you-go financing. Today's retirement benefits are paid mostly by today's workers' contributions, an arrangement that has functioned successfully for 60 years.
Measured against the goal of reducing old-age poverty, Social Security has been a notable success. It deliberately transfers resources from those with high lifetime earnings to those with low lifetime earnings. In 1996, each dollar of average (indexed) monthly earnings up to $437 translates into 90 cents of monthly retirement benefit. By contrast, each dollar of average monthly earnings between $437 and $2,635 adds only 32 cents, and each dollar of extra average monthly earnings beyond that adds only 15 cents. No private saving plan, annuity, or whole-life insurance policy has this redistributive, antipoverty feature.
Social Security has profoundly improved the economic well-being of older Americans. Nearly 30 percent of the elderly were poor as recently as 1967, more than twice the overall poverty rate. Social Security retirement benefits were raised substantially during the late 1960s and early 1970s, and poverty among the elderly plummeted. By 1974, the elderly poverty rate had fallen to half of its 1967 level, and it has been below the overall poverty rate since 1982. Social Security benefits provide over 80 percent of the income of the poorest 20 percent of elderly households (and nearly 80 percent for the next quintile), compared to only 20 percent of income for the richest elderly quintile.
Unfortunately this rosy record cannot continue without another round of tax increases, benefit cuts, or more fundamental changes. Precisely because it is so generous, at a time when people are living longer, Social Security is outstripping its means. And the fact that many people have come to expect generous public benefits may have discouraged private and pension savings as well.
Some analogize Social Security to a mandatory chain letter. As long as the economy was growing rapidly and the ratio of workers to retirees was large, the burden of financing ever more generous benefits could be passed along to the next generation. But as the population ages, the ratio of Social Security contributors to beneficiaries has declined, from 5 to 1 in 1960 to 3.3 to 1 today and projected to only about 2 to 1 by the year 2030. In an aging population with a pay-as-you-go retirement system, successive cohorts stand to receive a declining "rate of return" on their lifetime contributions.
THE VANISHING WINDFALL
Past and current cohorts of retirees have received much more back from Social Security than they and their employers contributed, even allowing for a reasonable rate of return. For example, a man with average lifetime earnings who retired at age 65 in 1980 could expect to receive in Social Security retirement benefits 3.7 times more than his contributions would have generated, had they been invested in low-risk government securities. For a similar woman, the ratio was even higher-4.4 times-because of her longer life expectancy. Thanks to the progressive benefit formula, these ratios were even greater for lower- income people. But in the past, even high earners received benefits more than 3 times the value of contributions made on their behalf. The income redistribution within cohorts of retirees was dominated by the income redistribution between cohorts (from workers to retirees).
As the ratio of workers to beneficiaries continues to decline, so will the ratio of Social Security benefits received to taxes paid. For example, most workers reaching age 65 in 1995 can still expect more in lifetime Social Security benefits than an alternative low-risk investment (earning a 2 percent real interest rate) would have provided, according to Eugene Steuerle and Jon Bakija of the Urban Institute. Within ten years, however, many retirees (primarily higher earners) can expect to receive less than a 2 percent real rate of return on their Social Security contributions. Although one-earner couples and those in the bottom half of the earnings distribution will continue to do well, there will be many more recipients who may judge their mandatory Social Security contributions to be a poor investment relative to what they could have done on their own.
The maturing of the system has made its redistributive aspect explicit-and thus politically vulnerable. In the past, the intergenerational income transfers-from workers to retirees-were so generous that they were a better deal than conventional low-risk investments, even for the affluent. But as that windfall gradually disappears, Social Security ceases to be so attractive to higher income earners, and its political base begins to splinter along income lines.
Social Security is also faulted for depressing the national saving rate, since it gives people a retirement check that is actually financed by transfers from the working population, rather than by interest on accumulated savings. The average 1996 Social Security recipient, with annual benefits of $7,700, would need a savings account (paying 5 percent interest) in excess of $150,000 to collect that much in interest. But no such savings account exists.
