Knowing how the United States will need to use force in the future is difficult, so selecting the most appropriate amount and mix of strategic lift is even harder. Would a greater emphasis on pre-positioning allow the United States to fight major contingencies at lower cost? Would such an approach force the Department of Defense to give up other types of military capabilities, such as conducting long-range, brigade-size airdrops? Is the ability to perform special airlift missions just as important for national security as delivering cargo to major conflicts? In order to look explicitly at the trade-offs between different ways to move forces, this chapter reviews the Administration's strategy for investing in strategic mobility and examines the costs and capabilities of five possible alternatives.



The Administration's Plan

The Administration's blueprint for modernizing strategic mobility is an ambitious one with two main areas of investment. First, it calls on the Navy to purchase large, medium-speed roll-on/roll-off ships and use some of them to preposition equipment for a heavy brigade closer to regions of potential conflict. That afloat pre-positioning would allow DoD to deliver heavy forces much more quickly than it did during Operation Desert Shield. To that end, the Administration's plan would also preposition equipment for one heavy brigade on land in South Korea and for two heavy brigades plus support units on land in the Persian Gulf region.

Second, the Administration recommends that the Air Force buy a total of 120 C-17 aircraft. Those planes would give DoD the ability to perform a variety of special airlift missions, such as conducting a large-scale airdrop over intercontinental distances or devoting one or two squadrons of C-17s solely to moving cargo within a theater. Modernizing strategic lift forces with those missions in mind provides considerable military capability. It also costs more, however, because it requires a larger number of planes than DoD would need for more traditional airlift missions.

The Congressional Budget Office (CBO) estimates that the Administration's plan for purchasing and operating LMSRs and C-17s and expanding prepositioning would cost nearly $3.8 billion in 1998 and $21.5 billion over the next five years (see Table 6). Buying, operating, and supporting all of those strategic lift forces would cost just over $54.2 billion between 1998 and 2020. Note that although those figures include acquisition, operation, support, and construction costs associated with C-17s, LMSRs, and prepositioned equipment, they do not reflect the cost of all mobility forces (as discussed in the next section).

About 84 percent of the costs in Table 6 would fall within the Air Force's budget since they result from purchasing and operating C-17s. Indeed, the C-17 acquisition program is the largest in the Air Force's budget until the end of the decade, when the service will begin procuring F-22 fighters in large quantities. Most of the remaining mobility costs shown in Table 6 are associated with prepositioning U.S. equipment in the Persian Gulf, on the Korean Peninsula, and on ships. The costs of building warehouses and operating and supporting prepositioned sets of equipment will largely fall to the Army, and the costs of procuring LMSRs will fall to the National Defense Sealift Fund, which is administered by the Navy.

A caveat about the Administration's plan is in order. The Mobility Requirements Study Bottom-Up Review Update recommended devoting one LMSR or two smaller ROROs to additional prepositioning rather than to surge sealift, as DoD originally intended. Such a step would allow DoD to move an additional 280,000 square feet of equipment for combat-support and combat-service-support units to a major regional conflict during its halting phase--which the study's authors deemed necessary to hold the risk of failing to achieve military objectives to a moderate level.

Military officials have been debating whether to carry out that recommendation since 1995, but DoD has not yet made a decision. Although the Army has told the Joint Chiefs of Staff that it might be able to preposition more equipment, it has not included the additional funding that would be needed for such a measure in its budget proposals. For that reason, CBO did not consider the additional prepositioning to be part of the Administration's plan.



Five Alternatives to the Administration's Plan

To examine the costs and capabilities of other investment strategies, CBO developed five alternatives to the Administration's plan for strategic mobility. Each was designed to meet the needs associated with DoD's current planning scenario: fighting two major regional contingencies that occur at nearly the same time. Most of the alternatives would cost significantly less than the Administration's plan. Thus, they could free up resources for different types of defense spending, for deficit reduction, or for other federal priorities.

The five alternatives emphasize different modes of lift (see Table 7). Options I and II would expand pre-positioning (afloat and on land, respectively) beyond what the Administration envisions, at the cost of fewer C-17s for airlift. Options III and IV focus on airlift. Option III would use a different mix of planes than the Administration's plan: fewer C-17s and more commercial wide-body jets. Option IV would add 20 C-17s to the Administration's planned airlift fleet, while scaling

back the number of LMSRs on which the Army would preposition equipment. Option V would emphasize sealift by purchasing an additional LMSR for surge sealift and fewer C-17s for airlift.

In many ways, those options represent only marginal changes from the Administration's plan. All of the alternatives would have the vast majority of strategic mobility forces in common. In each case, DoD would have 102 C-5 and 37 KC-10 aircraft devoted to airlift operations, all C-141s would be retired from service by 2007, and the military would have access to planes from the Civil Reserve Air Fleet. Similarly, the Administration's plan and all of CBO's options assume that if conflict arose, DoD could use its eight fast sealift ships, its fleet of roll-on/roll-off ships and other vessels from the Ready Reserve Force, and commercial sealift to transport cargo.

