Skip Navigation

Why Adjust Measures of the Total Surplus or Deficit?

Automatic Stabilizers

The Cyclically Adjusted Surplus or Deficit

The Standardized-Budget Surplus or Deficit

The Effects of Recent Tax and Spending Legislation

Measuring Budgetary Trends and Fiscal Stimulus

Appendix: Details of CBO's Projections of the Standardized Budget
 

Tables
1.	Adjusted Measures of the Surplus or Deficit, 1998-2003
2.	Fiscal Stimulus in Periods of Recession
A-1.	The Standardized-Budget Surplus or Deficit, 1998-2003
A-2.	Details of the Standardized-Budget Surplus or Deficit, 1998-2003 (In billions of dollars)
A-3.	Details of the Standardized-Budget Surplus or Deficit, 1998-2003 (As a percentage of potential GDP)
A-4.	The Standardized-Budget Surplus or Deficit and Related Series, 1960-2001 (In billions of dollars)
A-5.	The Standardized-Budget Surplus or Deficit and Related Series, 1960-2001 (As a percentage of potential GDP)

 

In August 2002, the Congressional Budget Office (CBO) released its updated baseline projections of federal revenues, outlays, surpluses, and deficits for the next 10 years.⁽¹⁾ Those projections show that the federal budget has gone from a surplus of $236 billion in 2000 to a deficit of about $157 billion in 2002 and that, if current policies continue, it is likely to show a deficit of $145 billion in 2003. The size of the budget surplus or deficit reflects temporary factors, such as the effects of the business cycle or of one-time shifts in the timing of federal spending and tax receipts, as well as the longer lasting impact of legislation, such as the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA) and the Job Creation and Worker Assistance Act of 2002 (JCWAA). To help separate out those factors, this report presents updated estimates of two adjusted budget measures. Those measures are the cyclically adjusted surplus or deficit (which filters out the effects of the business cycle) and the standardized-budget surplus or deficit (which removes the effects of the business cycle and other factors).

By those measures, roughly one-third of the projected decline in the total surplus between 2000 and 2003 results from "automatic stabilizers"--the automatic response of the budget to the business cycle. Most of the remaining two-thirds is attributable to legislative action: primarily EGTRRA, JCWAA, and increases in discretionary spending (including emergency appropriations enacted in response to the terrorist attacks of September 11). The largest increase in the deficit during that period occurs in 2002. Legislation accounts for roughly half of that increase, and automatic stabilizers for about one-quarter. The rest results from a variety of factors, including unusually large tax refunds this year (stemming from overpayment of estimated and withheld taxes in 2001) and, most likely, weakness in tax payments on capital gains realizations.

The year-to-year changes in the standardized and cyclically adjusted budgets suggest that fiscal stimulus (a reduction in government saving that increases total demand) during the recent recession was unusual in both its magnitude and timing. During some past periods of recession, by contrast, the passage of stimulative measures was delayed, resulting in fiscal restraint during the recession and stimulus that began only after the recovery was under way.
 

Why Adjust Measures of the Total Surplus or Deficit?

Despite some limitations, both conceptual and empirical, budget measures that separate out cyclical and other temporary factors are useful in a number of ways. For example, some analysts use those measures to determine whether fiscal policy is providing stimulus to short-term growth. Others use them to discern underlying trends in government saving. More generally, those measures provide an estimate of the extent to which changes in the budget are caused by normal movements of the business cycle and thus are likely to prove temporary.

Drops in revenues and increases in outlays occur automatically during a cyclical downturn and then correct themselves in a cyclical upturn. Those changes are called automatic stabilizers because they help even out cyclical swings in the overall level of demand for goods and services, adding to demand during recessions and reducing it during booms. The cyclically adjusted surplus or deficit shows the underlying outcome of the federal budget when those automatic stabilizers are removed.