Ultimately, future retirees' consumption will depend on the productive capacity of the economy. Increasing the economy's future productive capacity, in turn, requires higher rates of saving and investment. Sponsors of more radical reform argue that if Social Security were turned into a partly or wholly funded system, it could become a net contributor to savings. But the more that Social Security is revised to emphasize nontraditional goals-higher individual investment returns and more freedom of choice-the more it loses its redistributive character.
IS THERE A CRISIS?
Social Security is newly vulnerable because it is no longer a good deal for all, and because its projected revenues are inadequate to pay promised benefits. Around 2013, benefit payments will exceed payroll taxes. By about 2019, benefit payments will exceed all sources of OASDI revenues, including interest on the Trust Fund. After this date, less than 25 years from now, Social Security would have to sell Treasury securities to pay benefits and Trust Fund assets would begin to decline. If no further changes were made, Trust Fund reserves would be exhausted by about 2030, when baby boomers will be between 65 and 85 years old, and Social Security would be unable to pay the benefits retirees are due. Obviously, this will not occur, because Congress will act before then, either to raise taxes or change the program's benefit structure.
Over the next 75 years (the traditional accounting horizon for Social Security), the projected deficit will average about 2.2 percent of covered payroll per year. In other words, a payroll tax increase of 2.2 percentage points (i.e., from 6.2 to 7.3 percent for both employers and employees) could eliminate the current 75-year deficit. But solving the current 75-year deficit does not leave the system in true long-term balance, since the red ink rises again as the 75-year window moves forward.
When the system was founded, payroll taxes totaled 2 percent on the first $3,000 of annual earnings. By 1960, taxes had risen to 3 percent of the first $4,800, and by 1980, to 5.08 percent of the first $25,900 (excluding the Medicare tax begun in 1966). Since then, the OASDI tax rate has been increased several more times, reaching the current 12.4 percent in 1990, on covered earnings that are indexed annually and are currently capped at $62,700. Some of these increases were enacted to pay for expanded benefits, but for the most part they reflected a gradual decline in the ratio of workers to retirees.
In 1983, for the first time, some of the balancing occurred on the benefit side. Up to one-half of Social Security benefits were subjected to federal income taxation for high-income recipients, and the normal retirement age of 65 was scheduled to rise slowly to 66 and then to 67, beginning with those turning 62 in the year 2000. The 1983 amendments returned the system to long-range (75-year) actuarial balance. But the system is back out of balance only 13 years later, both because long-range economic and demographic assumptions are now less favorable than in 1983, and because the 75-year accounting period moves forward every year, each time replacing a current surplus year with a future deficit year.
RADICAL REFORM?
In 1994, President Clinton appointed a Social Security Advisory Council, as is done every four years. The committee, representing business (three members), labor (three), the self-employed (one), and the public (five), was chaired by University of Michigan economics professor Edward Gramlich. The council was directed to focus on the system's fiscal stability.
Reflecting the diverse views of its membership and the nation, the council has proposed three very different visions of Social Security reform. The first, termed the Maintenance of Benefits plan, is championed by Robert Ball, a former Social Security Commissioner, and five other members of the 13-person Advisory Council. It follows past tradition by raising revenues to meet current benefit obligations within the system's current structure. In a departure from the past, however, this plan also recommends investing up to 40 percent of Trust Fund reserves in private capital markets.
The second proposal, labeled the Individual Accounts plan, is favored by two members of the council, including Chairman Gramlich. This approach, similar to legislation proposed by Senators Bob Kerrey and Alan Simpson, trims benefits and adds revenues, and then adds a controversial new component-modest individual defined contribution accounts within Social Security, funded by an increase in the payroll charge, over which participants would have limited investment discretion.
The remaining five members, led by Sylvester Schieber, a business representative to the council, are championing a third option, called the Personal Security Account plan. This would replace the current Social Security system with a very different one, with flat benefits independent of earnings, and large mandatory personal retirement accounts, funded by a diversion of part of the current payroll tax, and held and managed by individuals.
Ball's Maintenance of Benefits plan maintains the currently legislated schedule of benefits, and finds new revenues needed to pay for them, from three sources: additional federal taxation of Social Security benefits and the diversion to the OASDI Trust Fund of the income tax receipts currently supporting Medicare; a 2 percentage point increase in the combined employer-employee payroll tax rate in the year 2050; and the allocation of up to 40 percent of Trust Fund reserves to the equity market, which, over the long run, has enjoyed a higher rate of return than government bonds. This proposal envisions no individual accounts and no benefit cuts beyond those already legislated.