Since many alternatives to the Administration's plan are possible, the options presented here are intended to be illustrative. Like the Administration's blueprint, each of CBO's alternatives includes sizable new investments in the military's own strategic lift forces. That is, none of the options would expand DoD's access to commercial airlift and sealift as an alternative to purchasing more military planes and ships. Nor would the alternatives invest in a program to install national defense features on commercial ships rather than buying LMSRs, although such a measure might be a reasonable approach to lower DoD's costs.

Rather than including costs for mobility forces that are common to all options, the tables that follow display costs for those elements of mobility forces that differ from the Administration's plan. For example, since DoD would incur the costs in any event, the tables do not reflect operating and supporting C-5 squadrons and SL-7 fast sealift ships or buying ROROs for the Ready Reserve Force. Thus, readers should bear in mind that although CBO's estimates allow for comparing costs among alternatives, they do not reflect DoD's total costs for procuring, operating, and supporting all mobility forces.

Option I: Buy Fewer C-17s and Preposition More Equipment Afloat

Under the first option, DoD would limit its purchases of C-17s to 72 aircraft, or 48 fewer than under the Administration's plan. In their place, DoD would buy one additional LMSR in 1999, which would carry preposi-tioned equipment with or near the Army's other prepo-sitioning ships in the Indian Ocean. Option I would raise the total number of LMSRs used for preposition-ing to nine, with a combined storage space of more than 2.25 million square feet.

CBO chose 72 as the size of the C-17 fleet in Option I (and several other options) because two 1995 analyses of how many C-17s DoD might require, conducted by the Air Force and the Office of the Secretary of Defense, included a total purchase of 72 as one alternative. By using the same number, CBO could apply the lessons learned in those analyses.

How does one ship substitute for 48 C-17s? Each newly constructed LMSR can preposition at least 250,000 square feet of cargo. That is equivalent to the floor space on 160 to 225 C-17s (depending on whether the space on the plane's ramp is included). If C-17s were flown at the maximum rate used by mobility planners, it would take a total inventory of 38 to 52 of them to deliver 250,000 square feet of cargo to the Persian Gulf over an 11- to 12-day period--the same amount of time it would take one LMSR to steam from Diego Garcia and unload its prepositioned equipment.⁽¹⁾

Airlift loads are constrained not only by the amount of available floor space on each plane but also by height and weight limitations. So the number of airlift missions required to move a comparable amount of heavy or bulky equipment on board one LMSR will almost always be larger than the square footage example described above. Based solely on the cumulative amount of cargo that each can move by the end of a two-week period, one additional LMSR devoted to prepositioning could arguably offset the deliveries of 48 C-17s.

What might DoD preposition on another LMSR? Given that the Army already plans to preposition a considerable amount of equipment, answering that question is not necessarily easy. To reduce airlift requirements, DoD would need to preposition equipment that it would otherwise airlift early in a contingency. If instead it prepositioned the equipment for a heavy brigade that was not needed until the end of the first month of a contingency (and thus would normally be transported by surge sealift), airlift requirements might actually increase rather than decrease. The reason is that having prepositioned equipment closer to hand could lead regional commanders to place those units higher on their list of priorities for deployment. And because most units include some equipment that is not suitable for prepositioning, airlift would be needed to transport that equipment.

The Army's current plans involve placing equipment for one heavy brigade, its support units, and theater- and corps-level equipment on board eight LMSRs. Those plans by no means exhaust the Army's possibilities for prepositioning. For instance, helicopters are not suitable for storing on ships because performing routine maintenance on them would be difficult. But at least half of the weight of a heavy division's aviation units comes from trucks, trailers, and other vehicles that could be placed on an LMSR. Likewise, a few critical units that would most likely deploy during the halting phase of a major conflict--such as air-defense and artillery units--could be prepositioned as well.

Other military services may be able to expand their prepositioning as well. The Air Force already prepositions a considerable amount of equipment on land so it can quickly establish air bases in remote or undeveloped regions. Yet other units might also be reasonable candidates. For example, Air Force engineering units perform jobs that aid the deployment of troops, such as creating and augmenting airfields or, as in Bosnia, building housing. Because of their numerous trucks, bulldozers, forklifts, cranes, and the like, those units weigh a considerable amount, but the Air Force plans to deploy them by air. By purchasing additional sets of equipment for engineering units and prepositioning it afloat or on land in the Persian Gulf and Korea, the Air Force might be able to conserve on demand for early airlift deliveries.

CBO identified nearly 260,000 square feet (over 11,400 tons) of equipment that could be prepositioned on board an additional LMSR--about 14 percent to 19 percent of the tonnage that DoD projects would otherwise be airlifted during the halting phase of a major conflict. CBO chose units that are likely to have high priority for early delivery during a major conflict to ensure that prepositioning the equipment would offset airlift requirements rather than add to them.

Roughly one-third of that weight is associated with Air Force units; the remaining two-thirds is composed of Army equipment.⁽²⁾ CBO assumed that DoD would purchase additional sets of equipment to preposition rather than use existing stocks, which results in a relatively conservative estimate of the option's cost.
 

CBO anticipates that the cost of Option I, including one additional LMSR plus new equipment for preposi-tioning, would total $14.4 billion over the 1998-2002 period or about $36.1 billion through 2020 (see Table 8 on page 61). Since Option I contains fewer airlift purchases than the Administration's plan, CBO estimates that it would cost $18.2 billion less over the 1998-2020 period.