Policy actions by the Congress and the President, such as legislated tax cuts or increases in spending during recessions, can also affect demand. Those actions create changes in the total budget surplus or deficit that are separate from the automatic cyclical movements. The cyclically adjusted surplus or deficit includes those legislated changes. It also reflects other factors, not directly connected with changes in policy, that alter revenues or spending. For example, drops in the stock market can cause revenue shortfalls because receipts from capital gains taxes decline, as has happened in the past two years. (Movements in capital gains tax receipts are not treated as cyclical because the correlation between those receipts and the business cycle is usually tenuous.) Similarly, even explicit budget decisions can produce temporary changes in the timing of tax receipts or government spending. CBO calculates another measure, the standardized-budget surplus or deficit, that attempts to adjust for those factors as well. To calculate that measure, CBO must judge whether each factor is likely to affect total demand significantly. As a result, the standardized-budget surplus or deficit is the more speculative of the two measures presented here.

Those measures consider only the short-run impact of the government's budget on total demand in the economy. But federal taxes and spending affect the economy in many ways and alter the prospects for economic growth in the longer run, by changing incentives to work, save, and invest. Frequently--as when EGTRRA was enacted--both short-term and longer-term effects are intended. However, budget measures are generally not useful in identifying the economic effects of changes in incentives.⁽²⁾ CBO's estimates of those longer-term effects are incorporated in its economic forecasts.⁽³⁾
 

Automatic Stabilizers

The automatic stabilizers built into the fiscal system help sustain total demand during a recession, thus dampening the severity of the recession.⁽⁴⁾ The main automatic stabilizer occurs when tax revenues decline as wages and salaries, corporate profits, and other tax bases shrink in a business downturn. Because of the progressivity of the tax system, revenues drop slightly more in percentage terms than those tax bases do. The effect of the drop in revenues is that consumers' after-tax income declines by less during a recession than their pretax income does, thus cushioning the effect of the recession on consumption.⁽⁵⁾ Another automatic stabilizer is the increase in spending for entitlement programs such as unemployment insurance, Food Stamps, and Medicaid that occurs as unemployment rises. Because those stabilizers are automatic, they do not require that policymakers recognize the downturn and act quickly--they automatically occur on time. Moreover, when a recession ends, the automatic stabilizers turn themselves off without the need for policy intervention.

The automatic stabilizers have provided a less-than-average amount of stimulus during the most recent recession period, but that simply reflects the fact that the recession was less severe than average. The automatic stimulus shows up as a decline in the cyclical surplus from 1.0 percent of potential gross domestic product (GDP) in 2000 to 0.2 percent in 2001, followed by a cyclical deficit of 0.4 percent of potential GDP in 2002 (see Table 1). Under current economic and budget projections, that stimulus is expected to shut off automatically next year (and instead be replaced by very minimal automatic restraint), as indicated by a slight decrease in the projected cyclical deficit between 2002 and 2003.
 

The Cyclically Adjusted Surplus or Deficit

Calculations of cyclically adjusted budget measures attempt to remove the effects of the business cycle on revenues and outlays (that is, the automatic stabilizers that are measured by the cyclical part of the budget). For example, cyclically adjusted revenues exclude the revenue loss that automatically occurs during recessions.⁽⁶⁾ Likewise, cyclically adjusted outlays exclude the additional spending that follows from a rise in unemployment. The difference between those two measures is the cyclically adjusted surplus or deficit.

Between 2000 and 2003, CBO estimates, the cyclically adjusted budget will decline from a surplus of 1.5 percent of potential GDP to a deficit of 1.1 percent (see Table 1). That decline equals roughly two-thirds of the change from the surplus of $236 billion recorded for 2000 to the deficit of $145 billion projected for 2003. It suggests that the other one-third is attributable to the business cycle. The difference between the total and cyclically adjusted surplus was smaller in 2001 than in previous years and turned into a deficit in 2002, reflecting the budgetary effects of the recent recession. That cyclical difference changes relatively little between 2002 and 2003.
 

