The military activities of the Department of Defense and the atomic energy activities of the Department of Energy (DOE) constitute most of the spending in function 050, which, after declining at the end of the Cold War, began to rise again in the late 1990s. Between 2002 and 2006, discretionary outlays rose from $349 billion to $520 billion (an increase of 49 percent). Some of that increase is attributable to operations in Iraq and Afghanistan and to activities related to the war on terrorism. So far, for 2007, function 050 is funded at $522 billion, including $70 billion for operations in Iraq and Afghanistan and for the war on terrorism. Further funding for such activities is likely to be provided in a mid-2007 supplemental appropriation.

Most components of defense spending have increased in recent years. Spending on pay and benefits for military personnel grew by 44 percent between 2002 and 2006, and spending for operations and maintenance—to meet many of the military's day-to-day costs—rose by 57 percent. (Most of the costs associated with military operations in Iraq and Afghanistan fall into those two categories.) Spending for procurement and for research and development of weapons systems and munitions also increased, from $107 billion in 2002 to $158 billion in 2006. Spending on DOE's atomic energy activities rose from $14 billion in 2002 to $16 billion in 2006.

The Army regards the Future Combat System (FCS) program as the cornerstone of its effort to transform itself into a force that can deploy combat units to respond quickly to crises anywhere in the world. With its current tanks and other armored vehicles, the Army typically would take three to four weeks to deploy a brigade to a remote location in Africa, Asia, or Eastern Europe. As envisioned, the next generation of combat vehicles that the FCS program would develop would be as lethal and as survivable as current weapons are but weigh as much as two-thirds less and require less fuel and other logistics support. The Army would develop eight new combat vehicle models as well as new unmanned aerial and ground vehicles, sensors, and munitions—all linked by advanced communications networks into an integrated combat system. The Army's fiscal year 2007 budget plan shows costs from 2008 through 2023 for the first FCS increment (to equip slightly more than one-third of the active Army's combat brigades) that could approach $150 billion. The 2008 plan may include less for FCS over the period, and it could develop fewer systems, equip fewer brigades, or both.

This option would cancel the FCS program and invest more in existing heavier combat vehicles that also have a proven record of utility. It would preserve a residual research and development effort for promising technologies that could be added later to existing systems. The option would expand the Army's programs for upgrading Abrams tanks, Bradley fighting vehicles, M113 armored personnel carriers, and M109 self-propelled howitzers—many purchased in the early 1980s—to keep those vehicles in service for another 20 years. Cancelling the FCS would reduce the need for Army budget authority for research and development and for procurement by a total of $23 billion over the next five years, but upgrading current systems would require about $7 billion in budget authority over the same period. On net, the need for budget authority would decline by about $16 billion between 2008 and 2012 and by $62 billion over 10 years.

The feasibility of the FCS program has been questioned by defense experts and by the Government Accountability Office. Many analysts have concluded that current technology does not permit the construction of light-weight combat vehicles that match or surpass current vehicles in reliability and invulnerability to enemy weapons. Furthermore, the Army's experience in Iraq suggests that its strategy for making lightly armored vehicles equally as survivable as the heavily armored Abrams tank may not be feasible. To achieve comparable survivability, U.S. combat vehicles would avoid being targeted by exploiting superior knowledge of enemy activities. The threat in Iraq has come primarily in urban settings from individually launched weapons, and the ability to identify attackers' locations may be beyond any technology now envisioned.

The primary argument against this option is that canceling the FCS program might preclude transforming the Army in any meaningful way. It would mean a significant portion of the Army would continue to use systems originally developed in the 1980s or earlier. Some of those weapons, notably the Abrams tank, are fuel inefficient and maintenance intensive. Improving the data processing and connectivity of those older systems would require the sometimes-difficult process of integrating newer components into old frames. Finally, retaining old systems might eventually lead the Army to lose its technological edge and military dominance.

The Army currently has 10 active and 8 reserve divisions, most of which include 4 maneuver combat brigades. In addition, the Army has independent combat brigades, which are not part of any division, and armored cavalry regiments that are similar to separate brigades. In all, the Army had 42 active combat brigades and 28 reserve combat brigades planned for 2007. The service draws on those forces for combat or for peacekeeping missions. Almost all other Army units are intended in some way to support those combat brigades and divisions.

Since the mid-1990s, the Army has been increasingly called upon to keep combat brigades deployed overseas for commitments that have included operations in Bosnia, Kosovo, Kuwait, Afghanistan, and Iraq. To keep forces deployed overseas while preserving high levels of training and readiness, the Army rotates units through those operations. Thus, the more commitments the service has, the more often any unit (and any soldier) can expect to be deployed.

This option would increase the Army's force structure by two divisions, or an additional eight combat brigades. One division would be a heavy mechanized infantry division; the other would be equipped with Stryker medium-weight armored vehicles. This option also would create support units that the new divisions would rely on in combat—corps support groups, artillery brigades, engineer battalions, truck companies, and the like. Some of those units would be part of the Army Reserve or National Guard. To man the units, the active Army's authorized end strength would be increased by 50,000, and the reserve component's end strength would be increased by 30,000.

The Army's recent reorganization into modular combat units and a robust program of remanufacturing its armored vehicles in recent years may have provided the Army with enough M1, M2, M3, M109, and M113-series armored vehicles to create the new units in the heavy division without purchasing additional vehicles of those types. If that were the case, fully recruiting, organizing, equipping, and training all of those new units would take about five years, the Congressional Budget Office estimates, and would require about $51 billion in budget authority over that period. If, however, the Army needed to purchase entirely new equipment, it would require additional budget authority of $55 billion over the same period. (Somewhat less than half of the $55 billion would be for procurement of new equipment in the first five years; the remainder would be for the recurring costs of personnel and for maintaining the new units.)

The main argument for this option is that the current Army may be too small to execute all of its assigned missions. The service's peacetime commitments have increased since the mid-1990s, especially since the beginning of the war on terrorism. When the Army must sustain significant overseas deployments, individual soldiers are separated from their families for long periods, units cannot maintain the training schedule the Army expects, and equipment is degraded by the stress of heavy use (and, in some cases, by exposure to harsh environments). Some proponents of adding two new Army divisions suggest that the pace of deployments has exacerbated those problems to an unacceptable extent and that the only way to slow deployment and preserve readiness is to add forces to the service. In the absence of new active-component divisions, the Army would need to mobilize and deploy more reservists, increasing stress on reserve-component units and personnel. Some defense experts argue that it is inappropriate to regularly mobilize and deploy reserve-component units, that the active Army should be large enough to handle peacetime commitments, and that the reserve component should be used only in exceptional cases.

An argument against this option is that the cost and time needed to increase the size of the Army's combat forces could make the addition of two divisions a poor response to what may be temporary pressures. Although the need to maintain large forces in Iraq has placed considerable stress on the active Army, that burden might be reduced before the new divisions become fully available in 2012. Increasing the force structure also would carry large long-term fiscal obligations, which could extend for many years after this option was enacted.

The Navy's proposed new guided-missile destroyer, the DDG-1000 Zumwalt-class (formerly DDX), is designed principally to provide volume fire support to Marine Corps units conducting operations ashore, although it will be able to perform other missions. Displacing some 14,500 tons, it will be larger than any other surface combatant in the Navy, and it will carry 80 missiles and two 155-millimeter advanced guns to provide support up to 83 nautical miles away. In the long-term ship construction plan it sent to the Congress in February 2006, the Navy proposed buying seven DDG-1000s between 2007 and 2013 at a total cost of about $20 billion. The Congressional Budget Office estimates the total cost of those same seven ships at about $30 billion.

