The Heritage Foundation

Center for Data Analysis Report #98-03 on Federal Budget

June 4, 1998

June 4, 1998 | Center for Data Analysis Report on Federal Budget

Budget Priorities May Have Serious Defense Consequences

Defense expenditures have been in steady decline throughout the 1990s, a trend that can be expected to continue unless budget priorities are changed substantially. Between 1990 and 1997, defense outlays dropped from $300.1 billion to $271.6 billion, a reduction of 26 percent in real terms and almost 10 percent in nominal terms. This Center for Data Analysis study shows that if the current pattern of budget priorities continues, the downward trend in defense would result in the virtual elimination of defense spending. To avoid this, there will need to be substantial changes in budget priorities.

The budget debates of the past two years suggest that Congress intends budget priorities to remain relatively fixed in the foreseeable future. If this is so, the Balanced Budget Act of 1997 (BBA of 1997) provides a reasonable indication of future federal revenue and spending priorities. A framework for future budgets appears that is shaped by four guiding principles:

  1. The federal budget will remain balanced. Last year Congress and the Clinton Administration agreed on a plan that would balance the federal unified budget by fiscal year 2002. Better-than-expected economic growth, however, has led both the Congressional Budget Office (CBO) and the Office of Management and Budget (OMB) to forecast a balanced budget by fiscal year 1999 if no further changes are made in current law (see Chart 1).1 The President's most recent budget takes this accelerated path to balance into consideration, and the budget produced by Congress is likely to do the same.

  2. Overall tax rates will not be increased and may be reduced. Last year's budget agreement resulted in a modest reduction in the level of taxation for selected groups, while maintaining the promise of a balanced budget. Current projections are that existing tax rates will result in revenue increases for each of the next five years (fiscal years 1999-2003) (see Chart 2).

  3. Entitlement and non-defense discretionary outlays will continue to grow at rates higher than the inflation rate. Entitlement expenditures have grown rapidly during the 1990s. In fact, outlays for federal entitlement programs, including Medicare and Social Security, have risen from $605.2 billion in 1990 to $880.1 billion in 1997. This is an average annual growth rate of 5.5 percent, which is well in excess of a no-real-growth path.2 (See Chart 3.) This rapid increase in entitlement outlays has been financed in part by new public debt. In 1990, outlays for interest payments were $184.2 billion. By 1997, they had grown to $244 billion, an increase of 32 percent (see Chart 4). This increase reflects the fact that while deficits have come down over the course of the 1990s, the national debt has continued to increase and currently stands at $5.5 trillion. Non-defense discretionary spending--all spending programs other than entitlements, interest on the debt, and defense--has grown steadily during the 1990s at about a 5 percent annual rate of growth (see Chart 5).

  4. Defense spending will be reduced in terms of current, non-inflation-adjusted dollars. Defense outlays were $300.1 billion in 1990. In 1997, they were $271.6 billion. This 9.5 percent reduction in current dollar terms represents a 26 percent reduction in real terms between 1990 and 1997 (see Chart 6).

The pattern of federal spending also can be seen by examining the percentage of total federal expenditures that is dedicated to each function of the government. This view indicates that entitlements were the highest priority of the federal government during the 1990s. Spending on entitlement programs (primarily Social Security and Medicare) was consistently greater than defense and non-defense discretionary spending combined. The second priority has been non-defense discretionary spending, which started the decade at a lower level than defense spending but surpassed defense spending in 1995. The lowest priority of the federal government has been defense (see Chart 7).

The budgets during the 1990s, reinforced by the BBA of 1997, thus follow a pattern. Specifically, the budget is brought into balance. The level of taxation stabilizes, even while revenues increase. On the spending side, expenditures on entitlements and non-defense discretionary spending grow at more than the rate of inflation. The budget for national defense is the only major spending category that decreases in both nominal and real terms. If left unchanged, these policies will have serious consequences for America's national defense (see Chart 8).

Projecting Future Spending Levels

The fiscal trends of the 1990s, as reinforced by the BBA of 1997, can help analysts project future levels of federal government revenues and spending. To produce this projection, fiscal priorities and current trends must be expressed as assumptions regarding budgetary decision-making. Heritage analysts have incorporated the following six assumptions into a model that projects annual revenues and expenditures.