The Standardized-Budget Surplus or Deficit

CBO routinely publishes another adjusted budget measure, the standardized-budget surplus or deficit. That measure excludes the effects not only of cyclical fluctuations but also of certain more-or-less-temporary factors that are likely to prove economically insignificant.⁽⁷⁾ Four such factors are unusually large discrepancies between tax payments and liabilities, swings in collections of capital gains taxes, changes in the inflation component of the government's net interest payments, and temporary legislative changes in the timing of revenues and outlays.

A substantial discrepancy between tax payments and liabilities emerged in 2001, when people did not reduce their estimated and withheld tax payments even though their tax liabilities declined because of such factors as stock market losses and smaller-than-expected bonuses at the end of the year. CBO estimates that those temporary tax overpayments in 2001 (and corresponding increases in refunds in 2002) totaled about $25 billion for individuals.⁽⁸⁾ Those overpayments should have little impact on the spending of taxpayers whose financial liquidity is not a binding factor--a group that accounts for perhaps 70 percent of total consumption.⁽⁹⁾ For that reason, in calculating the standardized budget, CBO treated 70 percent of those overpayments (and 70 percent of similar discrepancies between tax payments and liabilities in the past) as if they affected only the timing of tax payments and not consumption. That adjustment removes most of the temporary overpayments from 2001 and reduces refunds in 2002 by the same amount.

CBO removes capital gains tax receipts from the standardized budget because movements in those receipts are not captured by the cyclical adjustments to revenues. Moreover, removing those tax receipts avoids the misleading effects that can arise, for example, when a cut in the tax rate on capital gains stimulates taxable capital gains realizations by enough to increase the standardized-budget surplus, causing that measure to incorrectly indicate fiscal restraint.

The inflation component of net interest is removed because it primarily maintains the inflation-adjusted value of outstanding federal debt. Since it does not raise that value, it is likely to have little impact on total demand in the economy.

Legislation enacted by the Congress and the President sometimes temporarily shifts the timing of receipts or outlays (usually from the end of one fiscal year to the beginning of the next one). Those small timing shifts are excluded from the standardized budget because they are unlikely to alter significantly the timing of people's plans for investment or consumption.

In addition, the standardized budget excludes receipts from the government's sale of assets and from auctions of spectrum licenses as well as federal outlays for deposit insurance. The effects of asset sales and spectrum auctions are removed because those transactions are voluntary exchanges of existing assets that have no effect on private net worth and thus little economic impact (except on the businesses whose liquidity they may adversely affect). CBO removes outlays for deposit insurance because the economic impact of those outlays occurred in earlier years, when various thrift institutions failed.⁽¹⁰⁾

The standardized-budget surplus is projected to decline by a total of 2.4 percent of potential GDP between 2000 and 2003--slightly less than the decline in the cyclically adjusted surplus during that period (see Table 1). For both measures, the declines stem mainly from the revenue effects of the tax cuts enacted in those years and from additional discretionary spending (partly in response to the September 11 terrorist attacks). Because the standardized-budget surplus includes other adjustments, it shows a slightly larger decline in 2002 than the cyclically adjusted surplus does and some increase in 2003. More details of the standardized-budget surplus and its components through 2003 are contained in the appendix.
 

The Effects of Recent Tax and Spending Legislation

The estimated effects of legislation enacted since January 2001 account for roughly two-thirds of the decline in the total budget surplus between 2000 and 2003--twice as much as the estimated effects of the business cycle over that period. That legislation also accounts for most of the decline in the cyclically adjusted and standardized-budget surpluses.

Laws enacted since January 2001 lowered those surpluses by 0.5 percent of potential GDP in 2001, 1.8 percent in 2002, and 2.2 percent in 2003 (see Table 1). Almost all of those changes result from EGTRRA, from higher appropriations for defense, and from the tax cuts and additional spending enacted in the Job Creation and Worker Assistance Act.
 

Measuring Budgetary Trends and Fiscal Stimulus

Some analysts use the annual change in a cyclically adjusted surplus (measured relative to potential GDP) as an indicator of the federal government's fiscal posture. When viewed that way, a decline in the cyclically adjusted or standardized-budget surplus indicates fiscal stimulus. Other analysts use cyclically adjusted measures of the budget surplus over several years to discern the underlying trend in government saving and thus the government's contribution to longer-term movements in national saving. For both purposes, the cyclically adjusted and standardized-budget surpluses are probably better than the unadjusted surplus because they filter out changes that stem from the budgetary effects of the business cycle.