This option would have the Navy build only two DDG-1000s as technology demonstrators for the transition to the new class of cruisers, the first of which the Navy expects to order in 2011. Construction of the other five DDG-1000s in the Navy's plan would be canceled. To provide fire support to Marine Corps units, this option also would provide for five LPD-17 San Antonio-class amphibious ships, each modified to carry 16 vertical launch system cells and two Advanced Gun Systems (AGSs). The ship would have the same main battery as the Navy's DDG-1000, and its less expensive platform is already in production. This option would not lead to savings in 2008, but it would save nearly $4 billion in outlays through 2017. To generate additional savings, funding for two lead ships in 2007 could be combined into full funding for one ship, and a sixth DDG-1000 could be canceled.

Critics of the DDG-1000 have said it is too expensive for the amount of capability it will provide. Although the ship is expected to be stealthier than any surface combatant—thus able to operate closer to shore than other ships can—only three of the seven ships would be immediately available at any given time (the remainder would either be in maintenance or in use for predeployment training). Also, current surface combatants carry more long-range missiles than the DDG-1000s will. Aside from its guns (which the modified LPD-17s would have), the principal benefits of the DDG-1000 are its stealth and its new radar and combat systems, which will make it superior to other surface combatant ships at self-defense in the coastal regions where it will mostly operate. Advances in technology, however, might overtake that advantage as new or improved radar and combat systems are deployed. By incorporating the AGSs on the LPD-17 hull, the Navy could provide the same long-range fire support—a capability the service currently lacks—at much lower cost.

The disadvantage of modifying the LPD-17 to provide fire support is that the resulting ship would be less of a surface combatant than a gun platform capable of local self-defense. The ship would not match Zumwalt-class destroyers' stealthiness, it would not be able to embark helicopters, and it would lack a sophisticated combat suite for coastal operations. Such a vessel could be used only after the coastal waters had been made relatively secure by littoral combat ships or other surface combatants. Another disadvantage is that costs for the future cruiser could be higher because overhead rates at the commercial shipyards might rise as a result of producing fewer DDG-1000s.

Over the next seven years, the Navy plans to spend about $15 billion on a squadron of ships it calls the Maritime Prepositioning Force (Future), or MPF(F). Combined with several ships in the current fleet, the MPF(F) would allow the Navy to deploy a Marine expeditionary brigade to a hostile shore—and keep it supplied for almost three weeks—without seizing or establishing a land base. The Navy proposes to begin buying the MPF(F) in 2009 and to have the force operational by 2019 or 2020. The squadron would be an important component of Navy and Department of Defense plans for "sea basing"—an idea that is still evolving—which aims to increase the Navy's ability to respond to crises quickly, with a larger forcible-entry capability, and with more freedom of action than is currently possible.

This option would cancel the MPF(F) squadron, and nothing would be bought in place of those ships. The option would save $14 billion in outlays between 2008 and 2017. Some defense experts say its small benefit—the ability to transport and sustain one Marine brigade—will not justify its cost. In addition, at least six of the new ships, which would be built to less stringent commercial standards, would be more vulnerable to attack than are the Navy's amphibious warfare ships. The Navy would operate the MPF(F) along with amphibious ships in coastal areas where threats from enemy mines, antiship missiles, small boats, and submarines are more acute than they are on the open seas. Critics also argue that the technological challenges of deploying and sustaining a Marine brigade entirely from the sea will be insurmountable. Instead, the money would be better spent on traditional amphibious warships or on other equipment that could facilitate deployment of larger numbers of troops in hostile environments, albeit not as quickly as might be possible with the MPF(F) squadron.

The disadvantages of this option include disruption of the Navy's new shipbuilding plan. Senior Navy officials have identified stability in the shipbuilding program as a primary goal. In addition, this option would reduce, if not preclude, the Navy's ability to deploy substantial numbers of Marines ashore and to support them entirely from logistics ships at sea. Senior Navy leaders see that capability (and its concomitant freedom of action) as a paramount design objective for its new ships.

Canceling the MPF(F) squadron, however, does not necessarily translate to fewer ships being available for maritime pre-positioning. The Navy maintains three squadrons of ships overseas, each carrying the equipment needed by a Marine expeditionary brigade. To deploy those brigades, the Marines would be flown from the United States to converge with a ship at an established port where equipment would be unloaded. Under this option, the Navy would retain all three squadrons and the regular amphibious warfare ships in its fleet.

The F-35 Joint Strike Fighter program is the military's largest for aircraft development. In 2002, a team of manufacturers, led by Lockheed Martin, was awarded a contract to develop three versions of the stealthy aircraft: a conventional model for the Air Force; a longer range, carrier-based model for the Navy; and a short takeoff, vertical landing (STOVL) model for the Marine Corps. Navy and Air Force plans for 2008-2027 anticipate the purchase of about 2,400 F-35s, at a cost of about $230 billion, according to Bush Administration estimates. If research and development funding is included, more than $244 billion would be spent for the F-35.

This option would cancel the F-35 program and substitute upgraded fighter aircraft already in production: the Lockheed Martin F-16 Block 60 for the Air Force and the Boeing F/A-18E/F for the Navy and the Marine Corps. If those aircraft were purchased in the quantities and on the schedule currently planned for the F-35, this option would decrease outlays for development and procurement by $22 billion over the next five years, it would save $47 billion through 2017, and it would save $87 billion through the end of the planned program if each F-35 were replaced with an upgraded alternative fighter plane.

An argument in support of this option is that the new F-16 and F/A-18 aircraft—with upgraded radar systems, precision weapons, and digital communications—will be sufficiently advanced to meet the threats the nation is likely to face in the foreseeable future. The sophistication of the F-35 and the added technical challenges of building three distinct types of aircraft on a common airframe with the same engine model furthermore may result in costs substantially higher than current estimates would predict. In the past year alone, the cost estimate for the total F-35 program grew by about 9 percent. Experience suggests that additional growth is likely.

A disadvantage of this option is that F-16 and F/A-18 aircraft lack the stealth design features that will help the F-35 evade detection by enemy radar systems and thus enhance its safety in the presence of enemy air defenses. The armed services will maintain some stealth capability, however, with the B-2 bomber and F-22 fighter fleets and with planned development of new, highly stealthy unmanned fighters and long-range bombers. Also, substituting F/A-18s for the STOVL version of the F-35 (the F-35B) would make it impossible to include fixed-wing fighter operations from LHA and LHD amphibious assault ships of the Navy's Expeditionary Strike Group task forces—a capability the current AV-8B Harrier offers. The strike groups therefore would need to rely on armed helicopters (which lack the F-35's range, speed, payload, and survivability) or on the availability of other forces, such as aircraft carrier strike groups, for support.

Bush Administration plans call for the Department of the Navy to purchase a total of 680 F-35 Joint Strike Fighter planes in two variants. The Marine Corps will have the F-35B, a short takeoff, vertical landing (STOVL) aircraft; the Navy will have the F-35C carrier-based aircraft. (The Air Force's F-35A will be a conventional land-based fighter.) Although the F-35s have common design elements, there are substantial differences between them. The F-35B will have a lift fan, articulated engine nozzle, and special flight control systems for STOVL operations. The F-35C will have larger foldable wings and strengthened structures to withstand the particular demands of carrier operations. For 2008-2024, the Navy Department plans to spend about $7 billion to develop and $77 billion to procure its two versions of the F-35.

This option would maintain the Air Force's F-35A procurement as planned but cancel the F-35B and F-35C. Instead, the Navy and Marine Corps would purchase additional F/A-18E/F fighters that currently are in production. If those aircraft were purchased at the F-35's planned rates, this option would decrease outlays for development and procurement by $9 billion over the next five years and it would save $14 billion through 2017. These are net savings that account for the estimated cost increases that would be expected for the Air Force's F-35A as a result of lower total quantities and rates of production. (Higher unit costs for Air Force F-35s from 2018 through the end of the program in 2027 would reduce the total savings under this option to about $13 billion.)

