In their annual debate about the defense budget, Members of Congress focus primarily on whether the President's budget request will meet the military's immediate spending needs. But programs to develop weapon systems often run for a decade or more before those systems are fielded, and other policy decisions have long-term implications; thus, decisions made today can influence the size and composition of the nation's armed forces for many years to come. Recognizing the need for a longer view, the Senate Appropriations Committee's Defense Subcommittee requested that the Congressional Budget Office (CBO) analyze the long-term implications of the Administration's current plans for defense. This analysis examines those implications both for budgetary resources and for ages and inventories of weapon systems.

In the five years from 1997 to 2002, the annual U.S. defense budget grew from $274 billion to $345 billion. (All dollar amounts in this study represent total obligational authority expressed in 2002 dollars.) The defense program outlined by the Bush Administration for fiscal year 2003 and the following four years (the 2003 Future Years Defense Program, or FYDP) anticipates additional growth, with the defense budget averaging $387 billion over the 2003-2007 period and reaching $408 billion in 2007. If that program continued as currently envisioned, the demand for defense resources would continue to increase through 2012, CBO projects, and would average $428 billion a year between 2008 and 2020. Costs for day-to-day operations (running units, maintaining equipment, and providing pay and benefits to military personnel) would grow from $222 billion in 2002 to more than $280 billion by 2020. Demands for investment resources (primarily to develop and purchase new equipment) would rise from $110 billion in 2002 to $164 billion in 2012 and then decline to about $134 billion by 2020.

Those projections are based on the Department of Defense's (DoD's) current cost estimates for a host of defense programs and activities. CBO also projected long-term resource demands if costs for weapons programs and certain other activities grow as they have historically (a case it called cost risk). In that case, the annual cost of current defense plans would average $398 billion over the 2003-2007 period and could later reach $488 billion at its peak (see Summary Figure 1). In either case, future resource demands would be higher than defense spending has been at any time in the past 22 years--exceeding the peak of $421 billion in 1985--and would need to remain at such levels for a decade or more.
 

Summary Figure 1.

Past and Projected Resources for Defense

Source: Congressional Budget Office using data from the Department of Defense.

Notes: FYDP = Future Years Defense Program; RDT&E = research, development, test, and evaluation.

Portions of the budget titles for family housing and military construction pay for support activities, whereas other portions pay for investment. Thus, it is not possible to allocate all of the funding in each of those titles to either category.

Viewed in relation to the size of the economy, however, future demands for defense resources could be less than past spending. The share of gross domestic product (GDP) dedicated to defense declined from an average of 6 percent in the 1980s to an average of 4 percent in the 1990s (see Summary Figure 2). If current defense plans were implemented, defense spending would drop to 3 percent of GDP by 2007 and to 2 percent by 2020, assuming that GDP grows at the rates projected by CBO. (Because of the size of the economy, those levels are the same for CBO's projections with or without cost risk.)
 

Summary Figure 2.

Defense Spending as a Share of the Federal Budget and GDP

Source: Congressional Budget Office using data from the Department of Defense.

Note: FYDP = Future Years Defense Program; GDP = gross domestic product.

As a share of the federal budget, defense spending declined from an average of 28 percent in the 1980s to 16 percent in 1998. Since then, its share of the budget has increased to about 18 percent. Nonetheless, assuming that the rest of the federal budget continues to grow in real (inflation-adjusted) terms through 2020 at the rate that CBO projects in its current 10-year baseline, defense funding would compose a steadily smaller portion of the budget over the long run under current plans. In CBO's long-term projection of those plans, defense spending would fall to about 13 percent of the total budget by 2020 (15 percent using CBO's projection with cost risk), as the growth of mandatory programs for the elderly, such as Social Security and Medicare, and other entitlement programs, such as Medicaid, outpaced projected increases for the military. Those three programs are projected by 2030 to consume a substantial portion of what the federal government now spends on the entire budget. Thus, even if defense spending rose substantially over the long term, it would compose a smaller share of a rapidly increasing total federal budget.

In short, under CBO's projection of the Administration's current plans, defense resources could reach levels greater than those sustained during the Cold War. However, the fraction of the country's economic activity that they constitute, and their share of the total federal budget, could be significantly smaller than during the Cold War.
 

The Administration's Plans for Defense

CBO's estimates are current-policy projections, which assume that the policies and plans underlying the Bush Administration's 2003 defense program do not change in future years. That program is described in detail in DoD's 2003 FYDP, a report submitted with the President's 2003 budget request that presents DoD's estimated appropriation needs through 2007. Further detail comes from committee staff backup books, which DoD provides to the Congress to support its budget request for various accounts, and from Selected Acquisition Reports, which are submitted to the Congress for certain major defense programs. General descriptions of the Administration's current policies and plans are contained in the Quadrennial Defense Review Report from September 2001 and the Annual Report to the President and the Congress submitted by the Secretary of Defense in 2002.