The other two plans reduce future Social Security benefits. Gramlich's plan reduces benefits across the board, both by increasing the normal retirement age to age 67 by 2012 and then indexing it to changes in longevity, and more so for high-income workers through changes in the benefit formula. It then goes a significant step further by proposing an additional mandatory payroll charge-1.6 percent of covered earnings-to be dedicated to individual retirement accounts. Although the funds in these accounts would be held by Social Security, individuals would make their own investment choices from a limited set of broad options, such as bond funds and equity indexes.
At any time after age 62, retirees could claim an indexed annuity from Social Security, based on the total accumulations in the account; the funds could not be withdrawn in a lump sum. These individual accounts can be either be viewed as a modest beneficial addition to the current system, an encouragement to national saving, and a popular (and perhaps the only) way to raise the payroll tax-or as a major change in the redistributive philosophy of Social Security, one that might open the door to more significant changes later.
Schieber's option proposes more significant change in the system. Current earnings-related retirement benefits would be replaced by a low, flat-rate benefit, available to all workers with at least 35 years of contributions, regardless of their earnings histories. In today's dollars, the annual benefit would be about $400 per month, about two-thirds of the poverty level for an elderly individual, and significantly less than today's average Social Security benefit. (Prorated flat benefits would be paid to those with 10 to 35 years of coverage.) The mandatory personal security accounts (PSAs) would be funded by payroll taxes of 5 percent of covered earnings, diverted from the current 12.4 percent payroll tax. (About 2.4 percentage points would be needed to finance continuing survivors and disability programs; the individual's half of the remaining 10 percent would be allocated to the PSA, and the employer's 5 percent to the flat benefit.)
The PSAs would be owned and invested by individual participants, with some regulation over the types of funds people could hold. Accounts could be claimed only after age 62, but then could be withdrawn as a lump sum or converted into an indexed annuity provided by the Social Security Administration. Under this third plan, the redistributive and individual equity components of the current system would be completely separated. Workers aged 55 or older (in 1998) would continue under the current system, and enjoy benefits as currently promised, while workers under age 30 would be covered entirely under the new system. Those between 30 and 55 would receive hybrid benefits reflecting their accrued rights under the current system at the time of transition and prorated flat benefits, as well as the proceeds of their personal security accounts. Even with the significant Social Security benefit cut proposed, diverting 5 percent of current payroll taxes leaves a significant unfunded liability. Therefore, additional revenues equal to about 1 percent of aggregate consumption (or about 1.5 percent of covered payroll) over the next 70 years would be required to cover the flat-rate benefits.
Despite their significant differences, all three proposals contain common elements worth noting. They all recommend the maintenance of a mandatory, universal, public social insurance program, with retirement, survivor, and disability benefits. They all maintain a progressive benefit structure, with net transfers toward those with low earnings histories. Each envisions additional taxation of Social Security benefits, with the eventual inclusion of all Social Security benefits (in excess of contributions already taxed) in taxable income. All recommend mandatory Social Security coverage of all new state and local government employees (as is true in some states already).
All three plans also favor increases in the normal retirement age beyond 65, and two of them then index the age to changes in longevity. None proposes means-testing benefits or indexing Social Security benefits at less than the cost of living. All three eliminate the current 75-year Social Security imbalance, and go beyond this to stabilize the ratio of the Trust Fund reserves to annual expenditures between 2050 and 2070, the last two decades of the current planning horizon. Finally, all three plans rely on private-sector investments to generate additional revenues for future retirees.
This is the last quadrennial Social Security Advisory Council. The Social Security Administration is now a quasi-independent agency, and it will be advised by a permanent advisory board. How will this last group's efforts be judged? Some believe that this Advisory Council's lack of consensus will doom its efforts to irrelevancy. But it may well turn out to be just the reverse. By offering three diverse proposals, the council has clarified profound political and philosophical differences about the future of Social Security, which will make for a more informed national debate.
PERILS OF PRIVATIZATION?
