By Laurence J. Kotlikoff
The population of wealthy countries is getting much older. Between 2005 and 2035, the number of elderly in wealthy countries will more than double, but the number of workers will barely change. This historically unprecedented demographic change portends enormous fiscal stresses because of the high and growing cost of meeting government pension and health-care commitments to the elderly. Indeed, these projected payments are so high that collecting them may not be feasible, either economically or politically. The costs associated with the coming generational storm will bankrupt the governments of most wealthy countries unless major and painful adjustments are made now.
“Bankrupt” is a strong word, but entirely appropriate in this context. When countries’ governments go bankrupt, they stop paying what they owe. They may default explicitly by reneging on principal and interest payments on their debt. Or they may fail to pay promised benefits and meet other spending commitments; some people regard this as default. A particularly popular way of implicitly defaulting on spending obligations and on debt is to use inflation to do the dirty work. The government simply prints the money it needs to “meet” its spending obligations. The increase in the money supply generates inflation, which waters down the real value of the government’s spending and reduces the real value of its debt.
Although bankruptcies of national governments are not common, neither are they rare. Argentina’s government defaulted in 2001, and Russia’s government defaulted in 1998. Before Russia, Bulgaria and other countries in Eastern Europe did likewise. In the 1970s, Israel and Bolivia took their turn at printing money to pay their bills. Going back further in time, there were the notable hyperinflations of Germany, Austria, and Hungary. Indeed, governments have engaged in official and unofficial default since at least the time of Rome’s Emperor Diocletian, who ran what may well have been the mother of all hyperinflations.
One can assess whether fiscal policy is sustainable by examining the size of long-term fiscal imbalances. Doing so provides an early warning of explicit or implicit default and also indicates the magnitude of the adjustments needed to preclude default. Unfortunately, the main indicator of a country’s fiscal imbalance is its annual official deficit—the difference between annual expenditures (government purchases and transfer payments) and tax and nontax receipts. The use of the deficit to assess fiscal sustainability is not simply a matter of habit. It was enshrined in the Fiscal Responsibility Act of the European Monetary Union’s (EMU) Maastrict Treaty. This act limits entry to the union to countries whose deficits are less than 3 percent of GDP and imposes hypothetical penalties on EMU member nations that violate this limit. I say “hypothetical” because the treaty permits exemptions from the deficit limit that, as of 2005, appear more the rule than the exception.
But the deficit and its associated cumulate—the debt—are, economically speaking, meaningless because nothing in economic theory tells us whether any particular government receipt should be called (labeled) a “tax” or “borrowing.” Nor is there anything in economic theory that can tell us whether a government payment to any entity should be labeled a “transfer payment” or “debt repayment.” Thus, when the EU proclaimed that the difference between expenditures and taxes in a given year—the deficit—should be limited to 3 percent of GDP, it failed to ask itself whether there was any economic basis for distinguishing expenditures from debt repayment or taxes from borrowing.
An example helps clarify this. In 2005, the U.S. federal deficit was projected to total $325 billion. But because most people think of Social Security as a forced pension plan, we could legitimately label this year’s receipts of Social Security taxes a forced loan to the government rather than taxes and label the future payments associated with these “contributions” “return of principal plus interest less an old-age tax” rather than “transfer payments.” If we did so, the U.S. deficit would amount to $1.1 trillion.1 This alternative labeling is just as valid in terms of nomenclature and just as meaningless in terms of economics.
Because the labeling of government receipts and payments is arbitrary, governments can choose labels that generate whatever time path of deficits or surpluses they wish to report, regardless of the true underlying fiscal policy. Thus, they can announce huge surpluses precisely when they are engaging in the most fiscally irresponsible policies, and they can announce huge deficits when their policies are fiscally most conservative.
This license to define the deficit became apparent in the run-up to the introduction of the euro. France’s government, for example, met the deficit limit by selling assets of state enterprises while retaining liabilities. But those manipulations were generally viewed as isolated abuses of a fundamentally sound fiscal measure. The measure is arbitrary, though. The fact that a change in words, not actions, could, from one second to the next, generate a different picture of a nation’s fiscal finances should have made clear that there is, indeed, no unique way to label receipts and payments, and therefore no unique measure of the deficit.
Use of “the” deficit not only tells us nothing about the stance of current policy, it also tells us nothing about policy changes. The reason is that no one can say whether changes in the deficit arise from true changes in fiscal policy or simply a change in labeling conventions. Thus, the Bush tax cuts, which have increased the official U.S. federal deficit, can be viewed as a pure change in labels under which the government called more current receipts from the private sector “borrowing” rather than “taxes,” and will call more future receipts “taxes” rather than “borrowing.” And the fact that all but one lonely economist uses the former rather than the latter set of words does not endow the former set of words with any special economic content. The dictates of economic theory are not determined by majority rule.
I refer to the problem outlined above as “fiscal relativity” because the perception of a government’s fiscal position depends on the beholder’s reference point. The reason the stock of government debt and its changes over time are not well-defined economic measures is that they do not answer an economic question. An example of an economic measure that does fit this criterion is gross domestic product, which answers the question: What is the value of all final goods and services produced by an economy in a given year?
Compare that question with the question: What are a government’s official liabilities? Although the word “liabilities” has economic content, the word “official” does not. It is a legal or accounting categorization, not an economic one. And from an economics perspective, there is nothing sacrosanct about this categorization—that is, each of us is free to make up our own definition of “official liabilities.”