An argument in support of this option is that the relatively new F/A-18E/F design (which has improved radar, weapon, and communication systems) is sufficient to meet likely threats. Continued development of the advanced technology required for the F-35, especially the powered lift systems in the F-35B, may cause costs to grow substantially beyond current estimates. In the past year, the cost estimate for the Navy's remaining share of F-35 program grew by about $8 billion, or 10 percent. Experience suggests that additional cost growth is possible.

A disadvantage of this option is that although the F/A-18E/F was designed to incorporate stealth features not found on older aircraft, it is nevertheless far less stealthy than the F-35. Canceling the F-35 could limit naval aviation operations early in a conflict, before enemy air defenses are suppressed. This shortcoming could be mitigated if the Navy can develop stealthy unmanned combat aircraft. (Achieving that, however, could present greater technical challenges than remain for the F-35.) Moreover, if the F/A-18 is substituted for the STOVL F-35B, the Marine Corps would not have any fixed-wing fighters to operate from its LHA and LHD amphibious assault ships in naval Expeditionary Strike Group (ESG) task forces or from austere locations ashore. (It can do so now with the AV-8B Harrier.) Absent support from carrier- or land-based aircraft, the ESGs would have to rely on armed helicopters, which lack the F-35's range, speed, payload, and survivability.

The Airborne Laser (ABL) program, managed by the Missile Defense Agency (MDA), is working to develop a system to destroy enemy ballistic missiles by means of a high-energy chemical laser carried on modified Boeing 747 aircraft. Its mission is to shoot down ballistic missiles during the boost phase, which occurs in the few minutes after launch and before a rocket's motors burn out. Initially, the ABL was envisioned as a defense against short-range theater ballistic missiles; now it is seen as a defense against short-, medium-, and long-range ballistic missiles.

The ABL program was started by the Air Force in 1996 and transferred to MDA in 2002. From 1996 to 2001, the Air Force invested almost $1 billion in the program; MDA spent an additional $2.4 billion between 2002 and 2006. MDA is continuing the program in a series of two-year blocks: 2004, 2006, 2008, and 2010. Block 2004 provided for the integration and initial testing of the first aircraft. Blocks 2006 and 2008 would continue testing the initial aircraft and focus on integrating the ABL into the larger Ballistic Missile Defense System. MDA's current plans include funding for the purchase of a second ABL aircraft, although MDA states that, in a knowledge-based strategy, those plans are contingent upon positive results from a shoot-down test scheduled for 2009.

This option would end the ABL program, immediately saving $330 million in outlays in 2008 and saving about $1.7 billion through 2011. Over the 2012-2017 period, the savings would be larger if the costs to develop, buy, and operate a fleet of ABL aircraft also are considered. In the absence of definitive current information from the Department of Defense (DoD) about technical characteristics, production quantities, and deployment schedules, the Congressional Budget Office cannot definitively estimate the annual future costs of buying and operating an ABL fleet. In earlier budgets, the Air Force indicated that it would deploy an operational ABL fleet by purchasing up to seven additional ABL aircraft at a cost of about $500 million each. DoD recently indicated that the cost of developing and building the first ABL aircraft would exceed $3 billion. Assuming that each aircraft would cost about $1.5 billion, the savings from discontinuing the development program and from forgoing the purchase of seven additional aircraft could exceed $10 billion between 2012 and 2017.

Supporters of this option argue that the technical problems, rising costs, and schedule delays encountered over the past eight years fuel doubt about the program's chances of success. If the ABL must operate closer to a missile's launch site, it may be vulnerable to enemy air defenses. Moreover, the ABL program is not the only one in MDA's broader Boost Defense Segment. MDA also has another new program that is developing a kinetic-energy hit-to-kill interceptor (KEI) that would be launched from land or sea to intercept ballistic missiles during their boost phases. Those interceptors are potentially more promising for boost-phase defenses because they are less technically challenging to develop than is the ABL system. Analysis also indicates that three to four aircraft would be needed to maintain a constant presence at a single location to defend against a potential enemy missile launch. One ABL aircraft would be on station; one or two would be in transit between the base and the orbiting location; and another would be at the base for refueling, reloading laser chemicals, and maintenance. In addition, the ABL aircraft might require air-refueling tankers, depending on where the aircraft were based. A single, fixed, ground- or sea-based interceptor battery could provide similar coverage at lower cost.

Opponents of ending the ABL program argue that although the ABL poses large technical challenges, it will provide a leap in the nation's ability to defend against ballistic missile attack. Furthermore, even if the boost-phase interceptor program proved a more viable alternative (and some observers argue that its potential need for multiple basing sites around a hostile country would limit its utility), the KEI would not be operational before 2010. Hence, any capability that the ABL might provide in the interim would be useful. The Air Force also points to significant progress in overcoming the ABL's technical difficulties and remains confident that it will be able to build a laser that can disable threats at long range.

The Space Tracking and Surveillance System (STSS), in development by the Missile Defense Agency (MDA), is planned as a constellation of low-Earth-orbit satellites to track enemy ballistic missiles and distinguish enemy warheads from decoys. The program grew out of an Air Force effort initiated in 1996 to develop the Space-Based Infrared System-Low (SBIRS-Low), satellites in low-Earth-orbit for detection and tracking of enemy missiles. SBIRS-Low experienced cost and schedule overruns; two satellites in the flight demonstration system were partly manufactured but subsequently placed in storage.

In 2000, the Congress directed the transfer of SBIRS-Low to the Ballistic Missile Defense Organization (now MDA). In 2002, SBIRS-Low was renamed STSS, and its development continues in a series of four two-year blocks—Block 2006, Block 2008, Block 2010, and Block 2012. In Block 2006, MDA is completing construction of the two satellites that had been partially manufactured under SBIRS-Low. Those satellites, slated for launch in 2007, are intended to demonstrate the ability to track ballistic missiles in flight and to distinguish warheads on those missiles from decoys. Block 2008 will upgrade the initial system's software; Block 2010's goals are classified. Current plans call for launching five satellites in Block 2012, the initial constellation, with three or more satellites possibly added later. The first launch, scheduled for 2012, is timed to replace flight demonstration satellites at the end of their service lives, although MDA indicates that launch could be delayed until 2014.

By the time STSS Block 12 is completed, MDA expects to have developed other deployable surface-based radars for missile defense, including the Sea-Based X-Band (SBX) and the Forward-Based X-Band Transportable (FBX-T) radar systems. By 2012, MDA plans to have upgraded the Cobra Dane, Beale, Fylingdales, and Thule Early Warning Radars to enhance the nation's ability to track ballistic missiles. The Air Force also expects to improve its missile-warning capability with the Space-Based Infrared System-High constellation. The first launch of a SBIRS-High GEO (geosynchronous) satellite is planned for 2009. The sensors on those satellites will be able to track ballistic missiles early in their flight.

This option would terminate Block 2012 of the STSS program and replace it with ground- and sea-based radars. House Report 107-298 refers to an internal Department of Defense (DoD) study that "indicates that ground based radars not only provide a viable alternative to a space based system, but also provide this capability at significantly lower cost and risk."

To estimate the savings from canceling STSS Block 2012, the Congressional Budget Office has assumed that the initial STSS constellation would follow current DoD plans for five satellites and would subsequently be expanded to nine. Based on the expected capabilities of STSS satellites and DoD's estimate of the satellite mass, CBO estimates that each would cost $700 million (in 2007 dollars). Thus, CBO estimates that canceling STSS Block 2012 would save about $4 billion over the next five years and about $14 billion over a decade. The 10-year savings would come from not starting Block 2012 research and development (about $6 billion), from not buying and launching the new satellites (about $8 billion), and from not operating the constellation (about $100 million). However, MDA would still be able to use the demonstration satellites for technology testing and for gathering data from a planned series of tests. If DoD decided subsequently not to deploy a constellation of operational STSS satellites, more than half the savings CBO estimates would not be accrued.

