The Future of Social Security

An Address By Holley Hewitt Ulbrich, Ph.D.
1998

In 1995, the Social Security system took in revenue of $452 billion, including $31 billion in interest, and paid benefits (including administrative costs) of $352 billion, for a net increase in assets of $72 billion, which was added to the accumulated balance of $550 billion. The system covers 133 million workers and serves 50.8 million beneficiaries, of whom 37.8 million are getting retirement and survivors benefits. This is big business. It deserves careful attention. Most of that attention has been focused on these projections:

In the year 2019 benefit payments will begin to exceed current taxes and interest.

By 2029 the accumulated surplus will be exhausted.

As are any economic projections, these are based on assumptions about wage growth (1% a year adjusted for inflation), fertility, longevity, interest rates, marital stability, immigration, labor force participation (especially by women), and unemployment. Robert Kuttner1 points out that if wages were to grow at their average rate of the last 50 years (1.7%), then the doomsday scenario is replaced by one that envisions the system in the black for the indefinite future. Other assumptions, likewise, can be wrong because projecting for so long a period for so large a system is very sensitive to even minor adjustments in the assumptions.

Discussions of Social Security often get bogged down in issues relating to Medicare and SSI (Supplementary Security Income), both of which are important, but for the great mass of working Americans, the primary interest right now lies in the old age, survivors, and disability program (OASDI) put in place between 1935 and 1950. This core business of Social Security is an intergenerational and interpersonal compact, reflecting both individual and communal values. Every industrial country has something similar except Japan, which relies on private savings and families. Our design reflects the values/context of the 1930s: individualism, earned benefits, some redistribution, and married women not working outside the home. OASDI is a hybrid of insurance, pension, and redistribution programs. Hybrids can be the best of both parents or the worst, sometimes both. Consider the mule, which is stubborn, sturdy, and sterile!

The insurance component consists of survivors (1939) and disability (1950) programs. Disability is relatively new and much more complex than old age and survivors programs, especially in the establishment of eligibility. Recent publicity has focused on two issues: children getting disability payments (which are SSI, not OASDI), and payments to people disabled because they are substance abusers, which is currently under review.

The pension/annuity component is the earliest, and the one we think of when we say Social Security. If you return to the original discussions when this program was being designed you will be reminded of something current critics gloss over. This system was designed by the middle class to defend itself from the poor. Knowing that the middle and upper income groups were going to get stuck with the bill supporting the elderly poor in their declining years, because this group was not likely to save for old age, a purpose of the program was to force the working poor to contribute something while they were working. Thus, an element of the design was protecting the middle class and the rich from the poor!

Having said that, however, we must acknowledge that Social Security has a strong redistribution component (a much higher ratio of benefits to earnings for lower income earners). The lowest income recipients, or those not covered, are shunted off to SSI, which replaced Aid to the Aged, some disability, and General Assistance in the 1970s. This program is often confused with Social Security but is in fact separate, funded by general tax revenues.

An essential feature of the design was broad coverage, a breadth that has continued to grow as such groups as state and local employees, ministers, self-employed persons, farm workers and other groups were added to the system over the years. Today there is close to universal participation among those employed, a feature that results in two positive benefits -- low administrative costs (about 1.5% of benefits paid), and no adverse selection. Those who expect to live to a ripe and healthy old age share the system with those who expect to die young or to be disabled before reaching retirement age.

Other important design features include the wage ceiling or maximum amount subject to Social Security taxes ($61,200 in 1995, adjusted annually, the tax rate (currently 6.1% by employer and employee for OASDI and another 1.49% each for Medicare), the coverage of wages only (excluding other forms of income), the annual inflation adjustment, the loss of benefits for earning above a certain ceiling, and the marriage penalty for widows and widowers. This last feature is one of several problems related to the differential treatment of wives of retired or disabled workers and women workers, with an increasing proportion of women being eligible for either (but not both) types of benefits.

Many of the concerns about Social Security can be described as equity issues, relating to interpersonal equity and intergenerational equity.

Interpersonal equity is often measured as the ratio of lifetime taxes to lifetime benefits. These calculations, which appear in the popular press, are not easy to make; there is no typical earner/recipient, and each person's calculation depends on how long you live, what your earnings pattern was, and other factors. A recent issue of the Social Security Bulletin reviewed and critiqued various money's worth studies that looked at such measures as payback period, benefit/tax ratio, lifetime transfer, and internal rate of return. It is important, in making such comparisons, to compare apples with apples, considering among other issues the level of risk assumed in the investment mix (it is very low for Social Security Trust Fund investments), and being sure to include administrative costs (also very low for Social Security)2.

Within age cohorts (people born during the same period), it is estimated that the return is better for couples, women, minorities and the poor, because women and couples live longer and lower income groups (including a disproportionate share of minorities) receive a higher ratio of benefits to earnings. The present value of taxes less benefits is very positive for lowest income decile, remaining positive up to the middle of the income distribution, and is negative for the top half of the income distribution.

More attention has been paid recently to intergenerational equity; some critics describe its "pay-as-you-go" nature as a sort of Ponzi scheme, with the baby bust generation (1965 onward) left holding the bag. Again, looking at age cohorts, researchers found that the earliest cohort examined (1895-1903) earned on average a 12.5% inflation-adjusted return; the 1917-22 cohort received 5.9%, while the babies born in 1995 are projected to receive a 1.5% inflation-adjusted rate of return. All of these figures compare favorably to a long- term 0.6% inflation-adjusted return on government bonds.3

One important consequence of Social Security has been reduced poverty among the elderly. More than half of those current elderly receive more than half their income from Social Security . Among current workers, about 1/2 are covered by a pension plan; the others, apparently, are depending on Social Security to provide for their old age. What will happen to the benefits to the elderly and the rates paid by active workers down the road? Again, there is no simple answer: it depends on fertility, longevity, earnings growth, interest rates, marital stability, immigration, labor force participation, and unemployment.

