Social Security: A Social Insurance System

Laura D'Andrea Tyson

Council on Foreign Relations

November 15-16, 1996

Social Security has been described as the crown jewel of American federal government programs. It is widely recognized to be the major reason why the poverty rate among the elderly in the United States has fallen in half since 1959 and is lower today than the poverty rate for any other population group as a whole. Fifteen million older Americans are kept out of poverty by Social Security.

For the elderly with incomes in the lowest 40 percent of the income distribution, Social Security benefits account for over 80 percent of family income. Even for those elderly in the upper fourth quintile of the income distribution, Social Security benefits make up almost 50 percent of all family income. It is only for the top 20 percent of the elderly income group that Social Security benefits are a small portion-that is, about a quarter of total income. Most of the people at this meeting represent the minority of the population in the sense that they will not rely primarily upon Social Security as most elderly Americans do. In addition, 3 million children, mostly children of deceased or disabled workers, and over 4 million disabled adults in the United States are on the disability or survivors' part of Social Security.

Today there is concern that the United States, like the other advanced industrial countries, faces tremendous challenges in financing its public retirement system as a result of changing demographics. Such concern is warranted. On the other hand, too often in these discussions such challenges are converted to crises(from financing problems to bankruptcy. The reality is different, and we should keep reality in mind when trying to make public policy recommendations.

When considering the financial challenges facing Social Security, three facts should frame the discussion.

First, Social Security has sufficient resources to pay 100 percent of its promised benefits through the year 2029. Between now and 2012, the Social Security system is expected to raise more revenue than it pays out in benefits. Between 2013 and 2019, tax revenues plus interest on the trust fund balances are projected to produce enough resources to cover all benefits. After 2019, trust fund assets (government bonds) would have to be liquidated to pay expected benefits. But even with exhaustion of the trust fund in 2029, and with no changes, the Social Security system is projected to raise sufficient revenues to pay about three-quarters of promised benefits in subsequent years.

Second, it is certainly true that Social Security is running a deficit over a 75-year period, but the range of actions necessary to eliminate that deficit is well within the bounds of previous changes to the program. An illustrative statistic makes this point. Under the intermediate assumptions of the trustees of the Social Security system, the 75-year deficit equals about 2.19 percent of taxable payrolls. By comparison, between 1980 and 1990--the period covered by the Reagan and Bush administrations--the Old Age Survivors and Disability portion of payroll taxes increased by 2.24 percent, from 10.16 percent to 12.4 percent. This is a slightly larger increase than the 2.19 percent deficit of the system projected over a 75-year period. This comparison is not meant to imply that we should solve the future financing shortfall by a further increase in the payroll tax. It is meant to demonstrate that we are not looking at a sizable change over a 75-year period that our system cannot absorb.

Third, the Social Security system is clearly much better off than the retirement systems of the other advanced industrialized countries. A smaller share of the population is expected to be 65 years or older in the United States by 2010: the figure is 13.6 percent, compared to 21 percent in Japan, 20 percent in German, and 17 percent in the United Kingdom. We all have aging populations, but our elderly population accounts for a slightly smaller share of our total population projection. The standard retirement age is also lower in many European countries than in the United States, and the retirement systems in many other countries are much more generous than in the United States, permitting workers to begin drawing public retirement benefits at age 55, for example, compared to age 62 in the United States. And, finally, the U.S. system has a trust fund with a significant amount of assets; in contrast, the public pension systems in many other countries are run almost completely on a pure pay-as-you-go basis. The existence of a trust fund in the United States raises the possibility of diversifying the fund's asset holdings in order to increase its rate of return.

In thinking about the challenges confronting Social Security, it is important to distinguish between its financing challenge and what is called its "money's-worth," or rate of return challenge. The financing challenge is simply the challenge of addressing the projected 75-year deficit of the system and doing so in such a way that the trust fund is not exhausted in year 76.

The "money's-worth" challenge facing Social Security has developed because calculations show that future cohorts of Social Security recipients will receive relatively small real returns on their contributions. The real return on Social Security contributions over one's working lifetime has been declining, and projections show it will be about 1 to 2 percent in real terms annually going forward. This projected behavior of the cohort-wide return depends on the growth rate for aggregate wages, the growth of productivity in the economy, and the fact that the system has matured toward a pay-as-you-go system.

With a declining real rate of return, some high-wage workers may actually see a negative real rate of return on their Social Security contributions. That's because the Social Security system is a redistributive system; that is, it pays higher relative benefits to lower income workers than to higher income workers.

Social Security, by design, is also a social insurance system. It insures by pooling risk. Each and every covered working American is insured against inadequate retirement income should he or she become disabled or lose employment. In contrast, the simple money's-worth calculation implies that those who are unlucky enough to suffer a loss are treated better than those who are fortunate enough to avoid a loss. This is a misleading interpretation: both are insured, but only the former may need to rely on the insurance. The money's-worth concerns, as they relate to high-income workers, miss the point of insurance and redistribution.