In place of STSS, this option would provide for one additional SBX and four additional FBX-T radars (the same radars currently being purchased for the Ground-Based Missile Defense System). Because STSS is a space-based system, it offers global coverage (albeit with potential gaps). Although SBX and FBX-T have more limited range, they can be deployed to any region of concern because they are mobile. Nonetheless, the number of radars assumed by this option would not replace the capability of a nine-satellite STSS constellation. It would provide more limited regional coverage of ballistic missile threats than would STSS.

To estimate the cost of the SBX and FBX-T, CBO examined procurement expense for the initial versions of those radars. CBO estimates that one SBX costs $1 billion and that one FBX-T costs $200 million. CBO assumed the radars would be purchased in 2012 and 2013. Combining the two parts of this option, CBO estimates that the net savings over the next five years would be $2.8 billion in outlays and net savings over the next 10 years would total $11.7 billion.

An advantage of this option is the significant savings from not developing and acquiring the full constellation of STSS satellites. That constellation might not be needed because programs that MDA and the Air Force plan to operate simultaneously with STSS also would provide some ability to track and discriminate ballistic missile warheads. This option would augment that capability with additional ground-based radars, which may be more effective than the sensors in the STSS satellites for that purpose.

An argument against this option is that the STSS flight demonstration system could validate the use of space-based infrared sensors for tracking and discrimination of warheads launched on enemy ballistic missiles. Although technical issues associated with the STSS sensors remain to be solved, use of ground-based systems for discrimination also poses technical challenges. Moreover, ground-based radars cannot match the global coverage offered by a full constellation of STSS satellites. The Air Force's SBIRS-High GEO program also has experienced cost growth and schedule delays, and its capability would be insufficient for tracking ballistic missiles throughout all phases of flight.

The Ground-Based Midcourse Defense (GMD) Block 2004 segment of the Ballistic Missile Defense System had two components, a test bed and an operational segment. Among other elements, Block 2004 included interceptor missiles based at Fort Greely, Alaska, and Vandenberg Air Force Base, California; detection and tracking radars located around the United States; battle management command-and-control software; and a communications system used to relay information to and from the interceptors in flight. The Block 2004/2006 segment continued development and fielding of those capabilities, resulting in December 2005 in the completion of the Initial Defense Capability (IDC). Future block developments would extend the system beyond the IDC by providing more interceptors and radars and expanding GMD to a third ground-based interceptor site.

This option would cancel the development of the block upgrades to the GMD system after the Block 2004/2006 effort. The option would continue to operate the interceptors at the two sites and would spend about $300 million a year to develop improvements to the initial capability. This option would cancel additional interceptor missiles and development of a third ground-based interceptor site currently planned for later blocks. The Congressional Budget Office (CBO) estimates that this option would save $1.2 billion in outlays in 2008 and nearly $13 billion between 2008 and 2017. Although the Administration has provided no detailed information on its post-2011 spending plans, CBO's estimate for this option assumes that spending from 2012 to 2017 to operate and continue development of a three-site GMD system would be consistent with the 2007 Future Years Defense Program. If the Department of Defense subsequently changes those plans and decides not to pursue a three-site system, much of the savings CBO estimates for this option would not be realized.

Some defense experts believe that, without improvement of technology and absent testing of its components individually and as a whole, the GMD system is not ready to field. Fielding the IDC alone would allow testing and provide limited tracking and engagement capacity for ballistic missiles launched from North Korea toward Alaska or the West Coast of the continental United States. Moreover, the delay in additional deployments would allow time to improve missile defense technologies for incorporation into a more capable operational system, should the United States decide to deploy one.

Opponents of this option argue that ballistic missile launches from enemy nations pose a current threat to the United States. Thus, developing and deploying all currently planned GMD segments would provide urgently needed protection for the nation and its allies. In particular, only by fielding all GMD segments will the United States be able to defend all of its territory and extend its missile defenses to its allies and deployed forces.

The Space Radar (SR) program is intended to provide around-the-clock, all-weather, global surveillance for the U.S. military and intelligence community. SR would complement airborne radar (or sensor) systems, such as the Joint Surveillance and Target Attack Radar System (JSTARS), which provides surveillance and tracking of enemy forces over areas that are inherently more limited than those that space-based systems could cover. The proposed SR would provide periodic high-resolution imaging of large areas. Potentially, it also could detect moving targets, including enemy convoys and troops, to provide information about activities deep inside enemy territory.

This option would cancel SR and retain current surveillance systems, including JSTARS and Global Hawk, to provide battle-planning information. The Congressional Budget Office (CBO) estimates this option would save $340 million in outlays in 2008 and nearly $14 billion between 2008 and 2017.

According to House Report 108-553, the Department of Defense (DoD) is considering a system of nine low-Earth-orbit radar satellites, and current DoD plans call for the first SR satellite launch in 2015. CBO's estimate of 10-year savings for this option is based on the assumption that DoD will develop, procure, and operate the SR system according to those plans. A more detailed description of the information CBO used in constructing this option is presented in a recent CBO study, Alternatives for Military Space Radar (January 2007). If DoD subsequently decided not to deploy the constellation, much of the savings CBO estimates for this option would not be realized.

One justification for this option stems from the significant technical challenges and costs associated with collecting radar data over distances of thousands of kilometers compared with collecting data by aircraft over hundreds of kilometers. Other technical challenges are found in down-linking, processing, and analyzing large amounts of data quickly enough to support battle planning. The radar system also would require efficient, lightweight solar cell and battery technology, powerful on-board signal processing, high-bandwidth satellite-to-ground communications, and complex signal-processing algorithms for identifying moving targets.

Another argument for this option concerns the space radar's military value. House Report 108-553 states that the nine-satellite constellation proposed by DoD "would be unable to track vehicles effectively because of significant coverage gaps." This conclusion is supported by the CBO study cited above, which reports that a nine-satellite system, similar to that proposed by DoD, would be impractical for tracking individual ground targets, although the movement of large military units probably could be detected. Some would argue that those limitations reduce SR's tactical value to the military.

An argument against terminating the SR program is that the radar could be seen as the next logical and necessary step in military transformation, which emphasizes the use of superior intelligence to prevail in conflicts. The SR constellation would not require access to bases in the region of a conflict, nor would it be affected by operations delays during transportation of airborne sensors to an area of interest. SR also would be much less vulnerable to attack than airborne sensors operating close to areas of combat would be. Some proponents of SR also argue that the technology needed for power generation and signal processing is already mature and ready for operational use.

More than half of the federal government's cost of compensating military personnel falls outside military personnel appropriations for the Department of Defense (DoD). Other DoD appropriations pay for many noncash benefits, such as use of commissaries, DoD schools, base housing for military families, and some medical care. The Department of Veterans Affairs (VA) funds additional benefits, including veterans' health care and disability payments and benefits provided under the Montgomery GI bill.

Under this option, the DoD-funded costs mentioned above would become part of military personnel appropriations. Some VA programs also might be funded in the defense budget. That realignment would have two related goals: It would provide more complete information about how much money is being allocated to support military personnel, and it would give DoD managers a greater incentive to use resources wisely. The amount this option might save is unknown (so no table of year-by-year savings is shown). But with DoD-funded support of military personnel totaling about $140 billion in 2007, the potential savings from better management are substantial. For example, a savings of just 1 percent would equal more than $1 billion annually.

