The Future of Social Security

Henry J. Aaron and
Henry J. Aaron The Bruce and Virginia MacLaury Chair, Senior Fellow Emeritus - Economic Studies

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

April 19, 1998

Social Security has been a huge success. It provides benefits to 44 million Americans who are elderly, disabled or survivors of deceased workers. More than three-fifths of the elderly derive the majority of their income from Social Security. Without the program’s benefits, half of the elderly would live in poverty. Social Security also provides families of active workers with a form of life insurance worth more than $12 trillion—more than all private life insurance currently in force.

Its accomplishments notwithstanding, the current system has to change. The number of beneficiaries will double over the next four decades while the number of workers who pay the taxes that support the system will grow by only 17 percent. Although the program’s receipts now exceed its expenditures by more than $100 billion a year and it is adequately funded for the next three decades, the revenue will cover only 70 percent to 75 percent of promised benefits after 2029.

As policy makers look for ways to close the projected long-term deficit, some are suggesting that Social Security should be privatized—that is, replaced with mandatory saving in individually owned accounts. With the stock market soaring and many Americans gaining experience with mutual funds, individual retirement accounts and 401(k) plans, tying basic retirement benefits to the performance of mandated personal retirement accounts, look increasingly attractive. Such accounts would let future workers invest some or all of the money they now pay to Social Security in the private markets—allowing them, it is claimed, to earn higher returns than Social Security can offer.

But before we succumb to “bull-market frenzy,” everyone should insist on answers to three questions: Would the returns on personal accounts actually exceed what a reformed Social Security could deliver? Would benefits under a privatized system be safe? Would these privatized alternatives reliably fulfill the current program’s crucial social functions?

The answer to all three questions is no. Privatization is not the answer because Social Security, with changes that would leave its basic structure intact, can provide workers with higher returns than could any privatized alternative and can provide these returns with much less risk while continuing to advance important social objectives that privatized alternatives would jeopardize.

Social Security was created in 1935 to provide all American families with a dependable income base upon which they could build additional protection against income loss for themselves and their families after retirement, disability or the death of a breadwinner.

To this end, Social Security replaces a larger share of lost income for low earners than for high earners, grants more generous survivors’ benefits to larger than to smaller families, provides extra resources for retired couples in which one spouse has had no or limited earnings, and gives special consideration to divorced people whose marriages lasted at least 10 years.

In part because of these provisions, Social Security lifts twice as many people out of poverty than all other income-tested assistance programs, cash and in-kind, combined.

Under a privatized system, in which each participant’s benefit would depend on the accumulations in his or her individual account, there is no room for such social assistance. That burden would have to be borne by a separate program, possibly one requiring a demeaning means test.

Social Security provides a secure and predictable financial guarantee by tying benefits to the average wages workers have earned over their lifetimes. In privatized systems, however, benefits ride the financial market roller coaster.

A drop in asset value just before a worker reaches retirement or becomes disabled can decimate benefits. Some may dismiss the post-1929 stock market crash as ancient history and say that the 60-percent swoon in Japan’s Nikkei average between early 1990 and 1995 “couldn’t happen here.” Our colleague Gary Burtless has estimated that under a privatized retirement system, in which all workers were required to invest a fixed percentage of the yearly earnings in common stocks, the ratios of benefits to past earnings of those retiring in the late 1970s (after the sag in U.S. stock prices) would have been less than half those available to workers who retired at the start of that decade.

Market volatility may be an annoyance to the wealthy, but it is potentially catastrophic for the majority of Americans who have few financial assets other than Social Security.

Advocates of private accounts admit that they may be riskier but argue that the risk is worth bearing because the returns on average are so much higher. But if the balances in all personal accounts were invested in stocks and bonds, the accounts taken together would have to earn about the average return paid on stocks and bonds. There is no financial Lake Wobegon where all, or even most, of the returns are above average!

Of course, some investors would beat the averages—most likely those with market sophistication or who can afford to buy expert financial advice. Others would do poorly because they invested too conservatively or accepted some of the endless supply of bad financial advice that is available.

If Social Security accumulated reserves of a similar size and was freed to invest in private assets, however, it would earn even higher returns than the average private account. The reason for this is straightforward: Brokerage fees for buying and selling stocks and bonds and for sales and management fees charged by mutual funds chomp away at the returns of private accounts. On average, these expenses are likely to take 1 percent to 2 percent out of accounts’ balances each year before retirement. Such charges would reduce, by 20 percent to 40 percent, the amounts the workers would ultimately accumulate in their accounts.

