By Laurence J. Kotlikoff
Those concerned about large deficits usually argue as follows: deficits let current generations off the hook for paying the government’s bills. Therefore, current generations consume more. This reduces the amount Americans save and invest. A reduced rate of investment means less capital per worker and, therefore, lower productivity growth. When capital is scarce, its rate of return rises, causing interest rates to increase. Higher U.S. interest rates attract foreign investment to the United States and imply larger trade deficits, because increased foreign investment must increase the trade deficit (see Balance of Payments).
Yet, there is very little correlation between budget deficits and interest rates, saving and investment rates, or productivity growth rates. Some economists, led by Robert Barro of Harvard, claim that the absence of a correlation is evidence for their view that deficits do not matter. They take seriously an off-the-cuff remark by David Ricardo (one of the great nineteenth-century economists) that deficits may not matter because current generations will hand future generations the means to pay off the debt. Barro argues that parents and grandparents do this by making bequests and gifts to children and grandchildren. But for such transfers to be large enough, older generations must have strong altruistic ties to younger generations. Recent studies find that, on the contrary, the ties are weak.
Other economists, who worry about deficits, claim that the correlation between the deficit and other economic variables is so low because the deficit has been defined incorrectly. Two such economists are Robert Eisner (see Federal Debt) of Northwestern University and Stanford’s Michael Boskin, chairman of the President’s Council of Economic Advisers under George H. W. Bush. They point out that the government’s official debt measures only the government’s liabilities. It completely ignores the government’s assets. Using the government’s debt figures to assess its financial position is, in their view, akin to calling the owner of a $1 million property a debtor because he has a large mortgage on the property. These and other economists also fault the conventional deficit measure for failing to correct for inflation.
It is hard to know where these corrections should end. The research of Martin Feldstein, a Harvard economist and former chairman of the Council of Economic Advisers, suggests that the unfunded liabilities of government retirement programs, such as Social Security, should be included in the deficit. Including such liabilities would more than triple the measure of U.S. federal debt. But if the government’s commitments to pay Social Security benefits should be included, shouldn’t we also include implicit commitments to other federal expenditures, such as those for defense and national parks?
The debate over how to measure the deficit is not confined to academics. In recent years many members of Congress have noted that the traditional deficit (i.e., the one that gets reported) will, in the late nineties, be reduced because revenues from Social Security taxes will greatly exceed Social Security payments. They worry that including Social Security in the deficit may mask the true fiscal picture. To avoid this outcome, Congress in 1990 redefined the federal deficit to exclude Social Security receipts and payments.
This is not the first time Congress redefined the deficit, nor the last. Indeed, by 1993 Congress had restored the old definition of the deficit that includes the Social Security surplus. The manipulation of the definition should not be surprising. Since everyone is sure the deficit should be zero, but no one is sure how to measure it, the deficit’s definition has real implications for economic and budget policy. Choosing a definition that makes the deficit large will invite efforts to lower it by limiting spending or increasing taxes. The opposite will be true with definitions that make the deficit appear small.
The simple fact is that the deficit is not a well-defined economic concept. The current measure of the deficit, or any measure, is based on arbitrary choices of how to label government receipts and payments. The government can conduct any real economic policy and simultaneously report any size deficit or surplus it wants just through its choice of words. If the government labels receipts as taxes and payments as expenditures, it will report one number for the deficit. If it labels receipts as loans and payments as return of principal and interest, it will report a very different number.
Take Social Security, for example. Social Security “contributions” are called taxes, and Social Security benefits are called expenditures. If the government taxes Mr. X by $1,000 this year and pays him $1,500 in benefits ten years from now, this year’s deficit falls by $1,000 and the deficit ten years hence will be $1,500 higher. But the taxes could just as plausibly be labeled as a forced loan to the government, and the benefits could be labeled as repayment of principal plus interest. In that case there would be no impact on the deficit.
There are real problems to be concerned about, but the federal deficit doesn’t measure those problems. One thing that neoclassical economists are concerned about is intergenerational transfers. Many government programs—some of which increase the traditional deficit (because of our choice of words), and some of which, like pay-as-you-go Social Security, don’t increase the deficit (again because of our choice of words)—transfer resources from one generation to another. The redistribution of resources by the government occurs between existing and future generations as well as between young and old existing generations.
According to Franco Modigliani’s life-cycle model (the most famous neoclassical macroeconomic model), policies that redistribute from future generations to current generations, as well as policies that redistribute from younger people to older ones, cause national consumption to increase and national saving to fall. The reason is that older generations have larger propensities to consume than do younger generations. This reflects the fact that older people, who are closer to the end of their lives, want to spend their remaining resources more quickly.