CBO's estimate assumes that the Air Force will buy the 24 remaining C-17s it needs to reach a total of 72 more slowly than now planned. C-17 purchases would reach a maximum rate of only eight planes a year under Option I rather than 15, which would lower costs in the short run. But the average cost per plane would be somewhat higher since they would be built at a less efficient rate than anticipated and without the savings associated with a multiyear-procurement strategy.

Option II: Buy Fewer C-17s and Preposition More Equipment on Land

Under Option II, DoD would also purchase 24 additional C-17s over the next three years for a total of 72 planes. Instead of buying another LMSR for preposi-tioning, however, DoD would build climate-controlled warehouses in both the Korean and Persian Gulf regions and preposition nearly 240,000 square feet (or about 11,400 tons) of equipment at each site. As with Option I, CBO selected equipment for prepositioning that is likely to be of high priority in DoD's deployment schedule. That measure helps to ensure that the additional prepositioning offsets rather than adds to airlift requirements.

Although Option II would not buy any additional ships, its costs are not necessarily lower than those for Option I because DoD would need to purchase two sets of equipment--one for each site. Under Option I, DoD could send equipment prepositioned on ships to either major conflict. But in order to deliver roughly the same amount of cargo during the halting phase of each conflict as the Administration's plan, Option II includes additional warehouses and equipment in both theaters.

CBO estimates that the cost of building warehouses and buying additional equipment would push Option II's price tag about $1.4 billion above Option I's through 2020. However, Option II would still cost $16.8 billion less than the Administration's plan during that time (see Table 9). In all, CBO estimates that Option II would cost $15.1 billion over the 1998-2002 period or about $37.5 billion between 1998 and 2020.

In order not to underestimate the costs of buying additional equipment, CBO included purchases of some major weapon systems, including a Patriot missile battalion for each region. (Maintaining the Patriot's sensitive electronics on board ships is difficult, so that system was not included among the equipment purchased for Option I.) If DoD substituted other units for the more expensive Patriot battalion, Option II might cost $460 million less over the 1998-2020 period. Given uncertainties about the type of equipment that DoD would need to purchase for those two alternatives, readers could consider their costs to be roughly comparable.

As with Option I, CBO assumed that the Air Force would procure the remaining 24 C-17s more slowly than current plans allow. That assumption leads to lower total acquisition costs in the near term (since DoD would buy fewer aircraft), but a higher average cost for each one.

Option III: Buy a Mixture of Airlift Planes

This alternative would allow DoD to maintain the same level of theoretical airlift capacity as the Administra-tion's plan. But instead of procuring a total of 120 C-17s, Option III would purchase 24 more C-17s (for a total of 72) plus 30 commercial wide-body jets, which are larger and less expensive. (CBO used estimates of the cost and capabilities of the C-33, a military version of the Boeing 747-400 freighter, for this analysis.)

CBO estimates that total costs for Option III, including purchases of C-17s and commercial planes, would be about $4.1 billion in 1998, or $0.3 billion more than the Administration's plan in that year (see Table 10). Over the 1998-2020 period, however, Option III would cost $8.4 billion less. Nearly half of those savings would take place over the next five years.

As with Options I and II, DoD would purchase its C-17s at a slower pace than under the Administration's budget proposal--a maximum of eight per year rather than 15. Thus, the average cost of each C-17 would be higher under these alternatives.

Option IV: Buy More C-17s and Preposition Less

Under this option, DoD would rely more heavily on airlift to deliver cargo during the halting phase of two major regional contingencies. Specifically, Option IV would buy a total of 140 C-17s. That size fleet would provide the Air Force with the upper end of its desired range for theoretical airlift capacity--nearly 52 MTM/D by 2006. But the alternative would also scale back purchases of LMSRs by one, leaving seven rather than eight ships to house prepositioned Army stocks in the Indian Ocean. For that reason, Option IV can be thought of as the reverse of Option I, which would add an LMSR for afloat prepositioning and subtract C-17s.

Purchasing 20 more C-17s than the Administration's plan would lower the average procurement cost of each plane. But even with one less LMSR to buy, the larger number of airlift planes would add considerably to total costs. CBO estimates that Option IV's price tag would be $3.8 billion in 1998 and $21.3 billion over the 1998-2002 period (see Table 11). Through 2020, Option IV would cost $60.5 billion, or about $6.3 billion more than the Administration's plan.

CBO assumed that DoD would purchase C-17s at a maximum rate of 15 per year under a multiyear-procurement contract, as in the Administration's current plan. Under Option IV, DoD would buy its 20 additional planes in 2003 and 2004.

Comparison of the Alternatives by Various Measures

Because they emphasize different modes of lift, the alternatives presented here vary in capabilities as well as in cost. The mix of mobility forces that best serves DoD's needs depends on what types of capabilities de-cisionmakers believe the United States will need for the future, and how much they are willing to pay.

Cost

Four of the five alternatives would cost less to purchase, operate, and support than the Administration's plan. Total costs for Options I, II, III, and V would range from $8 billion to $19 billion less between 1998 and 2020, mostly because of smaller numbers or less expensive combinations of airlift planes (see Table 13 on page 67). Option IV, which includes a larger purchase of C-17 aircraft, would cost over $6 billion more than the Administration's plan during that period.

