Social Security's Surprising Turn

Social Security is in dire straits and must be fixed now. How many times have we heard that refrain from government leaders and bipartisan commissions? To be sure, strengthening the Social Security system has become a national priority, driven by the common perception that the system cannot survive in its current form and that dramatic measures such as an increase in payroll taxes, a rollback in future benefit payments or the introduction of personal investment accounts are the only workable remedies.

It may come as something of a surprise, then, to learn that the Social Security program, which pays benefits to 47 million Americans, has actually shown significant financial improvement in recent years. How can that be in the face of so many doomsday forecasts? Consider:

• In their latest report, released in March, the Social Security trustees estimate that the system's trust funds (its pool of assets from which benefits are paid) will not be exhausted until 2042. This is 13 years later than the 2029 depletion date envisioned in the trustees' 1997 report. The Congressional Budget Office report on Social Security earlier this month showed a similar direction.

• In like fashion, the estimated year in which Social Security's annual benefit obligations begin to exceed its tax revenue has also receded, from 2012 in the 1997 trustees' report to 2018 in the most recent accounting.

• A crucial indicator known as the "actuarial balance" (the difference between the program's income and cost expressed as percentage of payroll) has fallen from a deficit of 2.23 percent in 1997 to 1.92 percent today.

• In 1997 it appeared that Social Security would go bust before the last of the baby boomers had retired. Now the youngest boomer will be 78 by the projected depletion date, which means the system as it stands will be able to cover the bulk of the baby boomers' benefits. In 1997 this was not expected to be the case.

What does this mean in terms of the long-range health of Social Security, once described by President Bush as "the single most successful program in American history"? Is the prognosis not nearly as bleak as the public has been led to believe?

The honest answer is that no one knows. After an increasingly pessimistic outlook for Social Security in the early 1990s -- which intensified calls for reform -- there has been a noticeable improvement in the past seven years. This trend can be traced in large part to favorable economic factors in which real wages grew significantly (reflecting the dramatic increase in productivity) and the unemployment rate declined significantly (to less than 5 percent). The cumulative effect was more and better paid workers remitting payroll taxes, hence greater surpluses for Social Security.

If the tide seems to have turned in Social Security's favor, no one can say with certainty whether the improvements are permanent. Indeed, looking at a relatively short seven-year period may lead to incorrect conclusions about the pattern of long-term changes in the actuarial projections. If we are truly interested in ensuring the future viability of a program on which so many millions of Americans depend, we must search for more fundamental answers based on a thorough analysis of the long-term assumptions around key demographic and economic indicators, and the uncertainty inherent in these assumptions.

One of those indicators is the anticipated reduction in the worker-to-beneficiary ratio -- the steamroller behind Social Security's projected deficits. (The lower the ratio, the fewer workers available to support the system and the heavier the burden each worker must bear.) The Social Security actuaries project a permanent drop in the number of workers per beneficiary from its current level of 3.3 to 2.0 by 2040, and ultimately to 1.9. Most of the decline takes place from 2013 to 2030 as baby boomers retire.

How reliable, though, can such long-term projections be given the historical variability in fertility rates and immigration levels, both of which have an enormous impact on the worker-to-beneficiary ratio? While birthrates showed significant variation in the 20th century, Social Security projections are based on a conservative, "below-replacement" rate over the next 75 years. Likewise, the level of immigration assumed by the government over that period is well below current levels. Given that immigration is in large part under government control, any program that increases immigration levels should increase the worker-to-beneficiary ratio, helping to offset projected Social Security deficits.

Complex though these issues are, they must be at the heart of any national debate over the future viability of Social Security. The recent strengthening of the system's financial vital signs is encouraging, but it should also serve to reinforce the inherent uncertainty of long-term actuarial projections and encourage vigilance to keep Social Security sound. It's in this context that the modest changes Federal Reserve Board Chairman Alan Greenspan recently suggested to Congress -- to index the age at which retirement benefits are paid and rethink the inflation measure used to adjust those payments -- should be seriously considered. Before we make any dramatic changes to forestall an impending Social Security crisis, let's make sure we know just how broke the existing program really is.

The writer is a retired partner of Goldman Sachs and a senior fellow at Harvard University's Kennedy School of Government. He was assistant Treasury secretary under President Ronald Reagan.