There is no lack of proposals for reform. The following is a brief summary of some of the proposals currently on the table.

1. Delinking Social Security from the budget, so that the reported deficit does not count the net revenue of the trust funds against the operating deficit. This procedure, strongly advocated by Senator Hollings and others, makes balancing the budget much harder, since these revenues are currently covering about 1/3 of the deficit.
2. Capping the cost of living adjustment (COLA). There is considerable evidence that the Consumer Price Index currently used to adjust benefits is overstating inflation, and particularly the impact of inflation on the elderly, because of the strong role of housing and medical costs in the index. Many of the elderly are not affected by rising costs of buying new homes, as they live in homes that are paid for, and much of their medical expense is covered by Medicare. Proposals include a lower adjustment or a link to growth in wages rather than in prices.
3. Changing the assumptions. Congress and the President have argued this issue on projecting the budget deficit for a mere seven years; the projected problems in Social Security come in a period some 23 to 33 years hence, and are crucially dependent on assumptions about wage growth, fertility, labor force participation, and immigration. It only takes very small adjustments in some of these assumptions to make the system viable. Perhaps the cries of disaster are premature.
4. Adjusting the age for eligibility. With longer life spans, retirees are collecting much longer, and more are opting to collect reduced benefits at 62 rather than full benefits at age 65. A change scheduled in another five years beginning with the 1940 birth cohort, still permitting benefits at age 62 but gradually raising the age at which full benefits are available.
5. Reforming or delinking other components of the system, such as disability eligibility, and Medicare. Some argue that Social Security is bearing too many unrelated responsibilities, and should be pared back to its "core business"--a form of downsizing or reengineering. Generally, proponents of this reform are not arguing that those other components should be scrapped but that they involve other issues and should be separated and treated differently.
6. Treatment of employed v. not employed women. Since the payroll tax includes payments for survivor benefits, it can be argued that married couples pay twice, but can only collect once. Widows can collect either on their own earnings record or 80% of the benefit that would have been received by their deceased husbands; retired wives, likewise, can collect full benefits on their own records or 50% of their husbands' benefits. This issue will eventually self-correct as women being employed outside the home becomes the norm rather than the exception. In the interim, it is important not to penalize those who got caught in a values revolution.
7. Increasing the wage base. The wage ceiling has gone up, but the existence of a ceiling means that the fringe benefit cost doesn't continue to rise beyond a certain point. As a result, the tax is somewhat regressive. In fact, the Social Security payroll tax is biggest single tax burden for the working poor. Broadening the base would raise a more revenue, and make the tax more equitable.
8. Changing the investment mix. Some reformer would like to see at least part of the Trust Fund's $550 billion in assets invested in equities, for two reasons. First, the fund would experience greater growth, although at some cost in terms of risk and management expense. Second, this change would help to delink the trust fund from the budget deficit, since it would no longer be invested entirely in Treasury securities.
9. Making the system private and/or voluntary. Barry Goldwater suggested this idea in 1964, and never recovered. Today we see it in the Readers' Digest. Privatizing, while keeping participation mandatory, would create a need for monitoring and supervision, given the history problems with private pensions and with investment advisors. Making it voluntary will results in many of the poor opting out in order to provide for immediate consumption. Both would reduce or eliminate the insurance and redistribution functions of the system to function in its role as a pension provider. None of these problems are insuperable obstacles to making such a change, but they are real problems that would have to be addressed in some way.

   Like the former editor of The (Baltimore) Sun, I feel called upon to respond to those doubters and questioners of our age, but my correspondent is my 29-year-old daughter, who like Virginia O'Hanlon, has serious doubts about the future of this particular Santa. Yes, Carla, Social Security will be there when you turn 65 in 2031, because it is too politically costly to kill it. Benefits will be lower in relation to your income than they were for me and Dad, but it will be there, still financed largely by a dedicated tax on earnings of some kind, still favoring lower income workers over higher income workers. Part of the trust fund will probably be invested in the economy, in equities and corporate bonds, to provide diversity and growth and to delink SS from the broader question of the budget. The people paying into the system will be different: more women, more minorities, more recent immigrants. Even if you do get married, when you retire you will probably be eligible only on your own account: the few surviving dependent spouses will have been shifted to SSI, and the survivors' component of the program will have diminished, in order to focus on the core businesses of disability and retirement. But we as a people are committed to ensuring that our elderly do not have to move in with their children, die in dire poverty, or exhaust their limited personal assets to survive in old age, that there will be an opportunity to enjoy a little gold in the golden years. We do that in our own self-interest, looking after our elderly in hopes that the torch will pass to the next generation who will look after us. I think you can count on that mix of altruism and self-interest to ensure that there will be some provision, in some form, for you when your turn comes.

   Footnotes

   1Robert Kuttner, "Social Security: If It Ain't Broke, Don't Tinker," Business Week, February 22, 1995, p. 22.

   2Dean R. Leimer, "A Guide to Social Security's Money's Worth Issues," Social Security Bulletin, Summer 1995, p. 3-14.

   3David Pattison, "The Distribution of OASDI Taxes and Benefits by Income Decile," Social Security Bulletin, Spring 1995, p. 21-32.

   4Ibid.

   Dr. Ulbrich is Alumni Professor of Economics at Clemson University and Coordinator of the Community and Economic Development Program at the Strom Thurmond Insitute of Government and Public Affairs at Clemson.

This page is maintained by Thomas Rourke The person responsible for this web site server is Patrick Harris ©1998 Strom Thurmond Institute