  • The budget is brought into balance in fiscal 1999 and remains balanced thereafter. This is a conservative estimate, and does not allow for reductions in the national debt or changes in the social insurance trust funds beyond the levels currently forecast by the trustees of these funds.3 Any surpluses generated under current law are assumed to be applied proportionally to all types of federal spending, including defense.

  • The national debt stabilizes after fiscal 1999. Because the federal government maintains a balanced budget (but does not run a budget surplus), the national debt remains constant. However, the federal government must continue to service this debt by making annual interest payments. The level of these interest payments is determined by market-established interest rates. Heritage analysts used interest rates and annual levels of federal debt service used in the WEFA long-term U.S. macroeconomic model and released by WEFA in September 1997.

  • Overall tax rates are not increased. While tax rates remain constant, total revenues rise as the economy, and thus the tax base, grows. This is a conservative assumption because many in Congress, and even the President, have talked about reducing taxes through either wholesale reform of the tax system or targeted cuts.

  • Spending on federal entitlement programs, including Social Security and Medicare, is kept at current law levels. This assumption implies that any shortfall in the Social Security and Medicare trust funds will be covered through payments out of general revenues.4 President Clinton, in his 1998 State of the Union Address, stated his intention to use current and future budget surpluses on funding financial shortfalls in the Social Security trust funds.

  • Non-defense discretionary spending increases over the next 21 years at the level estimated by WEFA. Specifically, WEFA assumes an annual growth rate of about 5 percent, or about the same rate of increase as during the 1990s, and slightly greater than inflation over the same period.

  • Defense will receive all the funds remaining under projected revenues after interest on the debt, entitlement outlays, and non-defense discretionary outlays are set according to the assumptions described above. This assumption reflects the current consensus on budget policies, in which defense is the only account where nominal and real reductions in spending are deemed acceptable.

Based on the assumptions described above, Heritage Foundation analysts used the WEFA Mark XI macroeconomic model to estimate annual expenditures for interest on the national debt, entitlements, non-defense discretionary spending, and defense for the next 21 years. In order to meet our assumption of continuously balanced budgets, any difference between revenues and outlays is met in every case by reducing federal outlays for national defense.

A number of WEFA forecasts were used to project various aspects of federal spending, such as interest payments on the federal debt. However, it is important to note that Heritage economists constructed the projection of defense spending contained in this study; it does not come from a formal simulation of the WEFA U.S. model. "Simulating" the WEFA model would have meant changing federal spending levels and allowing the model to calculate new values for, say, interest payments on the federal debt. The reason such a "run" or simulation of the model was not performed is that the reduction in defense spending implicit in current budget priorities is so severe as to be outside the parameters of the WEFA model, making the results of a dynamic simulation highly suspect. Defense projections in this study, therefore, are static estimates, but good first approximations of the likely outcome of adhering to current law policies. (For a complete description of the analysis, see the Technical Appendix.) If a fully dynamic analysis could be undertaken, however, it is likely that a deficit would appear and that it would worsen over the forecast range. Therefore, if defense spending continued to be a "residual"--or the difference between total revenues and all other non-defense spending--it would reach zero more rapidly, and thus prior to the year 2020.

The Long-Term Result: The Defense Budget Eventually Would Be Crowded Out Entirely by Other Priorities.

Holding today's fiscal policy constant, using the assumptions noted, produces the following results:

  1. Interest payments on the national debt increase. By the end of the 21-year period, these payments will approach $450 billion annually, or 10 percent of the federal budget.

  2. Entitlement expenditures increase sharply between 1998 and 2020, rising to $3 trillion in fiscal year 2020, or 66 percent of the federal budget.

  3. Non-defense discretionary spending also increases, but at a more modest rate than entitlements. By 2020, non-defense discretionary spending is at $1.1 trillion, or 24 percent of the federal budget.

  4. Defense spending gradually declines over the 21-year period until no money is available for this basic function of government in the year 2020 (see Chart 9).

Note: It is important to recognize that this projection is not a prediction that the defense budget actually will go to zero by 2020. Rather, it is designed to show that this will be the "default" outcome if current and future administrations and Congresses hold to existing budget policies. Avoiding this result would require a significant shift in budget priorities. In effect, this projection is similar to several projections indicating that, under current policies, Medicare and Social Security eventually will be unable to pay benefits. In these cases, as in defense, it is inconceivable that Congress would allow this to happen. What all such projections provide is a warning that policy changes are needed.