There are already books, papers, speeches, and monographs written on what we can do to address the financing challenges of Social Security. I'm only going to address two specific options quickly.

One option is an increase in the normal retirement age. Under current law, the retirement age is scheduled to increase gradually from age 65 to age 67 by 2022. Various proposals have been made to speed that transition, to increase the retirement age above age 67, and then to index it to increases in longevity. To evaluate these alternatives, a couple of things need to be kept in mind. First, some statistics suggest that there has actually been an increase in early retirement in the United States. Many people have been leaving the work force well before age 65. Some recent research suggests that this trend may change in the future, but a consensus has not yet developed. It will be easier to increase the retirement age for Social Security purposes if more people decide to work longer than if more people choose to retire early.

A second point to keep in mind is that early retirees fall into two broad categories. One group includes those who are relatively prosperous--they have some wealth, they have employer-provided pension income, and they decide to retire. Another group of people retire early because they have irregular work histories, injury, poor health, or a job that requires significant physical exertion over their working lifetime. An increase in the retirement age will not affect these two pools of people in the same way. Moreover, these groups may have very different incomes as they approach retirement. Therefore, if we're going to consider changes to the retirement age, we're going to have to think about whether they would apply across-the-board or whether we should make distinctions for different kinds of work histories, different kinds of income levels, different kinds of individuals.

Another popular option for addressing the financing challenge of Social Security is the diversification of assets of the Social Security Trust Fund. The diversification argument notes that the annual real rate of return on equities has been about 7 percent over the long run, while the real rate of return on bonds has averaged about 2.5 percent. We could do better, this line of argumentation goes, if a portion of the Social Security Trust Fund were diversified into equities.

There are, however, some possible drawbacks associated with this approach. First, the higher return on assets held by the trust fund would be matched by lower returns earned by private investors, who would end up holding Treasury bonds in place of the equities they once held. More equities in the Social Security Trust Fund would mean that more bonds would be held somewhere in the system.

Second, the return on Treasury bonds would be bid up to entice private investors to hold the Treasury bonds released by the trust fund. The result would be higher interest costs for the government.

Third, the shift in bond and equity demand would narrow the equity premium over time, and this premium has already been eroding over the past few decades.

Fourth, there are some technical issues. The Social Security Trust Fund would eventually accumulate sizable positions in the stock market, even for a small share of trust fund assets, raising issues of public interference in private-sector activity and corporate governance. How would the government behave as a very large holder of corporate equity? Moreover, the timing of the trust fund equity purchases and sales could coincide with the savings and dissavings of the baby boomers, possibly exacerbating long-run movement in the price of equities.

Privatization proposals, as distinct from diversification proposals, would shift the burden and responsibility of investment decisions to the individual and, therefore, would also shift the risk of fluctuating investment returns to the individual. Benefits would depend on the investment selected, and the timing of the individual investment, retirement and/or disability. And while individuals would reap the investment returns, they would also pay the transactions cost. Private annuity markets would have to evolve substantially in order to provide retirees with a financial product approximating the current inflation-adjusted Social Security benefit annuity on an actuarially fair basis. This product does not currently exist in an affordable form.

Privatization proposals, in contrast to diversification proposals, move us away from the underlying principles of Social Security as it was originally designed. As I emphasized earlier, Social Security is a social insurance system that pools risk for everyone and redistributes income from those who are relatively well off to those who are less well off. Privatization would allow individuals to opt out of the system, curtailing the pooling of risk.

Proponents of privatization claim that it would increase the national savings rate. But there are other ways to achieve this goal without undermining the Social Security system. We should look first to reducing the near-term fiscal deficit, which we are attempting to do by getting the federal budget into balance by 2002. And we should do what we can to increase private savings or pensions in addition to, not in lieu of, the Social Security System.

More than 50 percent of all working persons in the United States are not earning any private-sector pensions. Many of these people work for small businesses, where 75 percent of workers have no private-sector pensions. Private pensions and Social Security benefits together replace 60 percent or more of pre-retirement income for only 50 percent of today's retirees.

There's no reason to think this situation will get better all by itself. Fifty percent of current retirees, therefore, need to supplement in some fashion their Social Security benefits in order to maintain their living standards. Unfortunately, many of those Americans have saved little or nothing for their retirement. So, clearly, we have a pension problem, a pension challenge. I would like to see us do the tough things that need to be done to address the deficit, to address the financing problems of Social Security while maintaining its basic premise, and to devise methods to encourage Americans to save more for their retirement(recognizing, I'm afraid, that economists know very little about what to do to solve the last of these three important but distinct challenges.