January 30, 1998
« More Economic Letters
The Budget Deficit
J. Bradford DeLong
The 1997 accounting year of the federal government ended last September 30, recording a budget deficit of $22 billion–not quite 0.3% of national product. President Clinton will submit a balanced budget for fiscal 1999. For all intents and purposes, the budget is in balance.
But “balance” is a fuzzy concept. There are plausible technical adjustments and corrections to the budget that would have put it into either surplus or deficit last year. For example, treating government investments as capital expenditures to be depreciated–as a business would–would have produced a budget surplus of some $43 billion. Other technical adjustments and corrections–for example, removing the Old Age Survivors Disability Insurance and Medicare trust funds from the unified budget totals–would have pushed the fiscal 1997 budget deeper into deficit. The important thing to note is no matter what concept of the deficit one prefers, or what technical adjustments and corrections one makes, the deviation of the budget balance from zero remains too small for it to have a substantial impact on the American economy.
On one level, the improvement in the deficit in this decade is a heartening and important achievement. Even though relatively little was done to affect the future course of the deficit in 1997–the President-congressional deficit-reduction agreement of 1997 was small compared to previous deficit-reduction agreements–the cumulative total reduction in the deficit so far this decade is a substantial economic policy victory. It shows that the American system possesses more flexibility than many had thought: a decade ago “structural” explanations of the deficit as rooted in the institutional interaction of the President and Congress, and thus as unavoidable, were common.
But just because this past year’s federal budget was in rough balance does not mean that the fiscal situation in the U.S. is stable. The long-run fiscal crisis of the social insurance state–the fact that Medicare and Social Security taxes are inadequate to pay the benefits that the government has promised over the next century–remains unresolved. And successive administrations and congresses keep passing up opportunities to begin resolving it.
In the summer of 1997 the President and the congressional leadership announced the third deficit-reduction agreement in this decade, which carried forward the work of the bipartisan 1990 deficit reduction agreement and the Democratic 1993 deficit reduction program. However, there was one significant difference between the 1997 agreement and the previous agreements. The amount of deficit reduction contained in the 1997 agreement was relatively small. The 1990 and 1993 deficit-reduction agreements appeared to be between six and eight times as large in the relative size of their effects on the economy (Figure 1).
The reason for its small relative size is that given the course of the American economy, there was simply not that much deficit left to reduce: in 1992 the federal government budget deficit had amounted to 4.7% of GDP, but in 1997 it amounted to only 0.3% of GDP. Thus the bulk of the work of eliminating the deficit had already been done–by the strong recovery of the U.S. economy from the recession of the early 1990s and by the 1990 and 1993 deficit-reduction agreements.
Nearly half of the reduction in the deficit since 1992 from 4.7% to 0.3% of GDP is due to the reduction in the unemployment rate and the increase in GDP relative to potential output. In fiscal 1997, the U.S. unemployment rate averaged some 2.5 percentage points below its level of fiscal 1992. According to Okun’s Law, such a reduction in unemployment reflects an increase in GDP relative to potential output of some 6.25%. Such an increase in output relative to potential has a striking effect on the deficit: at the margin an extra dollar of real GDP increases federal tax collections by some $0.25 and decreases federal spending by some $0.07. Thus the improvement in business cycle conditions since 1992 is responsible for reducing the deficit as a share of GDP by some 2.0 percentage points.
The remaining 2.4 percentage points of reduction in the deficit as a share of GDP are a reduction in the cyclically adjusted deficit. The bulk of this is due to policies enacted in the 1990 and 1993 agreements, such as increases in taxes, and reductions in spending growth below the growth rate of the economy as a whole. And a final component is due to “extraordinary” factors, like the end of expenditures by the Resolution Trust Corporation which had been set up to handle the consequences of the 1980s savings and loan crisis.
The success of U.S. economic policymakers in reducing the deficit without imperiling continued economic recovery is worthy of note. Standard estimates of the effects of a given change in government spending and taxes on real GDP (known as Keynesian multipliers) today are less than they used to be–in the range of 1.5 to 2.2. Nevertheless the substantial reduction in the cyclically adjusted deficit between 1992 and 1997 reduced aggregate demand by between 3 1/2% and 5 1/2% of GDP; under most circumstances this would trigger a mild to moderate recession. Yet in the U.S. the mid-1990s have not seen any signs of recession, in part because the economy in 1992 was poised for cyclical recovery and in large part because of the skillful conduct of monetary policy. The Federal Reserve’s efforts to keep interest rates relatively low managed to offset successfully any contractionary impact of reductions in the cyclically adjusted deficit without triggering renewed inflation. Given how often the Federal Reserve, the Congress, and the President are blamed for an inappropriate or faulty monetary and fiscal policy mix, it is worth pausing to note that in the mid-1990s the policy mix appears to have been exactly right.