--Baker Spring is a Senior Policy Analyst in The Kathryn and Shelby Cullom Davis International Studies Center at The Heritage Foundation.

--John S. Barry, during the preparation of this study, was a Policy Analyst in Economics at The Heritage Foundation.

Technical Appendix

This budget analysis makes use of the Mark XI Quarterly Macroeconomic Model created and maintained by WEFA Inc.5 Specifically, Heritage analysts used WEFA's September 1997 U.S. long-term forecast of over 1,600 economic variables for the period 1998-2020, including more than 25 variables that measure federal tax and spending priorities.6

The WEFA model is maintained to forecast the most probable economic conditions given historical trends, public policy, and likely policy changes. Therefore, in its September long-term forecast, WEFA assumes that payroll taxes earmarked for the Medicare Hospital Insurance (HI) trust fund will be increased 136 basis points by 2020 (evenly split between employers and employees). WEFA also assumes that payroll taxes earmarked for the Old-Age and Survivors Insurance (OASI) and Disability Insurance (DI) trust funds will be increased 348 basis points (evenly split between employers and employees) by the year 2020. In both cases, the increase in tax rates is phased in beginning in 2002. In other words, WEFA diverges slightly from current law in order to forecast the most likely economic outlook over the next 21 years.

However, this analysis depends on current law to determine what would happen to defense spending if no reform of entitlements is enacted and the commitment to a balanced budget is maintained. Therefore, Heritage analysts reset all payroll tax rates to their current law level. In light of these changes and others discussed above, Heritage analysts made very limited use of the WEFA model. Specifically, Heritage analysts excluded all stochastic variables in the WEFA model except those relating to the federal government's taxing and spending policies. That is, the analysis assumes no change in current law with respect to federal taxation and entitlement programs. Non-defense discretionary spending is maintained at the level forecast in the WEFA September long-term forecast. Interest payments on the debt (the level of which is held constant after 1999) are assumed to respond to WEFA-forecast interest rates.

Therefore, the analysis in this study can be considered "static" because it does not assume any macroeconomic response to a change in current-law federal tax and spending priorities. In other words, relevant data were extracted from the WEFA model for calculation in a Heritage-created spreadsheet model. Specifically, Heritage analysts decreased defense spending to accommodate (i.e., eliminate) forecast budget deficits. Thus, it is assumed that Congress and the President maintain a balanced budget policy by reducing only defense spending.


1. More recent estimates by the CBO indicate that the federal budget will be in surplus this fiscal year.

2. A no-real-growth path, in this context, means that spending will increase only as a result of inflation. For example, if inflation is 4 percent, "no-real-growth" would imply that spending increases by 4 percent.

3. The social insurance trust funds are the Old-Age and Survivors Trust Fund, the Disability Insurance Trust Fund, and the Health Insurance Trust Fund.

4. The Medicare trust fund is projected to be exhausted by 2007, and the Social Security trust fund is projected to be exhausted by 2019.

5. The estimates in this paper were developed by The Heritage Foundation using the Mark XI U.S. Macro Model of WEFA, Inc. (formerly Wharton Econometric Forecasting Associates). Nobel Prize-winning economist Lawrence Klein and several of his colleagues at the Wharton Business School of the University of Pennsylvania developed the model in the late 1960s. The model is used widely by Fortune 500 companies, prominent federal agencies, and economic forecasting departments. The methodologies, assumptions, conclusions, and opinions herein are entirely the work of Heritage Foundation analysts. They have not been endorsed by, nor do they necessarily reflect the views of, the owners of the model.

6. These estimates use budget numbers based on the National Income and Product Accounts (NIPA) accounting system, which is maintained by the Department of Commerce. NIPA accounts do not conform to the functional accounts Congress uses to set fiscal policy. Further, the Bureau of Economic Analysis, in calculating NIPA, uses a different accounting methodology than the Congressional Budget Office uses to calculate outlays. The reason for these differences stems from the fact that NIPA is intended to measure the impact of government activity on the economy as a whole. Heritage analysts used NIPA because the long-range economic and budget forecasts produced through the WEFA Mark XI computer model are based on NIPA figures. Despite the differences in accounting and accounting methodologies, the forecasts provided in this paper are accurate reflections of long-term trends for the federal budget under existing policy.

About the Author

Baker Spring F.M. Kirby Research Fellow in National Security Policy
Douglas and Sarah Allison Center for Foreign and National Security Policy