The adjusted budget measures give only a partial view of the effect of the budget on total demand. Changes in private saving may partly offset changes in government saving if some people think their future tax liabilities are affected by how much the government currently saves. Furthermore, those measures reflect the budget's bottom line rather than its underlying components of spending and revenues, although those components probably affect demand in different ways. For example, to the extent that tax cuts are saved, they have little effect on current consumption, whereas increases in spending on roads and bridges show up fully in demand. In a recent paper, CBO examined the fiscal stimulus that would result from several kinds of changes in taxes.⁽¹¹⁾ That analysis showed a wide range of short-term effects on demand from different policy actions, for reasons unrelated to their effect on the budget surplus (whether standardized or not).

In addition to those limitations, the estimates of cyclically adjusted and standardized surpluses do not take into account anticipations of fiscal stimulus. Economists generally agree that, in theory, anticipated fiscal stimulus will have an impact before the stimulus is enacted. However, it is not always clear how to separate anticipated from unanticipated changes in fiscal stimulus.⁽¹²⁾

The government's current fiscal stimulus appears to be unusual in both its size and timing. The declines in the standardized-budget and cyclically adjusted surpluses (relative to potential GDP) are larger for 2001 and 2002 combined (2.8 percent and 2.6 percent, respectively) than they were for any other recession years since 1960, including the period of 1981, 1982, and 1983 combined (see Table 2). (Historical estimates for the standardized budget appear in the appendix in Tables A-4 and A-5.) Moreover, in contrast to most past recessions, when significant lags accompanied the adoption and implementation of fiscal stimulus, a major source of stimulus this time began in only the fourth month of the recession (July 2001) in the form of tax rebates.
 

As noted above, the change in the cyclically adjusted and standardized surpluses between 2000 and 2003 is mostly explained by recent legislation. But changes in those budget measures also reflect movements in receipts and outlays attributable to factors that are not captured by the various adjustments discussed above. Those other factors include noncyclical movements in the proportion of taxable income going to people in the top tax bracket, in the size of various tax bases as a share of GDP, and in outlays for entitlement programs such as Medicare and Medicaid.
 

Appendix: Details of CBO's Projections of the Standardized Budget

In 2001, the budgetary changes that are measured by the standardized surplus or deficit tended to stimulate the growth of total demand. The standardized-budget surplus declined to 0.7 percent of potential gross domestic product (GDP) in 2001 from 1.1 percent the year before (see Table A-1). All of that decline can be attributed to a decrease in standardized-budget revenues, which fell from 19.2 percent of potential GDP to 18.7 percent. In turn, that decline in standardized revenues (0.5 percent of potential GDP) reflected the effects of the Economic Growth and Tax Relief Reconciliation Act of 2001 (not counting the $23 billion shift in corporate tax payments from September to October 2001). Movements in standardized-budget outlays in 2001 were relatively small and offset each other to some extent.

For 2002, the change in the standardized budget indicates substantially more fiscal stimulus than in 2001, reflected in a record shift--from a surplus of 0.7 percent of potential GDP to a deficit of 1.7 percent. That is the largest single-year decline in the standardized surplus, surpassing the previous records (1.9 percent of potential GDP) observed in 1976 and 1983. Standardized revenues account for roughly three-fifths of this year's stimulus, and the reduction in standardized revenues (1.4 percent of potential GDP) equals the decline registered in 1983 as a result of the Economic Recovery Tax Act of 1981. Nevertheless, only about half of the stimulus from the revenue side in 2002 reflects recent legislation; the rest results from a decline in the economy's effective tax rate that is not captured by the cyclical and other adjustments used to calculate standardized revenues.