As described in those sources, the Administration's current policies stress the importance of transforming the nation's military forces, but they make few substantial changes to those forces or to the defense programs that will sustain them in the long term. DoD describes transformation as shifting the composition of the armed forces from traditional approaches to new approaches. Those new approaches are intended to incorporate advanced technologies that will save money while retaining or improving combat capability to meet the threats foreseen for the 21st century. The Administration plans to initiate or increase funding for a number of programs that it associates with transformation, including space-based radar satellites, missile defenses, unmanned combat air vehicles, unmanned reconnaissance aircraft, small surface combat ships for the Navy, advanced-technology combat vehicles for the Army, and precision munitions.

However, the 2003 FYDP would also continue to fund nearly all of the major defense programs inherited from the previous Administration (except the Army's Crusader self-propelled howitzer, which was terminated in the spring of 2002). In addition, the Administration has not announced any major changes to military force structure--the number and composition of Army divisions, Air Force tactical fighter wings, Marine Corps expeditionary brigades, Navy carrier battle groups, and so forth--through 2007. Although transformation might enable DoD to make significant changes to its major investment programs or force structure in the longer term, programs started now are unlikely to provide operational capabilities for 10 to 15 years, if not longer.

Thus, in CBO's projection of current plans, demands for defense resources increase in the long term for three reasons: the transition from development to production or increasing production for a number of existing programs; continued growth (even without cost risk) in the costs to operate and sustain future forces, which are assumed to be essentially the same size as today's forces; and continued development and eventual production of those few new programs associated with transformation that are included in the Administration's current plans.

CBO projected the resource demands associated with current plans--but it did not evaluate the basis for those plans, the necessity of pursuing them, or possible alternative plans. For example, CBO has not considered alternatives to the national security strategy or to the defense strategy developed during the 2001 Quadrennial Defense Review. It has not assessed alternative plans that might achieve those strategies or examined their resource implications. In addition, CBO has not attempted to evaluate the efficiency of the way in which DoD does business or the need for a particular infrastructure to support combat forces. Those are all important issues that could have a major impact on future defense costs. But those issues would be associated with changes to current policies and plans and thus are excluded from CBO's current-policy analysis.
 

Resource Demands for Operation and Support: Paying for Day-to-Day Operations

Although investment in new weapon systems tends to get more attention, about two-thirds of the defense budget is devoted to operation and support (O&S) funding--a category that covers almost all of DoD's current operating costs. In 2002, O&S funding totaled $222 billion ($218 billion will be available for O&S in 2003).

Some 60 percent of O&S funding is appropriated to operation and maintenance accounts, which pay for the salaries of DoD's civilian employees, the services of contractors, and purchases of fuel, spare parts, and other goods routinely consumed by the military (see Summary Figure 3). Some of the activities funded by the operation and maintenance accounts, such as training for individuals and units, support combat capabilities directly, whereas other activities, such as central headquarters, support combat indirectly.
 

Summary Figure 3.

Past and Projected Resources for Operation and Support

Source: Congressional Budget Office using data from the Department of Defense.

The remaining 40 percent of O&S funding consists of appropriations to DoD's military personnel accounts. That funding covers the pay and allowances received by active-duty and reserve personnel as well as contributions to their retirement fund. Beginning in 2003, personnel spending also includes contributions to a fund to provide supplemental medical benefits for current service members when they have retired and become eligible for Medicare.

If the Administration's current defense plans were unchanged over the long term, annual O&S costs would grow by 14 percent between 2007 and 2020, CBO projects, rising from $245 billion in 2007 to $280 billion. The main sources of that projected growth are wage increases and rising medical costs.

As noted above, CBO's projection assumes that the size of the military remains constant from 2007 through 2020. It also assumes that military and civilian pay grows at the same rate as the employment cost index (a measure of wage increases in the private sector produced by the Bureau of Labor Statistics). Medical costs for active-duty personnel are assumed to grow at the annual long-term rate projected by the Department of Health and Human Services for growth in civilian health care costs (3.6 percent). Medical costs for retirees are assumed to grow at an annual rate estimated by DoD's independent board of actuaries and included in the 2003 FYDP (4.0 percent).

CBO's projection of the Administration's plans excludes certain risks of additional cost growth, such as potential costs that are not fully funded in current plans or elements of those plans that experience suggests may not come to pass. Those sources of risk include the cost of maintaining existing equipment as it ages and the cost of maintaining new, more-complex equipment as it is acquired and fielded. Another source of risk is that the round of base realignments and closures that the Administration has proposed for 2005 will not proceed as planned and thus not produce the anticipated savings in O&S costs.