The current distribution of personnel costs among different appropriations makes it difficult for DoD, the Congress, and taxpayers to track the total cost of supporting military personnel. In the absence of a total picture, it is difficult to assess the resources devoted to health care, housing, and education benefits or to compare military with civilian compensation.

DoD has some recent experience in consolidating costs into military personnel appropriations. In 2003, it adopted accrual funding for the cost of health care for Medicare-eligible retirees. Those payments, which represent the future cost of providing health care benefits to future retirees, were added into the military personnel accounts of each service. (The current costs of providing health care benefits to Medicare-eligible retirees were removed from DoD's operations and maintenance budget and paid out of a new fund.) This option would expand that approach by incorporating additional personnel support costs into military personnel appropriations.

Advocates of this option argue that further consolidation would encourage DoD managers to use military personnel effectively and to substitute less costly federal civilian employees, contractors, or labor-saving technology for military personnel where possible. This option also would help DoD and the Congress by highlighting the extensive array of noncash benefits in the military compensation package.

Critics of this option argue that implementation could be difficult. For example, new financial management systems and a new appropriations structure would be required.

The cash pay that military personnel receive includes basic pay, which depends on rank and time in service, as well as bonuses, allowances, and the tax advantage that arises because some allowances are not subject to federal income tax. Basic pay is the most important element of cash pay, averaging about 60 percent of total cash compensation. Lawmakers typically use the employment cost index (ECI) for wages and salaries of private-sector workers in setting the annual military pay raise. In the 1990s, the raise generally was set either at the annual rate of increase of the ECI or 0.5 percentage points below it. However, the Fiscal Year 2001 National Defense Authorization Act set the annual raise for 2001-2006 at 0.5 percentage points above the ECI. To improve retention, several increases in the pay table for officers and enlisted personnel in some pay grades also were authorized. Those legislated changes raised the average basic pay for all enlisted personnel 13 percent between 2000 and 2006 and raised the basic pay for senior enlisted personnel 15 percent in real (inflation-adjusted) terms. Real basic pay for officers has risen 10 percent over the same period.

Another tool the services have used to increase retention is the selective reenlistment bonus (SRB), a cash incentive typically offered to qualified enlisted personnel in occupational specialties with high training costs or with demonstrated shortfalls in retention. Each service branch regularly adjusts its SRBs to address current retention problems, adding or dropping eligible specialties and raising or lowering bonuses. In addition, the Army pays a deployed SRB to all eligible soldiers who reenlist while deployed in support of current operations. Depending on the service, eligible personnel receive the bonuses in a lump sum at reenlistment, or they receive half at reenlistment and the remainder in annual installments over the course of the additional obligation.

This option would substitute reenlistment bonuses for part of the basic pay increase. From 2008 to 2011, it would limit annual basic pay raises to 0.5 percentage points below the increase in the ECI and offer SRBs to service members in occupations where shortages exist. It would increase the services' spending on bonus payments by about $375 million annually from 2008 through 2011 and remove current restrictions on the maximum bonus. Between 2008 and 2011, service members receiving the additional bonuses would receive higher overall pay than would be the case under the current plan. This option would cost $10 million in 2008 and save more than $4 billion between 2009 and 2012. Because bonuses do not compound the same way general pay raises do, however, all service members would have lower overall compensation in 2012 and beyond, unless the bonus program was extended.

The rationale for this option is that increasing selected reenlistment bonuses is a more efficient way to address occupational mismatches than is giving general pay increases, because bonuses allow DoD to target compensation to specific occupational categories. On average, from 2000 to 2005, about 30 percent of enlisted occupations regularly had shortages, while about 40 percent usually were overstaffed. General pay increases would alleviate shortages in some occupations but would worsen surpluses in others. Unlike pay increases, bonuses would be more easily adjusted from year to year to match recruiting and retention goals. Bonuses also would not incur the heavy cost of "tag-alongs," the elements of compensation, such as retirement benefits, that are tied to basic pay.

Another advantage of this option stems from the flexibility of bonuses, which could be focused on the years of service in which personnel make career decisions. And larger bonuses could provide more meaningful differences in pay among occupations, which could be a cost-effective tool for improving military readiness.

An argument against this option is that expansion of reenlistment bonuses would amplify pay differences among occupations and thus counter the long-standing principle of military compensation that personnel with similar amounts of responsibility should receive similar pay. Moreover, the practice of increasing bonuses deprives service members of the retirement and other benefits that they would receive if the money were part of basic pay throughout a career.

Warrant officers account for about 1 percent of active-duty military personnel, serving as senior technical experts and managers in a variety of occupations and, in the Army, as pilots of helicopters and fixed-wing aircraft. In rank, they fall between enlisted personnel and other commissioned officers. They (like the Navy's limited-duty officers) tend to have long careers, during which they gain considerable expertise.

This option would slowly expand the number of warrant officers to help attract and retain highly qualified, skilled personnel, particularly in occupations with attractive civilian alternatives. To achieve savings, it would offer smaller pay raises to senior enlisted personnel than those prescribed by current law.

Programs to help the military meet its labor force needs tend to be more cost-effective when they are focused on particular occupations and skills. Some analysts point out that growing numbers of midcareer and senior enlisted personnel have substantial college training that current military pay scales may not adequately recognize. Recent defense appropriation acts have increased pay for senior enlisted personnel more rapidly than for other groups. Between 2000 and 2006, real (inflation-adjusted) basic pay for senior enlisted personnel rose by about 15 percent; real basic pay for enlisted personnel generally increased by 13 percent.

Instead of paying all midcareer and senior enlisted personnel more, however, the Department of Defense (DoD) could offer warrant officer positions to those people it most wanted to retain or to those in military occupations with the best-paying civilian alternatives. For five years, this option would limit annual pay increases for personnel in grades E-6 and above to 0.5 percentage points below the increase in the employment cost index for private-sector workers. It would convert 10,000 enlisted positions in the top four grades to warrant officer positions. For 2008-2012, the net outlay savings would total $1 billion. A program that expanded opportunities for warrant officers could focus on specific areas, such as information technology, in which a robust civilian sector can make military compensation noncompetitive. DoD has used enlistment and reenlistment bonuses to fill such positions, although it might be argued that current bonuses are too small to provide meaningful differences among occupations.

This option might offer advantages in efficiency that do not yield near-term budget savings. Expanded opportunities for warrant officers might attract two-year-college graduates who could enter as professionals rather than serving long apprenticeships in the enlisted ranks. Service as a warrant officer also might appeal to those who prefer technical specialties over leadership jobs. The resulting more-experienced workforce could reduce the size of the force that DoD needs.

Converting senior enlisted positions to warrant officer positions might create new problems, however. About 16,000 warrant officers were on active duty in June 2005. Adding another 10,000 could make the force top-heavy without providing a commensurate increase in leadership. Some within the military might object to having a larger group of senior technicians who do not have leadership responsibilities. Also, reducing pay raises overall could hamper military recruitment and retention generally.

The last fundamental reorganization of military headquarters occurred under the Goldwater-Nichols Act of 1986. That law gave the unified theater commands—such as the European and Pacific Commands—the lead in planning operations and executing policy and had them report directly to the President. When a crisis develops that requires additional military forces and support, a unified theater commander calls on the four military services, which recruit, train, equip, and support the forces; the commanders then employ the forces in their geographic areas of responsibility. The Department of Defense (DoD) is changing the locations of some of its combat forces overseas, moving some units from one base to another and returning some units to the United States. That effort does not affect the services' overseas component commands.

