TCS Daily

The Trillion-Dollar Question

By William Sterling - December 29, 2004 12:00 AM

If President Bush pushes ahead with his proposal to privatize some portion of Social Security, politicians will almost certainly be forced to engage in a heated debate about a fairly obscure and technical topic: should Social Security's initial benefits be tied to future changes in consumer prices or wages?

This sounds like an issue that would matter only to economic policy wonks. But literally trillions of dollars are at stake in resolving the issue so a spirited debate is warranted. Following a controversial act of Congress in 1977, the current system has generously tied future benefits to changes in overall wages. In contrast, The President's Commission to Strengthen Social Security proposes to tie future benefits to the consumer price index. It would use the substantial cost savings to assure the solvency of the system and to improve poverty protection for the lowest-income workers beyond what is provided by the current system.

Wage indexation of benefits is naturally more expensive to the government than price indexation because wages tend to rise more rapidly than prices over time. An ongoing rise in real wages is a welcome consequence of overall growth in productivity. If firms are able to produce more widgets per hour thanks to new technologies or better business processes, they can afford to pay more to workers and still earn healthy profits. The most widely accepted models of economic growth, which will be considered shortly, suggest that the observed link between real wage growth and productivity is far from accidental.

Price indexation should save the government trillions of dollars in the long run because of the magic of compound interest. Wage growth has historically outpaced consumer price inflation by about 1.1% per annum. That means that the real value of wage-indexed benefits will be twice as high as price-indexed benefits after 70 years. And that is why a seemingly modest technical change in the benefits formula is a potential "magic bullet" for putting Social Security on a sound financial footing.

The debate about whether to move toward price indexation will have many facets that raise thorny philosophical and technical issues. Proponents of the current system argue that using the wage index is essential for two reasons. First, wage indexation ties future benefits to the general rise in living standards. Secondly, it allows workers to participate in the economy's productivity growth that they contributed to during their careers

Note that wage indexation ties future benefits not to rises in the "cost of living," but to productivity-driven rises in overall "living standards." In that sense, the current system does not just attempt to provide future retirees with a certain absolute standard of living; it also attempts to guarantee a relative standard of living, measured against a tide of rising real wages over time. Wage indexation ties future benefits not just to overall inflation but also to rising real wages and productivity. That essentially guarantees workers a pro-rata share of the economy's future output. It is the open-ended nature of that commitment that makes wage-indexed benefits so difficult for the government to afford.

Aside from the philosophical question of whether the government should be in the business of guaranteeing relative living standards, policy makers need to consider a key technical question as well: if the public demands a system that links future benefits to relative living standards, how can it be funded on a fiscally sound basis? This question is fundamental to the debate over both wage indexation and the proposed move to private accounts.

When the question is framed this way, economic theory suggests that investing some portion of Social Security taxes into the stock market makes eminent good sense. Whether individuals should be permitted to do so in private accounts or whether the government should do it on their behalf is another question that needs to be considered separately.

To think about issues like this, economists invariably fall back on models of long-term economic growth put forward in the 1950s by economists like Robert Solow and Trevor Swan. These models argue that in the long run, or "steady state," real wages must grow in line with productivity. If that were not the case, then virtually all of the economy's output would eventually go to either workers or to the owners of capital -- and neither of those outcomes is consistent with sustainable growth. So economists can make a strong theoretical case, supported by a good deal of hard evidence, that the share of output going to workers and the owners of capital must be relatively constant in the long run.

This insight from growth theory has profound implications for the wage indexation debate. If long-run economic stability dictates that the shares of output going to workers and to the owners of capital are relatively constant, how can the government guarantee that it will provide a pro-rata share of Gross Domestic Product (GDP) to retired workers? Only by guaranteeing to extract an equivalent share of GDP from the owners of capital!

Accordingly, indexing future retirement benefits to wages is analytically equivalent to indexing them to the overall returns experienced by the owners of capital. Those returns will basically depend on a weighted average of the returns to "risk-free" short-term government bonds and riskier assets like long-term government bonds as well as stocks and bonds issued by private companies. Under a system of entrepreneurial capitalism, a good portion of the return to owners of capital is presumably compensation for undertaking risky ventures -- i.e., the ventures that generate productivity gains for both individual companies and the overall economy.

From a balance sheet perspective, the government faces a massive quandary if it tries to guarantee relative living standards via wage indexation while funding the program with less risky investments like short- or intermediate-term government bonds. In doing so, its financial structure essentially becomes that of a doomed hedge fund -- perpetually "short" the stock market and "long" low-yielding government bonds. If stocks continue to provide higher returns than bonds over the next 100 years -- as has been the case over virtually all long-term historical periods -- that financial structure is almost certain to fail.

In a move that may have Karl Marx spinning in his grave, George Bush's vision of an "ownership society" would secure the Social Security system by explicitly making workers the owners of capital. The catch is that only by accepting some degree of high-risk stock market investments will workers receive the benefits of higher returns.

Of course, if standard economic growth theory is correct, the current system of wage indexation means that workers are already implicitly the owners of capital because they are effectively guaranteed a fixed share of GDP in the future. But instead of facing the risk of a market collapse, they instead face the risk of a fiscal collapse -- or a drastic cutback in future benefits -- because the system is not fiscally sound.

Indeed, the controversy surrounding the enactment of wage indexation in 1977 was not due to fundamental objections to the idea of indexing initial benefits to the cost of living. The idea of indexing benefits to the cost of living enjoyed broad bipartisan support then, as it does now. Rather, the controversy was due to the fact that Congress enacted wage indexation despite the findings of its own special commission that clearly rejected wage indexation as unaffordable in the long run.

As is often the case, economics tends to reinforce one basic point: there's no such thing as a free lunch. It remains to be seen how ardently President Bush will pursue his vision of an ownership society. But he has already performed a valuable service by prompting a genuine debate about the sustainability of current guarantees to future retirees

William Sterling is Chief Investment Officer of Trilogy Advisors. The opinions expressed in this article reflect his personal views and do not represent the view of his firm.


TCS Daily Archives