Among the alternatives that cost less than the Administration's plan, the majority of savings would accrue in the Air Force's budget because of smaller purchases of C-17s. In the case of Options I and II, however, the Army and Navy would face higher costs because they would need to buy larger stocks of equipment for prepositioning, build warehouses, or add an additional LMSR to their budgets. Although the Air Force would buy fewer C-17s under Option III, much of the savings would be offset by purchases of C-33s. In

Figure 11.
Theoretical Airlift Capacity Under the Administration's Plan and Five Alternatives, 1996-2007

SOURCE: Congressional Budget Office.

a. In order to reach the Defense Department's requirement for 49.7 million ton-miles per day of theoretical capacity, the Air Force would supplement military planes with capacity from the Civil Reserve Air Fleet.

the case of Option IV, the Air Force would need to find an additional $6.3 billion in its budget over the next two decades to purchase and operate 20 more C-17s than the Administration now plans to acquire.

Theoretical Airlift Capacity

Theoretical airlift capacity provides a rough measure of the capability of an airlift fleet by showing how many ton-miles of cargo it could carry (under ideal conditions) in one day. The three CBO alternatives that emphasize prepositioning or sealift (Options I, II, and V) would buy a total of 72 C-17s. Including DoD's other military airlift planes, those options would provide a total of 26.5 million ton-miles per day of theoretical airlift capacity by 2007--16 percent less than the current level and 17 percent less than under the Administration's plan (see Figure 11). In order to reach DoD's requirement of 49.7 MTM/D of theoretical capacity, those alternatives would need to rely on the Civil Reserve Air Fleet for as much as 23.2 MTM/D.

Option III, which would substitute purchases of commercial-style transport planes for C-17s, would include roughly the same level of theoretical capacity as the Administration's plan. Although the C-33 is based on a commercial aircraft design that could be delivered fairly quickly, modifying the planes to carry heavier loads or wider pieces of equipment would take time. Thus, the alternative would probably not provide airlift capacity any faster than the Administration's plan. In order to achieve DoD's airlift requirement of 49.7 MTM/D, Option III would rely on approximately the same amount of Civil Reserve Air Fleet capacity as the Administration's plan--a maximum of 20.5 MTM/D between now and 2007.

Option IV, which features the largest purchase of C-17s, would produce the greatest capacity of all of the alternatives. By 2007, the theoretical capacity of military planes under that option would reach 34.1 MTM/D --nearly 9 percent more than the current level and 7 percent more than under the Administration's plan. Option IV would need to rely on the least amount of airlift from CRAF--just 15.6 MTM/D in 2007.

Cargo Deliveries to Major Regional Conflicts

For each option, CBO estimated how much cargo those mobility forces could deliver to a major conflict in Korea, followed shortly by a second regional contingency in the Persian Gulf--the most demanding scenario of the MRS BURU. CBO followed the assumptions of the MRS BURU as closely as possible with regard to warning, call-up of reservists, and the like. But unlike that study, whose scenarios were set in 2001, CBO's analysis assumed that DoD would no longer have C-141s available for strategic airlift operations.

The Congressional Budget Office estimated deliveries using two tools: the Airlift Cycle Assessment System, a spreadsheet model developed by the Air Force; and the Sealift Factors and Closure Approximation Tool, a simulation program that CBO developed. (For a description of both models and more details about CBO's assumptions, see Appendix D.) Data on how much prepositioned equipment DoD would deliver to a major regional conflict are classified, so in estimating cargo deliveries for each option, CBO included prepositioning only to the extent that it would differ from the Administration's plan.

CBO used one assumption in its estimates of cargo deliveries that deserves particular attention. Despite the recommendation of the MRS BURU, CBO did not assume that the Administration's plan includes an additional 280,000 square feet of combat-support and combat-service-support equipment prepositioned on one LMSR or two smaller ROROs. If defense officials carry out that recommendation in the future, DoD will be able to deliver considerably more equipment during the halting phase of a major regional contingency than it otherwise could. In that case, the alternatives that include additional prepositioning (Options I and II) or that emphasize surge sealift (Option V) would compare much less favorably with the Administration's plan than they do here.

Deliveries During the Halting Phase. If a second conflict broke out in the Persian Gulf on the heels of one in Korea, ships prepositioned at Diego Garcia

Figure 12.
Airlift Deliveries to a Conflict in the Persian Gulf, Plus Sustaining Operations in Korea, Under the
Administration's Plan and Five Alternatives (In thousands of tons delivered)

SOURCE: Congressional Budget Office.

a. Because Option IV includes one fewer large, medium-speed roll-on/roll-off ship than the Administration's plan, it would need to deliver additional cargo by airlift that would otherwise have been prepositioned. As a result, Option IV would deliver about 2,400 tons less by day 20 than the Administration's plan.

b. The total amount of prepositioned equipment that the military would deliver to a major regional conflict is classified information, so the prepositioning shown here is the amount that CBO added to or subtracted from the Administration's plan.

would take 11 to 12 days to reach the Gulf region and unload their cargo.⁽³⁾ Before that, airlift and land-based prepositioning would provide the only means of delivering unit equipment and supplies.