Moreover, any allocation of official debts and assets to particular programs is entirely arbitrary. Thus, in the U.S. context, we see an endless debate about whether the Social Security Trust Fund should or should not be included in assessing Social Security’s long-term underfunding. The answer is that there is no answer. Either practice is equally valid and equally meaningless. Some Americans want to claim that Social Security is in great shape because the Trust Fund is flush with government bonds that the rest of the government is responsible for paying. If they believe this, then they would have to believe that the rest of the federal government is in terrible shape because of the huge amount of money it owes the Social Security Administration. Other Americans want to claim that Social Security is in terrible shape because the bonds held in the Trust Fund are assets properly attributed to the rest of the government’s fiscal operations. No one can stop either side, but that should not prevent us from recognizing that both positions are mindless.
In contrast to “official” debt, “generational accounting” answers an economic question, namely: What is the net tax burden facing future generations assuming current generations pay no more in net taxes than current policy suggests? “Net tax” refers to the actuarial present value of all future taxes minus all future transfer payments.
This question can be understood by referring to the following formulation of the government’s intertemporal budget constraint, that is, its budget constraint over time:
A = C + D − B,
where A is the present value of net tax payments of future generations, B is the present value of net tax payments of current generations, C is the present value of government purchases, and D is official net liabilities.
Given any labeling convention, C + D can be viewed as the government’s bills and B can be viewed as the amount of those bills to be paid by current generations. The value of the difference D − B is invariant to labeling conventions, but the absolute sizes of D and B are not. Since C is a well-defined measure and D − B, which represents the sum of explicit plus implicit debt, is also well defined, their sum, A, is itself well defined. So, different choices of labels will increase or decrease D and B by the same amount but leave A unchanged. Stated differently, you can use any labeling convention you want and still arrive at the same measure of the collective fiscal burden facing future generations. Consequently, generational accounting makes full use of “official” government statistics in determining the size of A.
Once you calculate A, you can (a) determine the lifetime net tax burden facing individual future generations assuming each pays the same lifetime net taxes on a growth-adjusted basis, and (b) compare the growth-adjusted lifetime net tax burden of future generations with that of current newborns. This comparison is label free because the lifetime net tax burden of newborns is the same, regardless of the choice of labels. “Growth adjusted” refers to having net taxes rise for each successive future generation at the rate of labor productivity growth.
If the growth-adjusted lifetime net tax burden facing future generations is larger than that facing newborns, generational policy is referred to as “imbalanced.” If the burden facing future generations is larger than those generations are willing to put up with politically, fiscal policy is deemed to be both generationally imbalanced and economically unsustainable.
The Fiscal Gap
The “fiscal gap” is a closely related measure of long-term fiscal imbalances. The equation used to determine the fiscal gap, G, is given by
G = C + D − B − A*,
where A* is the net tax burden that would face future generations if there were no generational imbalance, that is, if future generations were to face the same lifetime net tax bill as current newborns after adjusting for growth.
Like the imbalance in generational policy, the fiscal gap is a label-free and well-defined measure of a nation’s long-term fiscal problem. In the United States, the current (2005) fiscal gap equals $65.9 trillion. This estimate comes courtesy of Jagadeesh Gokhale and Kent Smetters (2005), who based their calculation on U.S. government projections—which, incidentally, tend to be overoptimistic.
Some economists believe that if productivity in the next two decades grows at about 3 percent a year, a view Robert J. Gordon supports, then we can “grow our way out” of our fiscal problems.2 Unfortunately, this is false. Even if productivity grows at 3 percent annually, that would only slightly improve the picture, for one simple reason: all of the spending on Social Security payments is indexed to real wages, and although Medicare spending is not indexed to real wages, real wage growth causes nonseniors to demand more and better health care, which tends to become the standard for Medicare. High-productivity growth would do much to solve the problem if and only if the U.S. government took a step similar to one that former British prime minister Margaret Thatcher took in the 1980s, namely, indexing Britain’s version of Social Security benefits to consumer prices rather than to wages. Britain is almost alone in Europe in having taken care of its fiscal future.
One way to put the U.S. fiscal gap in perspective is to ask how much of a tax hike would be required to make the present value of the new taxes equal the gap. The answer is that U.S. federal personal and corporate income taxes would have to be doubled, immediately and permanently! Alternatively, the gap could be closed by immediately and permanently cutting by two-thirds the elderly’s Medicare health benefits as well as their Social Security pension benefits!
Either of these policies or any combination of them would impose a huge burden on current adults. But American adults appear in no mood to endorse any fiscal adjustments that either raise their taxes or cut their benefits. Of course, what people want is often far removed from what they can get. As the government’s intertemporal budget constraint reminds us, generational policy is a zero-sum game. So leaving today’s “grownups” off the hook means forcing young and future Americans to pay this bill in its entirety. Such a policy is not only ethically abhorrent but also appears to be economically infeasible because it would entail a doubling of the average lifetime net tax rates levied on today’s young and future generations.
Economic projections and measurement have now reached the point where people can understand the sustainability of fiscal policy by doing generational accounting and fiscal gap analysis. Some governments, like the United Kingdom’s, are beginning to do precisely that.
Note that the “old-age tax” would equal the difference between the actual payment and what would constitute interest on the “loan” given prevailing interest rates.
Robert J. Gordon, “Exploding Productivity Growth: Context, Causes, and Implications,” Brookings Papers on Economic Activity 2 (2003): 207–279.