The remainder of the stimulus in 2002 comes from increases in discretionary and mandatory spending, which each amount to 0.5 percent of potential GDP (see Tables A-2 and A-3). In large part, those increases were caused by emergency appropriations enacted in the wake of the September 11 attacks and by additional spending for unemployed workers in the Job Creation and Worker Assistance Act.

In contrast with the stimulus of 2001 and 2002, the standardized-budget projections suggest some fiscal restraint next year. That restraint results from a projected decline in inflation-adjusted interest payments (from 1.3 percent of potential GDP to 1.0 percent) and a small increase in standardized revenues (from 17.3 percent to 17.5 percent of potential GDP). Mandatory outlays are expected to remain essentially unchanged relative to potential GDP, while discretionary spending rises by 0.2 percent of potential GDP.
 

1. See Congressional Budget Office, The Budget and Economic Outlook: An Update (August 2002).

2. The supply-side effects on work and investment are generally thought to be smaller in the short run than in the long run. In the end, however, the impact of fiscal stimulus on the supply of labor and capital will largely depend on how the stimulus is financed. In general, if it is ultimately financed by reducing federal spending, supply-side effects will be enhanced; but if current tax cuts are financed by raising future taxes, the stimulus could have adverse supply-side effects.

3. See Congressional Budget Office, The Budget and Economic Outlook: An Update (August 2001), Box 2-3, pp. 34-35.

4. For theoretical and empirical discussions of automatic stabilizers, see Darrell Cohen and Glenn Follette, "The Automatic Fiscal Stabilizers: Quietly Doing Their Own Thing," Economic Policy Review, Federal Reserve Bank of New York, vol. 6, no. 1 (2000), pp. 35-68. See also Thomas J. Kniesner and James P. Ziliak, "Tax Reform and Automatic Stabilization," American Economic Review, vol. 92, no. 3 (June 2002), pp. 590-613.

5. By contrast with tax liabilities, many financial obligations of taxpayers, such as mortgage payments, do not decline when people lose their jobs.

6. Estimates of the cyclical component of revenues and outlays for 1998 through 2003 are presented in the appendix to this report. Those estimates differ from CBO's rule-of-thumb estimates of the effects of changes in economic assumptions, which appear in The Budget and Economic Outlook: Fiscal Years 2003-2012 (January 2002), Appendix A. The rule-of-thumb estimates attempt to capture sustained changes in the rate of GDP growth and other economic variables, whereas the estimates in this report are based on temporary, cyclical fluctuations.

CBO's estimates of the cyclical component of revenues and outlays depend on the gap between actual GDP and potential GDP, which is the level of output consistent with stable inflation. Those estimates of the cyclical component, however, may not capture all of the movement in revenues and outlays that some analysts might view as cyclical. For example, different estimates of potential GDP would produce different estimates of the size of the cyclically adjusted surplus or deficit. For a discussion of the relationship between the cyclically adjusted budget and potential GDP, see Congressional Budget Office, The Budget Adjusted for the Effects of the Business Cycle (July 30, 1999).

7. See Congressional Budget Office, The Standardized Budget: Revised Historical Estimates (June 2000), and The Budget and Economic Outlook: Fiscal Years 2003-2012 (January 2002), Appendix F, Tables F-11, F-12, and F-13.

8. See Congressional Budget Office, The Budget and Economic Outlook: An Update (August 2002), p. 15.

9. See Congressional Budget Office, The Standardized Budget: Revised Historical Estimates (June 2000).

10. The short-term fiscal impact of deposit insurance is discussed in Congressional Budget Office, The Economic and Budget Outlook: An Update (August 1991).

11. See Congressional Budget Office, Economic Stimulus: Evaluating Proposed Changes in Tax Policy (January 2002).

12. Several empirical studies find that, contrary to accepted theory, individual consumption rises when fully anticipated increases in after-tax income are realized. See, for example, Jonathan Parker, "The Reaction of Household Consumption to Predictable Changes in Social Security Taxes," American Economic Review (September 1999), pp. 959-973; and Nicholas Souleles, "The Response of Household Consumption to Income Tax Refunds," American Economic Review (September 1999), pp. 947-958.