If those and other cost risks are taken into account, annual operation and support costs could rise by an additional $26 billion, CBO projects, reaching $306 billion by 2020. That increase would represent growth of 25 percent from the 2007 level rather than the 14 percent increase without cost risk. Of the additional $26 billion in the cost-risk projection, about $16 billion would come from increases in the cost of maintaining older equipment and new, more-complex equipment, and $5 billion would reflect forgone savings from the anticipated 2005 round of base closures.

In CBO's projection including those risks, the introduction of new equipment does not offset the growth in maintenance costs that would otherwise occur if old equipment were retained. Although increases in costs because of equipment aging and the introduction of more-complex equipment do not dominate total projected growth in O&S costs, they are the largest component of O&S cost risk considered by CBO. The reason is that those costs are some of the largest that are not fully funded in DoD's current plans.¹

Other risks of cost growth may exist that CBO has not considered. Such additional risks would most likely be associated with changes to current policy, such as the provision of additional medical benefits to military personnel or increases in force structure to meet increased commitments overseas. Those sorts of risks were not considered in CBO's projection, which is based on current policies and plans. (In addition, there are some projections of growth in U.S. health care costs that, at least in the near term, greatly exceed the rates that CBO used in this analysis.  If those higher rates were realized over the long term, they would represent a cost risk associated with current plans that CBO has not considered.)
 

Resource Demands for Investment: Paying for a Modern Force

DoD currently devotes about one-third of its budget to researching, developing, testing, and buying weapon systems. In 2002, investment funding (which this analysis defines as the sum of accounts for research, development, test, and evaluation and for procurement) totaled $110 billion. (About $128 billion will be available for investment in 2003.)

Under the 2003 FYDP, the Administration plans to increase funding for investment by about 5 percent per year, reaching $142 billion by 2007. If the Administration's plans were carried out over the long term, resource demands for investment would continue to grow--peaking at about $164 billion in 2012, CBO projects. Unlike operation and support, which would increase continuously through 2020, investment would decline after 2012 as the purchases projected to begin just beyond 2007 are completed. Annual investment would fall to about $134 billion by 2020, or $8 billion less than the amount planned for 2007. Over the entire 2002-2020 period, resource demands for investment would average $144 billion a year--about $20 billion more than average annual funding during the past 22 years (see Summary Figure 4).
 

Summary Figure 4.

Past and Projected Resources for Investment

Source: Congressional Budget Office using data from the Department of Defense.

Note: FYDP = Future Years Defense Program; RDT&E = research, development, test, and evaluation.

Those projections are based on DoD's current estimates for the costs and numbers of weapon systems that it expects to buy. If costs for such systems exceeded current estimates to the extent that they have historically, the resources needed to carry out current investment plans could rise to $156 billion by 2007 and top $190 billion in 2012, CBO projects. In that case, investment demands through 2020 would average about $165 billion--or $41 billion more per year than the average investment funding provided since 1980.

Aging Equipment
In addition to projecting the long-term cost of the current defense program, CBO examined whether planned purchases would support DoD's planned forces and halt the aging trends that are evident for many types of military equipment since the armed forces were reduced after the Cold War. CBO projected the size of DoD's equipment stocks through 2020 and their average age if currently planned purchases occur. A useful indicator of whether the age of equipment stocks is an issue is how close that age is to one-half of the equipment's expected service life (what this analysis calls the half-life). An average age well above the half-life generally implies an inventory with many pieces of equipment that might soon have to be replaced or refurbished over a short span of time.

If DoD received enough funding through 2020 to execute its current plans, as projected by CBO, aging trends would eventually be halted or reversed for many types of weapons, including Army and Marine Corps combat vehicles and helicopters (see Summary Figures 5 and 6). In addition, current plans would sustain the Navy's fleets of ships and aircraft at average ages that the service considers acceptable. However, average ages would remain a concern for Air Force weapons such as bombers, airlifters (transport aircraft), and tankers.
 

Summary Figure 5. (Revised 1-13-03)

Average Age of Selected Army and Air Force Weapon Systems

Source: Congressional Budget Office using data from the Department of Defense.

 

Summary Figure 6.

Average Age of Selected Navy and Marine Corps Weapon Systems

Source: Congressional Budget Office using data from the Department of Defense.

Of course, to achieve those results, purchases and procurement budgets would, in many cases, have to match or exceed the levels seen during the defense buildup of the 1980s. Future budget pressures--arising from growth in other federal programs, such as Social Security and Medicare, as well as from continued growth in the military's O&S costs--might make it difficult to sustain those levels of procurement spending.

Steady-State Procurement
Another way to measure the adequacy of planned purchases is to compare those purchases with steady-state levels (the procurement needed to sustain planned forces indefinitely).² CBO estimated the annual level of steady-state purchases by dividing a planned inventory of weapon systems by the expected service life of those systems. Multiplying those steady-state purchases by the estimated unit costs of various systems (either DoD's estimates or those estimates plus historical cost growth) yields an overall projection of annual steady-state procurement costs (see Summary Table 1). In making that calculation, CBO used two alternative estimates of service lives. One assumes DoD's current projections for service lives, which are generally longer than those planned for during the Cold War. The other uses the shorter service lives incorporated in DoD's Cold War-era planning.