Under that scenario, two of CBO's five alternatives might be able to deliver a slightly larger amount of equipment during the halting phase than the Administration's plan. CBO estimates that roughly two weeks after the start of deployments, Options I and II would deliver 90,700 tons by air to a Persian Gulf conflict while also carrying sustainment supplies to combat operations in Korea. That amounts to about 6,700 fewer tons than under the Administration's plan, and the shortfall would grow to 8,300 tons by the third week (see Figure 12). But with an additional 11,400 tons prepositioned either afloat or on land, those options might allow DoD to deliver 3,100 more tons overall. That advantage amounts to an extra 3 percent of the total deliveries that DoD might be able to achieve over the first three weeks of airlift operations.

Between Options I and II (land- and sea-based pre-positioning), it is unclear which would deliver forces more quickly. At first glance, land-based preposition-ing seems to lend itself to faster deployments, since DoD would only need to transport troops to meet up with the equipment. Prepositioning ships, by contrast, would need to steam to the region before their cargo could be unloaded--a process that, in total, could take 10 days to two weeks. But the relative speed of the two deliveries depends to a large degree on where an enemy initiates its attack. If prepositioning sites on land were far from a foe's assault, it might take just as long or even longer for DoD to move that equipment over land than to steam prepositioning ships to the nearest port and unload their cargo.

Perhaps a more important issue than the location of a conflict is whether the Army would be able to regenerate its afloat prepositioning package for a second contingency, as the Joint Chiefs of Staff suggested in 1995. Thus far, the Army has not developed such a doctrine. Without that preparation, the risk exists that if two major conflicts did overlap, DoD might not be able to react quickly enough. In that respect, Option II, with its larger amount of equipment prepositioned on land, might have real advantages over Option I.

The amount of time required to complete airlift deliveries is sensitive to assumptions about how many planes each airfield can accommodate at once, or the "maximum on the ground" (MOG). Constraints on MOG include the amount of ramp space that military planners devote to airlift deliveries and the availability of resources such as fuel, maintenance personnel, and equipment to unload cargo. In most airlift analyses, those factors are lumped together in one or two variables to indicate how well an airfield can accommodate each type of airlift plane at a given time. The physical amount of space that each plane requires is the key constraint that MOG captures. Thus, as that variable is used in airlift models and simulations, an airfield cannot simultaneously handle as many large planes (C-33s or C-5s) as it can smaller planes (C-17s or C-130s).

During the early part of Operation Desert Shield, however, airlift deliveries were constrained not only by lack of access to airfields but also by too few trucks and drivers to move fuel from storage facilities to aircraft that needed refueling.⁽⁴⁾ Once military officials realized the problem, they sent additional trucks and personnel to ease the situation. Some officials believe that during the war, availability of fuel constrained airlift operations more tightly than airfield ramp space did.⁽⁵⁾ If that were true in the future, smaller planes such as the C-17 might not fare any better than larger ones such as the C-5 or C-33.

To estimate delivery times conservatively, CBO constrained the MOG of airfields in the Persian Gulf and Korean Peninsula to roughly the same levels that the Institute for Defense Analyses used in its 1992 study of the C-17's cost and operational effectiveness.⁽⁶⁾ As a result, CBO estimates that the airlift fleet proposed under Option III (72 C-17s and 30 C-33s) would deliver its cargo at a somewhat slower pace than a fleet that included 120 C-17s. By the 15th day after the start of deployments, CBO estimates, Option III would deliver about 4,200 fewer tons of cargo than the Administration's plan. By the 20th day, that difference would amount to 5,300 tons, or roughly 4 percent less.

Because Option IV would purchase the largest number of C-17s (140), it would allow DoD to deliver cargo by airlift more quickly than would the Administration's plan or the other alternatives. CBO estimates that Option IV would move 104,300 tons by air in 15 days of deployments and about 129,000 tons in 20 days. The latter amount is about 9,000 tons more than under the Administration's plan. However, that margin would not be large enough to offset the lower amount of equipment delivered by prepositioning ships. With one fewer LMSR, Option IV would require DoD to transport an additional 11,400 tons of cargo by air during the halting phase. Overall, CBO estimates, an airlift fleet with 140 C-17s would leave DoD with 2,400 fewer tons than the Administration's plan 20 days after the start of deliveries.

Because Option V emphasizes surge sealift rather than airlift, it would deliver cargo at a slower pace than the Administration's plan. In 10 days of deployments to the Persian Gulf, Option V's airlift fleet (like those in Options I and II) would deliver about 4,900 fewer tons of equipment than a fleet that included 120 C-17s. By day 20, the shortfall would be nearly 8,300 tons.

Without additional airlift or prepositioned stocks of equipment to make up for slower initial deliveries, that shortfall would continue to grow until sealift ships began arriving from the United States 25 to 30 days after the start of the deployment. Would an additional LMSR surging from the United States be able to make up for the early shortfall? The answer depends on assumptions about how such a scenario would unfold.

According to military planners, the United States would be more likely to lose territory if it could not deliver forces as quickly as planned. And those U.S. troops who did deploy quickly would be exposed to a higher level of risk, potentially leading to more casualties. If the assumptions behind mobility analyses like the MRS BURU are credible, Option V's shortfall in deliveries could lead to higher costs for the United States--in terms of the number of forces it would need to deploy for a counterattack and quite possibly in the number of lives lost.