In practice, unified commanders constitute another management layer over existing overseas service component commands, such as the U.S. Army Europe and the Pacific Fleet. The commanders' requests are relayed through component commands to the services' U.S. headquarters. Because each service maintains its own headquarters in a given region, there are redundancies in many management functions. In some regions, the only personnel in a particular service branch are those at the component command headquarters. Those various overseas headquarters now are staffed by 6,000 personnel, or 10 percent of all headquarters staff.

This option would reorganize the military's command structure by eliminating the overseas component headquarters, a change that could release 4,000 troops for critical missions. This options would not cut end strength. Instead it would free those military personnel for assignment to different duties. Some operating costs might be saved, but because estimating those savings is not straightforward, no year-by-year table is shown.

An advantage of this option is that eliminating overseas component commands would tighten command and control and free troops for other duties. It would streamline communications by eliminating a management layer between the services and the unified commanders. This option would retain some personnel, however, given the assumption that some command responsibilities could not be eliminated.

An argument against this option is that the overseas component commands provide essential support, including dedicated and responsive support for staging operations and integrating personnel and equipment deployed to a region. The unified commanders are thus freed to concentrate on their combat responsibilities. Overseas component commands also bolster theater support services (medical support, engineering, intelligence, fuel handling, and supply transport, for example), and they plan and execute joint and coalition military exercises and treaty obligations as directed by the North Atlantic Treaty Organization and under bilateral agreements.

Another argument against this option is that the envisioned restructuring would be the largest since the Goldwater-Nichols Act, and it could eliminate as many as 45 general-officer positions overseas. Some observers, however, including some senior staff members in the Office of the Secretary of Defense, argue that despite the difficulty, the new threat environment and the need for additional combat troops demand consideration of just such a widespread reorganization.

Over four years, this option would replace 20,000 of the more than 500,000 uniformed military personnel in support jobs with Department of Defense (DoD) civilian employees and make those military positions available for combat functions. The Congressional Budget Office has identified some jobs that one service branch considers "military essential" that the others do not and some in all branches that could be filled by civilians. Those jobs are in military units that do not deploy overseas for combat, and they do not involve sensitive functions that might raise security concerns.

Some analysts say as many as 90,000 positions could be converted. This option would convert 20,000 of those jobs, making that many military personnel available to satisfy demands for combat units. Although costs would increase overall, some savings would occur as fewer civilians were substituted for a given number of military personnel. Because the civilians would not be encumbered with military-specific duties, they would have more time to perform their jobs.

Nevertheless, the addition of civilians could decrease outlays by $3.3 billion between 2008 and 2012 and by $8.9 billion between 2008 and 2017, as indicated by DoD's experience with similar conversions. That cost could be lower if some converted positions were opened to contractors. In 2004, DoD approved a plan to convert 10,000 Army military to civilian positions between 2006 and 2011, replacing military personnel with fewer civilians than assumed in this option. Depending on the extent to which the conversions are implemented as planned, the cost of implementation would be lower than shown here.

Although proposals to convert military to civilian positions have been made in the past, only a small percentage of DoD's total personnel have been subject to review. In 2006, DoD made an inventory of civilian and military positions, categorizing them by function; determining whether they were inherently governmental; and, if so, deciding whether each had to be filled with a military service member. That inventory could be used to identify new positions for civilian employees of DoD.

The Air Force categorizes as military 54 percent of its positions in the functional category of morale, welfare, and recreation services. Removing that designation could open about 2,000 jobs to civilians. The Army fills 32 percent of its positions in legal services and support with military staff. In contrast, the Navy has 61 percent and the Air Force has 79 percent of those positions staffed with military personnel. Converting the Air Force and Navy jobs in that category could open more than 4,000 jobs to civilians.

Opponents of this option argue that defining, evaluating, and then redesignating positions would be a cumbersome process with hard-to-define savings. They point out that comparisons among the services can be misleading because some functional areas are service specific. The Navy, for example, must rely on military personnel to fill shipboard support positions. Finally, substituting DoD civilian employees for military personnel without reducing end strength would increase DoD's total costs. Proponents of transferring military personnel out of nonmilitary tasks argue that even if military end strength were not reduced, personnel would still be freed to fulfill their primary mission of military combat.

In 1996, the British Parliament authorized the United Kingdom's Ministry of Defence to establish "sponsored reserves" to permit peacetime military contractors to become activated reservists when they are deployed overseas. The United States has a similar system for dual-status federal employees who serve with Reserve and National Guard units. While a unit is at home, those employees work as federal civilians; when their units are deployed overseas, they are mobilized to active duty.

A new sponsored-reserve program would require Department of Defense (DoD) contractors that supply services or equipment to have on their payrolls a specific portion of employees who also are members of the inactive reserve component of the military. Sponsored reservists would be contract employees while performing routine tasks at home but would agree to be activated to military status and to perform the same jobs during deployment overseas. Currently, many contractors' employees also are reservists, but when they are deployed overseas, they do other jobs or they work with units that are different from those of their peacetime employment.

This option would gradually institute the program to attract and retain highly qualified, skilled personnel in functions that already rely extensively on contractors. It would reduce by 20 percent the number of active-duty personnel in logistics, real property maintenance, and installation and facilities management. Over four years, about 20,000 active-duty personnel would be replaced with sponsored reservists. Converting those positions and reducing active-duty end strength by that amount could save about $1.6 billion in outlays from 2008 through 2012. In addition to their peacetime responsibilities as contractor employees, sponsored reservists would have military responsibilities but only when they were called to active duty. Some of the savings would accrue because, absent those military responsibilities during peacetime, a smaller number of sponsored reservists could replace a larger number of full-time military personnel.

This option would bridge the gap between wholly privatized functions performed by contractors and functions performed by the military. It would place deployed contract workers within the military chain of command (better ensuring military command and control) and afford them the protections of military status. In particular, sponsored reservists would be protected by the Geneva Conventions. Sponsored reservists also could provide military capability in hard-to-fill occupations or in jobs that require exceptional technical expertise. As members of the inactive ready reserve, those personnel would not count against legislated caps on end strength.

Converting active-duty to sponsored-reserve positions could create some difficulties, however. Although DoD has considered creating such a program, there might be concern that its ramifications have not been explored fully. As a first step, smaller demonstration projects might be preferable to the creation of a new personnel category. There also could be concern about creating a class of uniformed personnel that had not had the same training or leadership development afforded the regular military.

If DoD implemented a sponsored-reserve program without reducing active-duty end strength, active-duty personnel would be freed to perform other functions, but the savings shown in the table would not be achieved.

Many families may be overinsured under the current Department of Defense (DoD) military health care system. Given a choice, some might trade generous health benefits for cash compensation. This option would have DoD provide family members of active-duty personnel with a special cash allowance for health coverage. The allowance would be nontaxable (like the current housing allowance), and it could be used in one of three ways. Under the first option, family members could purchase one of the current TRICARE plans (Standard, Extra, or Prime). The second alternative would allow families to use some of the allowance to purchase a new lower cost, low-option TRICARE plan and keep the remaining funds. Like TRICARE Prime, the low-option plan would have managed care features, but it would incorporate substantial deductibles and copayments for health care services obtained either at military facilities or from civilian providers. (Low-option TRICARE would include a "stop loss" component to limit annual out-of-pocket expenditures and thus control the financial consequences of catastrophic illness.) Under the third alternative, military family members could show proof of insurance and apply the allowance to their share of the premiums, copayments, and deductibles of another health insurance plan.

This budget option would save nearly $1 billion in outlays over the next five years. That estimate incorporates the cost of the cash allowances and accounts for the decreased demand for health care by enrollees in the new plan. The low-option plan's higher out-of-pocket expenses would be expected to encourage restraint in health care purchases. The estimate also accounts for the increased cost of the benefit for eligible family members of active-duty personnel who, because they are not using TRICARE, currently cost the system nothing but who would be likely to apply for the cash allowance.