Summary Table 1.
Steady-State Procurement Costs Under a Variety of Assumptions
(In billions of 2002 dollars of total obligational authority)
		Using DoD's Cost Estimates		Assuming Historical Cost Growth
Steady-State Procurement Costs for Currently Planned Forces
	Assuming long service lives	85		99
	Assuming short service lives	111		130
Average Annual Procurement Costs to Carry Out Current Plans
	2003-2007 (As requested in the FYDP)	77		81
	2008-2020 (As projected by CBO)	101		121
Source: Congressional Budget Office.
Notes: FYDP = Future Years Defense Program.
In the long-service-lives assumption, the Department of Defense keeps weapons as long as it currently projects, which is generally longer than it has in the past. In the short-service-lives assumption, the Department of Defense operates systems only as long as it has previously.

In the past, the combination of new military equipment funded during the Reagan Administration and the drawdown in U.S. military forces that occurred after the Cold War enabled DoD to sustain its forces with much smaller annual equipment purchases. But during the 1990s, those purchases fell below the steady-state level, and as a result, the average age of many types of weapons increased, as shown in Summary Figures 5 and 6.

DoD's planned procurement budgets for the 2003-2007 period, as outlined in the current FYDP, are below CBO's estimates of the steady-state procurement costs needed to sustain currently planned forces. That is why average ages generally rise through 2007 in Summary Figures 5 and 6. But if current procurement plans were carried out through the 2008-2020 period, procurement would be within the steady-state range (between the estimates for long and short service lives), CBO projects. That is why average ages for many weapon systems decline beginning around 2010 in Summary Figures 5 and 6.

As noted earlier, the Bush Administration has stated that it expects DoD's forces to ultimately be transformed. The programs associated with transformation that the Administration has just started will not provide substantial numbers of operational weapons for 10 to 15 years. Therefore, transformation is unlikely to be fully realized before the 2020 endpoint of CBO's projection.

Whenever it occurs, transformation could take many forms, depending on the success or failure of concepts that have only begun to be explored. Research on unmanned aircraft, advanced-technology combat vehicles, and smaller ships has started. In addition, the military services and the U.S. Joint Forces Command are experimenting with new ways of organizing and operating forces that might be possible if such systems can be developed. If those (or other) concepts succeed, military forces could eventually be very different from the ones assumed in DoD's current plans and in CBO's year-by-year projections. And if the cost goals now asserted for those systems can be achieved--which would be at variance with past experience--steady-state procurement costs might eventually be lower.

To assess the potential impact of transformation on future procurement needs, CBO estimated steady-state procurement costs for one example of how a transformed force might look. The example that CBO chose assumes that many of the transformation initiatives now being undertaken by the Bush Administration ultimately succeed and are fully embraced by the military services. Thus, CBO's hypothetical transformed force uses unmanned aircraft to replace today's manned systems for both ground-attack and reconnaissance missions. It also assumes that many ships can be operated with two alternating crews or from forward bases, which would allow a smaller fleet of surface combatants to perform the current level of peacetime missions, and it assumes that the Navy converts additional nuclear-missile-carrying submarines to perform conventional missions. Such a military force would be a radical departure from today's force--and is unlikely to be realized within the period of CBO's projections.

CBO's estimate of steady-state procurement costs for a transformed force suggests that DoD might eventually be able to reduce its annual procurement needs below the level associated with current plans if it can ultimately change those plans significantly. For the particular example of transformation examined by CBO, procurement savings could equal 10 percent to 15 percent compared with the annual steady-state costs of current plans shown in Summary Table 1.

Of course, those estimates hinge on the particular assumptions that CBO made about the composition of a transformed force. Larger or smaller savings--or higher costs--might result depending on how transformation actually occurred. Although CBO's transformed force is radically different in composition from today's force, it contains about the same total numbers of aircraft and other major weapon systems. If transformation enabled DoD to substantially reduce its number of weapon systems, greater procurement savings than CBO projects might be possible. (Likewise, if transformation allowed forces to be cut or significantly changed peacetime operating concepts, savings could accrue in O&S costs as well as in procurement. CBO did not analyze the potential O&S savings that might result from transformation, which could be substantial.)

Conversely, savings could be much less if--as happened in the past--the currently hoped-for cost goals of new systems cannot be met. Such goals are generally very optimistic relative to historical experience with cost growth in DoD systems. For example, the Global Hawk unmanned reconnaissance aircraft now in operation was originally intended to cost $20 million apiece. Actual costs have exceeded $60 million per aircraft.