Hence, redistribution from younger to older generations can raise consumption, lower saving, lower investment, raise interest rates, increase trade deficits—in short, do all the bad things that have been ascribed to deficits. But using the federal deficit as a measure of U.S. generational policy is like driving in Los Angeles with a map of New York. While the map may be highly detailed, with overlays and multiple colors, it will, nonetheless, get us lost.
For measuring the government’s generational policies, neoclassical economics suggests an alternative to deficits: generational accounts. Generational accounts indicate in present value what the typical member of each generation can expect to pay to the government, minus benefits from the government, now and in the future. A generational account is thus a set of numbers, one for each existing generation, indicating the average remaining lifetime burden imposed by the government on members of the generation. Used properly, these accounts help assess generational policy, independent of the labels the government gives to receipts and payments.
Generational accounts indicate not only what existing generations will pay, but also what future generations are likely to pay. The burden on future generations is determined by working through the government’s “intertemporal budget constraint.” This constraint says that the present value of the government’s spending on goods and services cannot exceed the sum of three items: (1) the government’s net wealth, (2) the present value of net payments to the government by current generations, and (3) the present value of net payments by future generations. Translation: the government cannot spend more than the sum of what it has and what it can raise. At any point in time we can project the present value of the government’s spending and also estimate items 1 and 2. By subtracting 1 and 2 from the present value of government spending, we can determine the aggregate present value taken from future generations. By one set of estimates, as of 1989 the present value of future government spending was $25.4 trillion, the government’s net wealth was—$0.5 trillion, and the present value of net payments of current generations was $21.2 trillion. This left $4.7 trillion to be paid by future generations.
An analysis of U.S. generational accounts for 1991 indicates that unless U.S. economic policy is decisively altered, the typical member of future generations will end up paying roughly 71 percent more over his or her lifetime than will the typical member of current young generations! This figure is above and beyond the fact that future generations will pay more because their incomes will be higher due to economic growth. This 71 percent figure is extraordinarily high and indicates that U.S. economic policy is, generationally speaking, very badly out of balance.
Generational accounting leads to a radically different interpretation of postwar economic policy than does reliance on the deficit. From the perspective of generational accounts, the fifties, sixties, and seventies were periods of quite loose fiscal policy (policy that placed larger burdens on future generations). The reason was the buildup of our unfunded pay-as-you-go Social Security, civil service, and military retirement programs. The eighties, in contrast, were marked by rather tight fiscal policy. While the Reagan tax cuts did increase the burden on future generations, other policies, particularly the 1983 Social Security reform, greatly reduced the projected burden on future generations. By raising the retirement age in stages to sixty-seven from sixty-five, and by gradually subjecting all retirees’ Social Security benefits to income taxation, the 1983 reforms reduced the present value of Social Security benefits to be paid to current adults by about $1.1 trillion.
Generational accounting automatically deals with each of the major concerns raised by those who think the deficit is conceptually sound but simply needs to be adjusted. By measuring all current and projected payments and receipts in inflation-adjusted (constant) dollars, the proposed accounting deals with changes in the price level. It uses the government’s assets minus its liabilities to form the value of government net worth, which is ultimately used to help determine the “hit” on future generations. In considering future government payments and receipts to and from individuals, it accounts for the commitments to pay future Social Security benefits and to spend on other items, such as defense and parks. In projecting the future level of government payments and receipts, it takes into account economic growth. Finally, it considers the fiscal actions of all governments—federal, state, and local.
Generational accounting thus represents a sensible alternative to the deficit delusion that has misled postwar fiscal policy.
Laurence J. Kotlikoff is an economics professor at Boston University and a research associate at the National Bureau of Economic Research. From 1981 to 1982 he was senior economist for taxation and Social Security with the Council of Economic Advisers.
Kotlikoff, Laurence J. “Deficit Delusion.” The Public Interest (Summer 1986): 53-65.
Kotlikoff, Laurence J. “From Deficit Delusion to the Fiscal Balance Rule—Looking for a Meaningful Way to Describe Fiscal Policy.” National Bureau of Economic Research working paper no. 2841, February 1989.
Kotlikoff, Laurence J. “Generational Accounts—A Meaningful Alternative to Deficit Accounting.” In Tax Policy and the Economy, vol. 5, edited by David Bradford. 1991.
Kotlikoff, Laurence J. Generational Accounting: Knowing Who Pays and When for What We Spend. 1992.