In the United States, the era of large government deficits dates from 1974 (Figure 2). The era of overwhelming deficits–deficits so large that they are not just a serious economic problem but the economic problem–dates from 1981. By combining tax cuts with increases in defense spending, the Reagan administration and its congressional supporters made a mistake in budgetary policy that amplified the fiscal difficulties that had been created by slow growth in the 1970s and that gave the United States some 15 years of unprecedented peacetime budget deficits.
It is difficult today to understand the thought processes of those who set fiscal policy for the Reagan administration. Certainly no one intended to create large budget deficits that would drain the pool of capital for investment and retard the growth of the American economy. And even today the story is not clear, in large part because for more than fifteen years those who developed Reagan administration fiscal policy have argued among themselves over just who made the key mistakes and over just what the key mistakes were.
By absorbing capital that otherwise would have funded private investment, the deficits left the U.S. with a lower capital stock, a less productive economy, and a debt owed to overseas investors that must now be amortized. How destructive were these deficits? Different assumptions about the structure of the U.S. economy and different methodologies lead to different results.
Higher estimates come from models that characterize the U.S. as a “closed economy”–that is, one in which international capital does not flow in or out, so that no part of the deficit can be financed from abroad. Assuming a 10% real pretax social rate of return on investment and assuming that technological change and savings behavior do not respond to the deficit lead to the conclusion that U.S. real production today is 5% less than had the federal budget been balanced since 1981. But the “closed-economy” assumption is highly inappropriate, even though the savings assumption may not be.
Lower-end estimates come from models that characterize the U.S. as an economy in which all budget deficits can be thought of as financed from abroad. Such models lead to the conclusion that U.S. real GDP today is some 1.5% less than had the federal budget been balanced since 1981. But the assumptions of these models are inappropriate as well.
These estimates of between 1.5% and 5% provide us with boundaries to calculate the lost annual income for the economy as a whole–roughly between $1,000 and $3,500 for the average American worker–as the net consequence for economic growth of the era of deficits. It is possible to obtain estimates that are lower (or higher), but they require making assumptions about the structure of the economy that are even more speculative.
The end of the era of deficits means that the total drag on the economy inflicted by cumulative budget deficits is no longer increasing. It also means that fears common in the 1980s that the U.S. political system had broken down and was no longer capable of producing rational fiscal policy decisions have turned out to be overly alarmist. Indeed, looking back, perhaps the single most important step in bringing the federal deficit down was a procedural step, the Budget Enforcement Act (BEA) of 1990 imposed by President Bush as part of his price for agreeing to the 1990 budget deal. It changed the rules of Congress, making any proposal that would increase the deficit automatically out of order, and requiring supermajority votes to suspend this rule. It also made the Congressional Budget Office (CBO) immensely powerful, for it was the CBO that decided whether a legislative proposal would increase or reduce the deficit, and therefore whether the proposal was in or out of order. By and large, all agree that the CBO has handled its power well, delivering decisions that have been professional and technical, rather than political. And the CBO’s ability to use its power has been very effective in the past seven years in preventing Congress from passing bill after bill that each adds a drop or two to the deficit. The U.S. political system has demonstrated some flexibility and competence. And that is certainly cause for some celebration.
While we should celebrate the achievement of rough balance in fiscal 1997, we also need to look further ahead and recognize the temporary nature of the current fiscal policy success. Further into the future the fiscal outlook turns downward again, with renewed and growing budget deficits beginning late in the next decade.
This worsening scenario is projected because the United States today has a social insurance system, a Medicare and Social Security system, that was designed back in the 1970s for an economy that would be growing at a measured average rate of 2.5% per person per year. But the United States today in fact grows more slowly–at an average speed of less than 1.5% per person per year.
Thus the taxes that have been earmarked to pay for Social Security and Medicare as the U.S. population ages and the baby-boom generation approaches retirement are not going to cover the costs of providing currently promised benefits for far into the next century. At some point before the baby-boom generation reaches retirement age, the country will have to decide either to cut Social Security and Medicare benefits below levels that have been implicitly and explicitly promised or to raise social insurance taxes. The sooner the American political system makes this choice, the easier the process of adjustment will be. The longer the choice is delayed, the more disruptive and difficult will be the process of transition and adjustment to a sustainable social insurance system.
J. Bradford DeLong
More Economic Letters
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to