Commentary: Social Security: The Elderly Should Bear Some Risk

Allocating risk fairly between young and old is crucial for a viable benefits policy

By Michael J. Mandel

The clouds of confusion surrounding the Social Security debate are getting thicker by the minute. The latest piece of news to roll in: a Congressional Budget Office study showing that the federal budget surplus will be only $153 billion this year, down from the $275 billion projected in May. As a result, some $9 billion in Social Security revenues will have to be diverted this year to pay for general expenses.

The estimates, released on Aug. 28, ignited a political firestorm. Democrats immediately geared up to attack the Administration for raiding the retirement trust funds. Meanwhile, in an odd reversal, Republicans are defending the diversion as a necessary evil in an economic slowdown.

But both Democrats and Republicans are missing the bigger point. If we can't reliably forecast the surplus four months in advance, how can we know how much is needed in the trust funds to pay for retirement and medical costs 30 or more years in the future?

STAGGERING. The simple answer is, we can't: As policymakers attempt to gaze into the future, they are confronted not merely with enormous economic uncertainties but also with staggering questions regarding medical advances. No one knows how fast productivity will grow in the years to come. And there's no way of predicting how much new medical technologies will extend life spans, or how much these advances will cost. Depending on the assumptions you make, the trust funds will start running out of money in 15 years, 30 years, or never.

Faced with this unknown, it's pointless to worry about $9 billion today. Instead, we should be debating the key question: How much of the future uncertainty should be shouldered by the elderly, and how much should be borne by younger workers? Finding a fair allocation of risk between young and old is essential for making Social Security and Medicare politically and fiscally viable.

Certainly there are a wide range of alternatives. The current system of old-age entitlements, for example, effectively protects the elderly from many risks, such as an inflationary surge or an economic slowdown, after they have retired. But it also means the elderly don't share in the gains when the economy booms.

By contrast, fully privatizing the Social Security and Medicare systems, as some have suggested, could boost returns if the economy and stock market did well--but it would also shift much of the risk to the elderly. They would be exposed to the turmoil of the marketplace, including possible financial loss and escalating medical costs.

It could be possible to find a politically palatable middle ground, where the elderly bear a piece of the economic risk, rather than being relatively immune as they are now. That would mean making old-age benefits more sensitive to economic conditions while offering retirees some protection.

In the case of Social Security, for example, that could mean tweaking the payout formula. Right now, benefits are set by a complex formula when a person first retires, and then are adjusted for inflation using the consumer price index. Instead, benefits could be raised each year by an index based on wages. That means the elderly will see their real benefits go up faster whenever real wages do--but if real wages stagnate or fall, as they did in the 1980s, so will Social Security payouts. Similarly, Medicare spending could be better linked to the ability of the economy to support it.

Politically these won't be easy decisions to make. Yet past attempts to find a permanent solution to Social Security and Medicare funding have consistently fallen short because the future came out different than expected. For example, increases in Social Security taxes enacted in 1977 were expected to "restore the financial soundness of the cash benefit program throughout the remainder of this century and into the early years of the next one," in the words of the Social Security Trustees. But the trust funds quickly came close to running dry because of the unexpectedly high inflation and deep recession of the next few years, forcing another rescue attempt in 1983.

VERY WRONG. More broadly, projections of actuarial soundness based on forecasts of productivity growth run the risk of being very wrong. Since the end of World War II, average long-term productivity growth has fluctuated widely, going as high as 3% annually and as low as 1.1%. Faster productivity growth means a bigger economy--and that's what ultimately determines our ability to support the elderly. That's true whether or not benefits are paid out of trust funds, general tax revenues, or private accounts.

The future length of Americans' life spans is also fundamentally unknowable. Past demographic projections of the number of elderly have been way off the mark. For example, in the mid-1960s, the Census Bureau projected that there would be 28 million people 65 years and older in 2000. In fact, they underestimated by 25%; the actual number was 35 million. Now medical scientists are working on a wave of fundamental technological advances, such as stem cell research, that could lead to big increases in life expectancy. Or life spans could turn out to be hard to extend past the biblical three score years and 10.

Risk is not the easiest thing for the political system to deal with. If the economy falters over the next 30 years, old-age benefits will have to be cut--there's no way around it. But only by acknowledging the uncertainty can we make policy that makes sense.

Mandel covers economic policy from New York.

    Before it's here, it's on the Bloomberg Terminal.