Yet some analysts question the underlying assumptions of combat simulations, including whether the United States can realistically deliver forces in the short span of time laid out by mobility planners. If, for example, weather delayed airlift operations, as it did recently in Bosnia, the initial deployment of U.S. forces could take considerably longer than a two- to three-week halting phase. If deployments stretched toward three to four weeks, the first sealift ships would already be arriving from the United States. Thus, barring similar delays in sealift operations, some analysts might consider a strategy that relied more on surge vessels than airlift or prepositioning to be a reasonable approach.

Deliveries of Outsize Cargo. Besides looking at initial airlift deliveries, CBO analyzed how well each alternative would airlift outsize cargo--the largest pieces of equipment that can fit only on C-5s or C-17s. DoD expects outsize pieces to make up a much smaller share of total cargo in a future contingency than they would have in a conflict with the Soviet Union. Nevertheless, some analysts contend that in order to keep military units together during deployment, it is important to have an airlift fleet with planes that can carry all sizes of equipment--bulk, oversize, and outsize.

CBO analyzed early deliveries of outsize cargo to a Korean conflict, the scenario for which DoD officials believe airfields would be most congested. Since Option IV includes more C-17s than the Administration's plan or any of the other alternatives, it would deliver the most outsize cargo by air during the first 20 days of a Korean deployment--5,700 tons, or 26 percent, more than the Administration's plan (see Figure 13).

Option III, with its mix of planes, would airlift only slightly less outsize cargo than the Administration's plan: 300 tons, or about 1 percent, less according to CBO's estimates. By contrast, Options I, II, and V would deliver about 1,200 fewer tons of outsize equipment by air over the first 20 days of deployments. That amounts to about 6 percent less than outsize deliveries under the Administration's plan. However, if military planners selected units for prepositioning that had large amounts of outsize cargo, Options I and II might not result in as large a deficit during the first 20 days.

Flexibility to Handle Deployment Changes. Options I and II rely to a greater degree on prepositioning for major deployments than the other alternatives do. But with such an approach, military commanders would need to be certain of which units they would want to deploy early in each conflict so as to preposition the appropriate types of equipment. In other words, Options I and II might not permit military leaders to rearrange the order in which they deployed particular units as quickly as the other alternatives.

Of the two, however, Option I might allow the greater flexibility. With equipment prepositioned on ships rather than on land, military leaders could make significant changes in where they chose to deploy those units for a major regional conflict. For example, they would have the flexibility of unloading prepositioning ships in Kuwait or in ports farther south in the Persian Gulf, depending on circumstances. They could also use equipment prepositioned on ships for smaller operations elsewhere in the world without the complexity of redeploying units prepositioned on land.

Option III includes more airlift capacity than Options I and II, so it would provide military commanders with more flexibility than either of those alternatives. But depending on what airfields were available to the United States early in a major conflict, Option III might not allow war fighters to adapt as quickly to changing circumstances as the Administration's plan would. Although each commercial wide-body jet in Option III can carry a larger load than a C-17, it requires a long runway with special equipment to unload cargo. The Korean Peninsula and the Persian Gulf both have modern airports with such features. But if the United States could not gain access to those facilities early in a con-

Figure 13.
Airlift Deliveries of Outsize Cargo to a Korean Conflict Under the Administration's Plan and Five Alternatives (In thousands of tons delivered)

SOURCE: Congressional Budget Office.

flict, Option III would not provide as much flexibility as the Administration's plan.

Correspondingly, Option IV would be more flexible because it includes a larger number of C-17s than the Administration's plan. But readers should also bear in mind that by including one fewer LMSR, Option IV would need to rely on airlift to deliver the equipment that would otherwise have been prepositioned afloat. The greater demand on airlift might limit how much of that flexibility military commanders could actually use.

With an additional LMSR used for surge sealift, Option V would allow commanders to alter their priorities about which forces to send based on changing circumstances. But sealift is a slow way to deliver cargo and thus would not help military leaders if they needed to adjust the delivery schedule of units that would deploy during a conflict's halting phase. Only those options with more airlift planes provide both speed and flexibility for making changes in the earliest part of delivery schedules.

Vulnerability to Enemy Attack. If an enemy wanted to slow the pace of U.S. deployments to a major regional contingency, it would be wise to target both airfields and ports. On the surface, then, all types of mobility forces appear equally vulnerable to attack. Yet, on balance, those alternatives that include relatively more airlift might be somewhat less vulnerable. Most regions of the world have more airfields than ports, so an enemy would be less certain of precisely where the United States would send its forces. And each sealift ship or warehouse for prepositioned equipment would constitute a more concentrated load of cargo than would any single airlift plane, providing a greater incentive for attack.

Option II would probably be the most vulnerable among the five alternatives because it relies the most on land-based prepositioning. When U.S. equipment is kept at a fixed site, potential foes have more time to plan how they might stymie its use. Those alternatives that include more LMSRs (Options I and V) are the next most vulnerable because the concentrated loads of those ships make them more attractive to attack than larger numbers of airlift planes, such as in the Administration's plan and Options III and IV.