The two individual-account plans would create an explicitly double-deck system, with a lower deck focusing on income adequacy and income redistribution, and an upper deck generating benefits directly related to contributions. The lower deck is to be a defined benefit plan, while the upper deck is a defined contribution plan, like a 401 (k). The overall degree of redistribution in the system would depend on the generosity of the lower deck and on the returns earned on investments in the upper deck.
With privatization, Social Security's twin goals of income adequacy and individual equity would be separated and become more explicit and transparent. Over time, the commitment of more affluent taxpayers to the lower deck of the system might erode. The system as a whole might become far less redistributive.
Questions also remain about the effect of privatization on work, saving, and the distribution of income and risk. Privatization might increase the incentives to work at later ages and therefore delay retirement decisions, since one's own contributions would be more directly linked to personal pension accumulations and eventual benefits. Under the current system, working longer and earning more often increases Social Security taxes more than it does eventual benefits. Given the increasing health and longevity of older Americans, this change would be a plus.
Would individuals save more or less under these privatization plans than they now do? Under the Individual Accounts and the Personal Savings Accounts plans, total tax collections would rise, directly financing an increase in saving. Of course, those already preparing for retirement on their own could offset this mandatory change by saving that much less in other ways. Under both of these plans, having more personal control over a sizable portion of their payroll taxes might make individuals feel more confident about the future. Some might respond to this by saving less. On the other hand, allowing people some control over their retirement portfolios might improve financial literacy, generate more serious thought about retirement planning, and thereby increase personal saving rates.
Many other important details remain to be clarified, particularly under the PSA approach. Could the accumulations be tapped for medical, home foreclosure, or other emergencies prior to retirement? Who bears the ultimate risk of poor investment returns, either because an individual chose badly or because the economy as a whole performed poorly? If lump-sum withdrawals are permitted at retirement, what happens to those who squander these savings and then find themselves in need? If annuities were offered but not required, how would we deal with the adverse selection problem, that those choosing the annuity option would tend to be that subset of the elderly with the longest life expectancy?
The concerns are both economic and political. Although equities, on average, have outperformed bonds in the postwar period, there is no guarantee that the future will repeat the past. Novice or poorly educated investors may fare poorly with increased choice and exposure to financial markets. Some critics worry that there will be inadequate public education and regulation to avoid swindles and scandals. Lump-sum allocations may induce some elderly to consume their assets too quickly and then outlive their means, which is less likely with the current mandatory indexed annuity provided by Social Security.
The political future is even more problematic. Under the current Social Security system, both the income adequacy and individual equity components are intertwined in one complex set of payroll tax and retirement benefit regulations. This single system has commanded broad political support. Under privatization, the antipoverty component would be explicitly separated from the rest of the system, where it could survive or wither, depending on the political climate ahead. Privatization itself would contribute to a less universalistic conception of the program. Critics of privatization fear that, once on its own, the minimum benefit could lose its political support, and the antipoverty gains of the past decades may be lost.
Sponsors of privatization argue that everyone will be better off, because of increased saving and the higher rates of return to be earned in private capital markets. But prospective retirees at the lower end of the income distribution had hoped to be the beneficiaries of a highly progressive Social Security benefit structure. Under the PSA plan, they would receive a smaller basic benefit, and despite the personal savings accounts, some will end up losers.
In the past, defenders of Social Security have relied on its universalism, bolstered by the fact that Social Security was a good financial deal for nearly everyone, despite its redistributive character. This was widely appealing as long as it lasted, but demographic changes are making the universalism harder to sustain.
It is inevitable that stabilizing the Social Security system will require some combination of tax increases and benefit delays or cuts, and all three plans acknowledge this. If the goal is simply to render the Social Security system solvent, the Ball plan does the trick, mainly by raising taxes. The Gramlich and the Schieber plans also stabilize the system's finances, by raising taxes and reducing benefits, but introduce a new philosophical twist-personal retirement accounts. These individual accounts are intriguing and should not be dismissed out of hand. They have the potential of increasing saving incentives, while giving people more control and responsibility for their own economic future. They may help rebuild young workers' confidence in what they believe to be an insolvent system by reducing the burden of the public promise in the future. Yet privatization, by explicitly separating income redistribution from earnings replacement, might significantly weaken Social Security's important antipoverty role.