This option would offer several advantages. First, families of active-duty personnel would have more flexibility in choosing the mix of benefits and cash they receive. Second, enrollees in the low-option plan would have an incentive to use medical services prudently because they would be responsible for a significant share of the cost. Third, some of the cost would be shifted to the civilian employers of military spouses, thus reducing DoD spending. Finally, because family members would commit annually to a health insurance plan, total utilization would be easier to predict than it is under the current system, which allows users to join or leave at any time. Thus, this option would improve resource planning within the military health care system and allow DoD to negotiate firmer contracts for pharmaceuticals and civilian medical services. That advantage would exist even if most beneficiaries chose to remain in one of the three traditional TRICARE plans.

This option also would entail potential disadvantages. Enrollees who chose low-option TRICARE coverage would assume additional risks and might face financial difficulties, despite the low-option's stop-loss limit. Families who obtain health insurance through a spouse's employer might have their coverage disrupted in the event of the relocation of the active-duty member to a new post. DoD would have to develop methods to prorate cash allowances and deductibles for beneficiaries forced to change health plans midyear.

TRICARE For Life (TFL) was introduced at the beginning of fiscal year 2002 as a supplement to Medicare for military retirees and their family members over age 65. The wraparound program pays nearly all of its users' remaining medical costs and carries few out-of-pocket fees. Because the Department of Defense (DoD) is a passive payer in the program—it neither manages care nor provides incentives for cost-conscious use of services—it has virtually no means to control the program's costs.

This option would help reduce the costs of TFL as well as for Medicare by introducing small copayments for services and by increasing copayments for prescription drugs to match those commonly charged by civilian plans. Because the program is a wraparound benefit, lawmakers or DoD would need to establish new rules to ensure that users paid minimum out-of-pocket charges—for example, $20 for an office visit or $100 for the first day of a hospital stay—before coverage would begin.

Introducing such charges would reduce the federal spending devoted to TFL (including Medicare savings) by about $1 billion in 2008, by $6.3 billion over the next five years, and by $16 billion between 2008 and 2017. Much of those savings would come from reduced demand for medical services rather than from a transfer of spending from the government to military retirees and their families.

Introducing copayments into TFL would increase beneficiaries' awareness of the cost of health care and promote a concomitant restraint in the use of medical services. Research has generally shown that introducing modest cost sharing can substantially reduce medical expenditures without causing measurable increases in adverse health outcomes.

Among its disadvantages, this option could discourage some patients (particularly low-income patients) from seeking medical care and thus negatively affect their health. Beneficiaries who require treatment for chronic conditions, such as hypertension, might forgo purchasing necessary drugs. Some recent research indicates that rapid increases in copayments can lead to significant reductions in beneficiaries' use of prescription medicines.

The Department of Defense (DoD) operates three chains of military exchanges—the Army and Air Force Exchange Service, the Navy Exchange Command, and the Marine Corps exchange system. The chains provide an array of retail goods and consumer services at military bases for combined annual sales of about $12 billion, the Congressional Budget Office estimates.

This option would consolidate the three systems into a single organization. In addition, it would encourage more efficient operation by requiring the combined system to pay all of its operating costs from sales revenues, rather than relying on DoD to provide some services free of charge. After a three-year phase-in period, those changes would save about $180 million annually.

Studies sponsored by the Office of the Secretary of Defense show that consolidation could lead to significant efficiencies by eliminating the costs of maintaining several purchasing and personnel departments, warehouse and distribution systems, and management headquarters. Although consolidation would entail some one-time costs, CBO estimates that the required spending would be offset by inventory reductions.

CBO estimates that DoD provides the exchanges with about $400 million in free services each year. DoD maintains some parts of buildings, transports goods overseas, and provides utilities at overseas stores. DoD also provides indirect types of base support, such as police and fire protection. Under this option, the combined system would reimburse DoD for the costs of direct support and would thus have an incentive to economize on its use. Furthermore, the requirement for the system to pay all of its own operating costs would improve the exchanges' visibility in the defense budget.

When the exchanges' revenues exceed full operating costs, a portion of the surplus goes to fund military morale, welfare, and recreation programs. The surpluses would likely be smaller under this option, so it is assumed that lawmakers would appropriate about $80 million per year in additional funds for those programs.

One obstacle to implementing this option would be the need to find an acceptable formula for allocating among the individual services the funds for morale, welfare, and recreation activities. There could be concern about fair distribution—either of the earnings or of any additional appropriations. There also could be fear that lawmakers would gradually reduce the amount of additional funding for those activities.

Some critics of consolidation argue that the Navy Exchange Command and the Marine Corps system, with their unique service identities, meet the needs of their patrons better than a larger, DoD-wide system could. But consolidation proponents point to the Army and Air Force Exchange Service, which has served both branches for many years. People who shop in exchanges say their main concern is the availability of low prices and a wide selection of goods—a concern that a consolidated system might be able to satisfy more effectively.

The Department of Defense (DoD) operates four retail systems on military bases: a network of grocery stores (commissaries) for all of the service branches and three chains of general retail stores (exchanges) for the Army and Air Force, the Navy, and the Marine Corps. This option would consolidate those systems into a single retail chain that would operate more efficiently, without any appropriated subsidy. Like the current separate systems, the consolidated system would give military personnel access to low-cost groceries and other goods at all DoD installations, including those in isolated or overseas locations.

The current commissary and exchange systems operate under very different funding mechanisms. The commissary system, which is run by the Defense Commissary Agency (DeCA), has annual sales above $5 billion, but it also receives an appropriation of about $1.2 billion a year. The three exchange systems have annual sales totaling about $12 billion. They do not receive direct appropriations; instead, they rely on sales revenue to cover their costs.

The exchanges can operate without an appropriated subsidy because they charge customers a higher markup over wholesale prices than commissaries do. The exchange systems also are nonappropriated-fund (NAF) entities rather than federal agencies, so they have more flexibility in business practices for personnel and procurement. Because DeCA is a federal agency, its employees are civil service personnel and it follows standard federal procurement practices. This option assumes that consolidation would eliminate duplicative overhead headquarters functions and that DeCA's civil service employees would be converted to the NAF workforce.

Under this option, the commissary and exchange systems would be consolidated over a five-year period. At the end of that process, the budget authority required to operate the combined commissary and exchange system would be lower by about $1.4 billion per year. Of that amount, about $1.2 billion would come from eliminating the subsidy for commissaries and $200 million would come from eliminating duplicate functions. This option would return half of the $1.4 billion to active-duty service members through a tax-free grocery allowance of about $500 per year, payable to service members who are eligible to receive current cash allowances for food. The grocery allowance would be phased in to coincide with the consolidation of commissary and exchange stores at each base. The remaining $700 million would represent savings for DoD.

To break even without appropriated funds, the consolidated system would have to charge about 14 percent more for groceries than commissaries do now. At the current level of commissary sales, a 14 percent price increase would cost customers an extra $720 million annually.

Active-duty members and their families would benefit from consolidation. Those families would pay about $200 more per year for groceries—but that amount would be more than offset by the new grocery allowance. (A military family would have to spend at least $3,500 per year on groceries in commissaries before a 14 percent price increase outweighed the benefits of a $500 allowance.) Cash allowances would be particularly attractive to personnel who live off base and could shop more conveniently near home or online. All military families—active-duty, reserve, and retired—would benefit from longer store hours, one-stop shopping, access to private-label groceries (which are not currently sold in commissaries), and the greater certainty of a military shopping benefit that did not depend on the annual appropriation process. Another advantage is that the $500 average grocery allowance could be targeted to specific pay grades or groups, with larger allowances given to enhance retention or to benefit junior enlisted members with large families.

