A Hidden Fiscal Crisis

• Author: Laurence J. Kotlikoff
• Position: Professor of Economics at Boston University.

EVEN as the United States experiences continuing fallout from a terrible financial crisis, a more alarming fiscal problem looms. The world’s largest economy faces a daunting combination of high and rising costs for health care and pension benefits and constrained sources of revenue that will put enormous pressure on its fiscal soundness.

So far, the markets seem to be focusing on U.S. official government debt relative to its gross domestic product (GDP). That number stands at 60 percent, roughly half that, say, of beleaguered Greece. Consequently, the financial wolves are circling Greece, not the United States—driving up yields on Greek securities and driving down yields on U.S. treasury securities.

But the debt-to-GDP ratio is not a useful guide to a country’s true fiscal position. Because of something economists call the labeling problem, every dollar a government takes in and pays out can be labeled in an economically arbitrary manner. So what is reported as the size of a deficit or surplus is independent of a country’s actual underlying fiscal policy (see box).

What’s in a name?

The labeling problem attached to government deficits is a matter of theory, not simply practice. Consider the equations of any economic model with rational agents—that is, agents who pay no attention to language and instead make decisions based on fundamentals. Whether these equations are talked about by French, English, or Chinese speakers will not affect the model’s behavior, which is dictated by the math.

Attaching particular fiscal labels to a model’s variables is simply a matter of choosing an internally consistent language to discuss the equations. But each internally consistent labeling choice produces a different measure of the debt and its changes over time—the deficit.

In a recent paper, Jerry Green and I referred to the labeling problem as “the general relativity of fiscal language,” to emphasize that in economics, as in physics, certain concepts aren’t well defined (see Green and Kotlikoff, 2009). Time and distance aren’t well defined in physics, and government debt and the deficit aren’t well defined in economics. Nor for that matter are taxes, transfer payments, private net wealth, disposable income, private savings, and personal savings.

All deficit accounting, then, is inherently arbitrary. Substituting one set of arbitrary fiscal labels for another will not tell us anything worth knowing if we continue to act as if government debt measures a fiscal fundamental rather than what it really does: reflect our nomenclature.

All in a name

For example, take payroll taxes targeted to pay future pension and health care benefits in the United States. These receipts, now labeled taxes, could just as well be labeled borrowing. And the future benefits could be called repayment (with interest) on this borrowing (minus a future tax if the benefits fall short of principal plus interest). This alternative—but no less natural—language describes the same underlying reality: taxes are much lower and the projected 2010 deficit is 15 percent, not 9 percent, of GDP.

The Chilean pension “reform” of the early 1980s illustrates the arbitrary nature of fiscal labels. The reform funneled receipts, which had been called payroll taxes, into private pension funds, which the government then borrowed to cover pension payments. The same money was still flowing from workers to retirees, but was called borrowing.

If the standard debt-to-GDP ratio fails to measure a country’s long-term fiscal prospects, what does? The answer is the fiscal gap, whose value is the same no matter which labeling convention a country adopts. The size of the U.S. fiscal gap, as recently measured by the IMF...