Flexibility to Airlift Cargo to Smaller Operations

In terms of early equipment deliveries, the scenario of two major regional conflicts would place the greatest demands on U.S. mobility forces. But some military analysts argue that the United States is far more likely to become involved in smaller operations. Since those missions usually take place without the United States calling up reserve aircrews or commercial planes, they can be difficult to conduct.

For that reason, CBO applied the results of a 1995 DoD study that focused on such missions to each of its five alternatives for modernizing mobility forces.⁽⁷⁾ DoD's analysis examined how well fleets with various numbers of C-17s could deliver cargo to smaller airlift operations and how well they could perform special airlift missions. The study involved a large number of classified assumptions about how such operations would proceed and how quickly the United States would need to complete airlift deliveries to keep risk to a minimum. Because of those complexities, CBO could not conduct a similar analysis independently. Thus, the reader should bear in mind that the assessment of risk in this and the following section is DoD's, and its analysis is subject to all of the uncertainties described in Chapter 5. In DoD's study, risk referred to the risk of failing to complete a delivery mission in the required time.

DoD's analysis looked at four representative cases of smaller airlift operations taken from the Administration's Defense Planning Guidance: a peacekeeping mission, a humanitarian assistance operation, an evacuation of noncombatants from a foreign country, and a peace enforcement operation. In the first three cases, the study concluded that airlift fleets with as few as 40 C-17s could conduct deployments with little risk of failing to complete deliveries as quickly as war planners would like. Based on that standard, the Administration's plan and all five of CBO's alternatives would complete such airlift deliveries with minimal risk.

A peace enforcement operation, such as the 1995-1996 deployment of U.S. forces to Bosnia, would be a more difficult undertaking. To protect U.S. troops from warring parties, DoD planners would almost certainly deploy heavier equipment than they would to a more benign peacekeeping mission. Depending on circumstances, the military might also need to complete its deliveries on a quicker schedule.

Based on DoD's 1995 analysis, defense officials concluded that airlift fleets with 72 or 86 C-17s could complete deliveries to a peace enforcement operation with moderate risk of significantly exceeding their deadline. If airlift forces included 100 or more C-17s, military officials believe, they could complete deliveries quickly enough to keep that risk to a low level.

Based on those findings, the Administration's plan and Option IV (with 120 and 140 C-17s, respectively) could conduct a similar peace enforcement mission quickly enough to keep risk low. Options I, II, III, and V, which each include 72 C-17s in their airlift fleet, would complete deliveries more slowly and thus raise risk associated with a longer timeline to a moderate level. Is that amount of risk tolerable? Analysts will undoubtedly differ in their opinions. But for other analyses, such as the MRS BURU, the Joint Chiefs of Staff deemed a moderate level of risk acceptable given budget constraints.

Special Airlift Missions: Long-Range Airdrops, Intratheater Deliveries, and Direct Deliveries

In its 1995 study of smaller airlift operations, DoD also analyzed how quickly various airlift fleets could air-drop U.S. personnel and equipment over long distances. Based on the results of that study, DoD officials argue that the Air Force needs at least 100 C-17s to air-drop the personnel and equipment for today's brigade-size forces at a moderate level of risk. A fleet with 120 C-17s could do so at low risk, they contend.

By DoD's standards, the risk is high that airlift fleets under Options I, II, III, and V could not air-drop brigade-size forces quickly enough to meet the time-lines set by military planners. If one accepts those schedules, only the Administration's plan and Option IV would be able to conduct such missions with low risk. Thus, moving to an investment strategy for mobility forces with fewer C-17s would mean losing the ability to conduct large, long-range airdrops. But readers should weigh whether that capability is one the United States is likely to need for the future.

During a major contingency, military commanders might decide to use some C-17s to move outsize cargo within a theater of operations. That diversion of planes could slow the pace of strategic airlift deliveries from the United States. Based on DoD's study, a combination of 86 C-17s and 30 C-33s could complete the strategic deployment quickly enough to keep risk at a moderate level. But if military commanders chose to dedicate one or two squadrons to intratheater deliveries, defense officials conclude, the Air Force would need to buy additional C-17s. Thus, a fleet with just 72 C-17s would not have enough planes to use some in that way.

That analysis suggests there is high risk that the airlift fleets in Options I, II, III, and V could not simultaneously move equipment within a theater and complete the strategic deployment as quickly as military leaders would like. The Administration's plan and Option IV, by contrast, include enough C-17s to perform both tasks with only a low risk of slowing the overall deployment. Of course, military commanders could continue to rely on ground transportation and smaller airlift planes like the C-130 to deliver cargo within a theater, while dedicating all C-17s to strategic airlift deliveries.

Similarly, DoD assessed how well various airlift fleets would perform if they used some of their C-17s to deliver equipment from the United States directly to forward bases in a Korean conflict rather than to staging areas in the theater. Based on the results of that study, defense officials contend that a fleet with 86 or more C-17s and 30 C-33s could dedicate some C-17s to direct-delivery missions and still run a low risk of completing the overall deployment more slowly than planned. But for fleets with just 72 C-17s and no C-33s, DoD would probably judge that risk to be high. Based on that analysis, using the airlift fleets in Options I, II, III, and V for both direct deliveries and strategic deployments would entail high risk for the latter. The airlift fleets in Option IV and the Administration's plan could perform both such missions with low risk.