If Social Security hired private managers to invest its large reserves in private assets, the costs would be negligible, less than one-hundredth of a percent of the funds under management each year—leaving more of the total return to support pensions.

An even bigger bite would await owners of private accounts who upon reaching retirement age wanted to convert their private account balances into annuities that would guarantee them a fixed yearly income until they die. Privately purchased annuities are expensive. Many financial institutions that sell these contracts charge fees that average about 10 percent of the amount invested to cover sales costs, account maintenance and company profits, and an additional 10 percent to allow for the fact that people with long life expectancies are more likely to buy annuities than are people with shorter life expectancies. Faced with such charges, many retirees might choose to withdraw funds from their accounts periodically, thereby running the risks of either outliving their assets or, fearing that fate, depriving themselves needlessly.

A 20-percent bite out of what workers have saved as the price for an annuity is pointless and wasteful, since Social Security can convert savings into annuities at a negligible cost.

Not all privatization plans would generate such large administrative costs. But all would be more costly to run than Social Security and would therefore pay out smaller and riskier benefits than under a reformed Social Security program. And all would make important social objectives more difficult to achieve than they are under the current system.

So what needs to be done to fix Social Security? First, coverage should be made truly universal by covering all newly hired state and local government workers, one-quarter of whom are now outside the system. Extending coverage would provide additional protections to these workers and help Social Security’s finances. Furthermore, benefits should be treated like other retirement income by subjecting them to the same income-tax rules that apply to private pensions, that is by taxing benefits that exceed what the worker has contributed. These tax revenues could be credited to the trust fund to help finance future benefits.

Second, benefits should be slightly reduced by increasing the number of years of earnings averaged to compute a workers’ benefit and by accelerating scheduled increases in the normal retirement age. In combination with announced corrections in the Consumer Price Index, which would lower annual inflation adjustments to benefits, the changes listed so far would close two-thirds of the projected long-run deficit.

Third, the requirement that Social Security reserves be invested in relatively low-yielding Treasury securities should be scrapped and the trustees empowered to invest in a diversified portfolio that includes private as well as government assets. The responsibility for managing these funds should be transferred to a new, quasi-private agency modeled on the Federal Reserve. The chair and members, appointed for lengthy, staggered terms, should be required to invest reserves only in broad index funds and to make sure that shares were voted solely to reflect the economic interest of participants.

This structure would save money and make possible higher average returns for beneficiaries than private accounts could offer.

It would also ensure that Social Security reserves could not be used as an instrument of political control over private business decisions.

And moving the reserves to a quasi-private entity would help guard against the possibility that growing fund surpluses would be used to justify tax cuts or spending increases.

The foregoing menu of changes would fully close the projected long-term deficit in Social Security and for most people, generate higher and more reliable pensions than would private accounts. Of course, other measures could be substituted or added to further strengthen the long-run financing position of the program or to adjust the structure of benefits. The normal retirement age could be raised as life expectancy rises. The payroll tax could be hiked by perhaps 0.5 percentage points each on workers and employers sometime after the baby boomers are fully retired. The age at which benefits are first available could be raised gradually from 62 to 64 to encourage later retirement, particularly if the requirements for disability benefits are somewhat eased for this age group. Spouses’ benefits might be lowered in exchange for increased benefits to elderly widows and widowers.

This menu shows that fixing the financial problems of Social Security is not rocket science. There are many ways to do it that don’t require intergalactic thrust. As the Social Security debate continues, it is important to realize that privatization is not a free lunch. Taxes have to be raised, benefits for current retirees and older workers cut, or both. The unpleasant reality is that the current payroll taxes, 80 percent of which are needed to pay current benefits, do not generate enough money to fund meaningful deposits into private accounts unless benefits are slashed deeply.

If the advocates of privatizing Social Security were making valid claims, it might be worth paying those higher taxes. But it is hard to see why American workers should be asked to fork over more for a new system that would deliver lower average benefits for each tax dollar they pay, that would subject workers to financial risks they are ill-equipped to bear, and that would place in jeopardy the social assistance on which millions of Americans depend.

We need to tune up Social Security, not trade it in for a new, but flawed, model.

The authors are senior fellows at the Brookings Institution and co-authors of a forthcoming 20th Century Fund-Brookings Institution guide to the debate on Social Security reform.