The retail system would benefit as well. Commissaries and exchanges must now compete with online retailers and the large discount chains that have opened discount grocery and general merchandise stores just outside the gates of many military installations. Recent increases in base security procedures and changes in the civilian retail industry have made it more difficult and costly for DoD's fragmented retail systems to provide those services. This option would allow a consolidated system staffed by NAF employees to better compete with civilian alternatives.

Nonetheless, some people might oppose the change, arguing that low-cost shopping on bases has long been a benefit of military service. Under this option, about $425 million of the price increase would be borne by the military retirees who now shop in commissaries but who would not receive grocery allowances. As a result, this option could face strong opposition from associations of retirees. The average family of a retired service member would pay an additional $200 per year for groceries.

When vehicle transmissions, radar equipment, and other weapon system components need repairs, unit commanders can have the work done at their own facilities or send equipment to central maintenance depots. Under current policies, the depots' repair charges exceed actual repair costs, and that can raise total costs to the Department of Defense (DoD) because there is less incentive to use the depots, even when doing so would save money overall.

This option would allow depots to charge only for the incremental cost of repairs (that is, the costs attributable to the specific maintenance action). Currently, repair charges for components (called depot-level repairables, or DLRs) include incremental costs for labor, materials, and transportation and a share of the fixed costs of overhead. Under this option, the DLR charges would include only those costs that change with the number of DLRs in the depot—for instance, materials, transportation, and direct labor costs. Fixed costs, including overhead, would be covered by an annual flat fee to customers. The new pricing policy could save about $1 billion in outlays over five years because commanders would have stronger incentives to send the work out.

A two-part pricing structure, similar to that used by some utility companies, has been proposed by the RAND Corporation, the Center for Naval Analyses, and others. One RAND study concluded that two-part pricing can reduce depot charges by more than a third. The reduction could shift the workload to depots, and that in turn could reduce DoD's total repair expense. According to RAND, the Navy, and the Office of the Secretary of Defense, local maintenance can cost from 25 percent more to twice as much as repairs done at the depots.

DoD estimates local-facility repair costs at $54 billion. If two-part pricing shifted just 2 percent of the workload to centralized depots, about $1 billion in repair costs also would shift each year. DoD could save $240 million in annual outlays, on average, between 2008 and 2017.

Shifting repair work also could improve quality because local facilities often are not as well equipped for some tasks as depots are. The depots' higher prices can give local facilities an incentive to scavenge parts and, eventually, scavenged DLRs could be sent out for repairs, resulting in labor charges from two facilities for one unit.

A disadvantage of this option is that it could be difficult to develop accurate two-part prices. Depot managers, eager to attract work by keeping prices as low as possible, might try to move variable costs into the flat fee or use direct appropriations to pay for variable expenses. They might be reluctant to separate variable repair costs from fixed costs if doing so could highlight excess capacity. Such influences on prices would cloud cost comparisons between depots and local repair facilities. Two-part pricing also would eliminate a primary benefit of current DLR pricing: total cost visibility. By including fixed and workload-dependent costs in charges, the current system is intended to boost cost-consciousness and encourage commanders to be prudent in their use of DLRs. The system has worked, but it also creates an unintended incentive for unit commanders to use local facilities.

Currently, the Department of Defense (DoD) spends about $27 billion annually for equipment maintenance and repairs provided at its central maintenance depots or at facilities operated by private-sector contractors. The "50/50 rule" specified in 10 U.S.C. section 2466 allows DoD to award contracts for up to half of its depot maintenance appropriations to private-sector bidders, although some public-private partnerships are excluded from the calculation. Generally, work that is assigned directly to government depots without competitive bidding from the private sector costs more. Historically, opening depot work to private-sector bidders has been estimated to save at least 20 percent of costs, including cases in which the government depot wins the work. Studies that have tracked post-competition costs have shown that the savings from competition persist beyond the initial contract award.

DoD currently uses the private sector to perform as much depot work as the 50/50 rule allows. If lawmakers were to relax the rule to a 60/40 split, DoD could open more depot work to competitive bidding and stay within the new rule as long as the private sector did not take more than about $2.7 billion worth of work per year. With the new rule, an additional $3.6 billion in repair work could be opened to competitive bidding each year, assuming the private sector wins three-quarters of the contracts. The estimate of future savings is inexact, yet a conservative assumption that competition saves about 20 percent of costs predicts average annual savings through 2017 of about $370 million. Savings would not occur immediately and would be less in the near term because it would take the depots time to prepare for additional competition and to adjust to changes in workload. Alternatively, the 50/50 rule could be eliminated or redefined so the calculation applied to all maintenance (that would include organic and intermediate maintenance now performed mostly by DoD personnel). Savings would be larger under those changes because the depots could subject even more work to bidding.

Proponents of this option argue that the current limits are arbitrary and reduce DoD's flexibility in determining which source is best to provide maintenance. Easing the restrictions would allow DoD to seek the most efficient and most cost-effective source of support.

Opponents are concerned that DoD should maintain an organic skill base within its operational units to perform depot maintenance. They also consider it important that DoD retain the capacity to sharply increase depot maintenance when required, although private contractors often can meet sudden increases in demand. Some opponents also question the comparability of government and private accounting methods (mainly because of the government depots' limited capability for cost accounting) and so question the fairness of the competition. Finally, opponents of this option express concerns that it might lead to the loss of federal civilian jobs at the depots.

The Defense Base Act (DBA) requires that Department of Defense (DoD) contractors purchase workers' compensation insurance for employees who work overseas. Firms traditionally have purchased coverage on the competitive market for each DoD contract separately. But there is evidence that insurance premiums, commonly listed as a rate per $100 in direct labor cost, currently are higher than historical trends would predict. The higher cost of the premiums, which is passed on to DoD as overhead charges, is likely attributable to the increase in the number of contractor operations in the Middle East and to the heightened risk associated with working in dangerous locations.

This option would permit DoD to negotiate with a single broker to provide a large-scale DBA insurance pool for all contractors. The blanket coverage would provide a worldwide DBA rate for a specific period. Creating a larger DBA insurance pool would lower risk premiums and strengthen the buyer's negotiating position. The Department of State and the U.S. Agency for International Development (USAID) secure blanket coverage now, and their contractors pay lower DBA insurance premiums than DoD contractors do. A similar program is in development for Army Corps of Engineers contractors.

The savings generated by this option would depend on the cost advantages of an insurance pool as well as on the number of contractors deployed and the dangers associated with their locations. Under the assumptions that contractors would pass savings along to DoD through reduced overhead charges and that the pace of military activities in support of the global war on terrorism eventually will slow, the Congressional Budget Office estimates that this option would save an average of $36 million in annual outlays between 2008 and 2017.

The major rationale for this option is that pooling risk is an effective way to lower insurance costs. Firms with small numbers of deployed contractors would especially benefit from being included in a pool; when DBA insurance rates are negotiated independently, small firms tend to pay more for premiums than do larger companies.

An argument against this option is that a DBA insurance pool essentially would provide a subsidy to contractors in more-dangerous locales. Moreover, the creation of a DBA insurance pool would present several administrative challenges and would not guarantee savings for DoD. The State Department and USAID are much smaller agencies, and their use of blanket DBA insurance may not extrapolate to defense contracts. It is unclear whether a single insurance provider, or even several providers working together, would be willing to underwrite DBA insurance for all DoD contractors. Firms with large numbers of deployed employees, particularly those in relatively safe locations, might be reluctant to participate in an insurance pool if doing so would limit their negotiating leverage and flexibility. Finally, the costs of initiating and administering a large-scale DBA insurance program (which are not reflected in the estimates shown here) could greatly diminish the savings.