Conclusions

Each of CBO's five alternatives reflects a unique combination of costs and capabilities. How to balance the trade-offs between risk and cost depends on the likelihood that the United States will become involved in major or smaller regional conflicts, as well as whether U.S. forces will need to perform special airlift missions. Ultimately, those are subjective judgments that defense officials and the Congress must make.

Alternatives that include more prepositioned equip-ment--either on board ships (Option I) or in foreign countries (Option II)--would cost less than the Administration's plan. Based on CBO's estimates, those savings could be considerable: approximately $18.2 billion or $16.8 billion, respectively, over the 1998-2020 period. However, those alternatives would require DoD to accept some additional risks. For instance, military leaders would need to be sure which units they would dispatch first to major conflicts, since rearranging the priority for deploying units would be more difficult. Of course, redeploying prepositioned equipment to other regions, if needed, would also take time, although Option I would be more flexible in that regard than Option II. And sending troops to meet up with prepositioned gear makes deployments more complicated; the services would need to conduct more exercises in which personnel practice unloading, distributing, and deploying with prepositioned sets of equipment.

A key difference between Options I and II is that prepositioning equipment on an ally's territory can be more sensitive politically than storing it on board ships that steam through international waters or are based at a friendly port such as Diego Garcia. Thus, one hurdle to carrying out Option II would be securing agreements with allied nations to host prepositioning sites. Although such agreements have not been difficult to obtain with South Korea, there have been more sensitivities in placing U.S. equipment in Saudi Arabia and other parts of the Persian Gulf region. Moreover, even if the United States obtained approval to expand prepositioning in a region, those political sensitivities might continue to limit how and when the equipment was used.

However, one benefit of prepositioning equipment on land rather than afloat is that it can be tailored to the scenario at hand. For example, prepositioning sites in the Persian Gulf would most likely include larger numbers of heavy-equipment transports or desalination units than equipment sets in South Korea.

Alternatives that substitute more sealift ships for airlift--either as a platform on which to preposition equipment (Option I) or to surge cargo from the United States (Option V)--also introduce some risks. Since each LMSR carries such a large concentration of cargo, mines or attacks on harbors could cause greater delays in deliveries than if one or two airfields were closed. Deliveries might also be delayed if U.S. ships were unable to transit key choke points such as the Suez Canal or the Strait of Hormuz.

Yet despite those risks, LMSRs provide considerable delivery capacity at relatively low cost. For example, even if DoD purchased a total of 140 C-17s (as in Option IV), the addition of 20 planes beyond the Administration's planned level could not offset one fewer LMSR prepositioned in the Indian Ocean. Thus, according to CBO's estimates, Option IV could not deliver as much cargo to two major regional conflicts during the crucial halting phase as the Administration's plan.

Option III would give DoD some of the advantages of airlift, such as more flexibility to adjust deployment schedules to changing circumstances, at a savings of $8.4 billion over the 1998-2020 period. But like Options I, II, and V, Option III would not include enough C-17s to perform certain types of special airlift missions, such as strategic airdrops of brigades or cargo deliveries within a theater.

Although they cost the most, alternatives that include the largest numbers of C-17s (such as the Administration's plan and Option IV) have some distinct advantages. They would permit the Air Force to conduct a wider range of smaller deployments or special airlift missions with more confidence. And military commanders would have more flexibility to rearrange the order in which they deploy units to a major regional contingency. But whether those advantages are worth the considerable cost of the C-17 depends on whether decisionmakers believe that DoD is likely to need to perform those tasks or to require the flexibility of 120 C-17s in the years ahead.

1. A prepositioned LMSR would need 14 to 15 days to steam from Diego Garcia to the Korean Peninsula and unload its equipment there. The comparable inventory of C-17s needed to move 250,000 square feet of cargo to Korea over the same amount of time (assuming no height or weight constraints) would be 26 to 35.

2. One disadvantage of that arrangement is that the Army and Air Force would need to develop guidelines over which service had control of the ship. Ammunition deliveries are one precedent where military services have shared ship space.

3. That estimate assumes that the Army would withhold its package of prepositioned equipment for the second conflict or regenerate the set after unloading it at the first conflict.

4. Eliot Cohen and others, Gulf War Air Power Survey, vol. 3, Logistics and Support (Department of the Air Force, 1993), p. 101.

5. Jean Gebman, Lois Batchelder, and Katherine Poehlmann, Finding the Right Mix of Military and Civil Aircraft: Issues and Implications, vol. 3, Appendixes, MR-406/2-AF (Santa Monica, Calif: RAND, 1994), pp. 37-38.

6. W. L. Greer, Cost and Operational Effectiveness Analysis of the C-17 Program, Report R-390 (Alexandria, Va.: Institute for Defense Analyses, December 1993), pp. D-22, D-23.

7. "Tactical Utility Analysis: Lesser Regional Conflicts," a classified briefing presented by the Office of the Secretary of Defense, Director for Program Analysis and Evaluation, to the Defense Acquisition Board in November 1995.

---