Managing Materials for a Twenty-first Century Military (2008)

The National Defense Stockpile (NDS) was conceived after World War II as a vehicle for managing risks to the U.S. military-industrial complex posed by shortages arising from military conflicts—see Chapter 4 for a discussion of the evolution of military planning. Over the lifetime of the NDS, its composition and inventory levels of strategic and critical materials have changed considerably as it prepared for a potential mobilization. In the early days, these materials included minerals as well as other supplies such as rubber, whale oil, and goose down. But the environment the NDS operates in is much different. U.S. defense strategy and wartime planning have changed substantially; the globalization of manufacturing and the supply of raw materials has taken hold; and supply chains and the tools for their management have evolved.

Currently, the U.S. maintains in peacetime the military forces that are considered necessary for potential conflicts, reducing the need for a major mobilization and expansion of force levels. To the extent that military production needs to be expanded in response to a conflict, the expansion will be to replace lost equipment or manufacture new equipment to address new threats experienced during the conflict. In response to these changes in force planning and in estimates of the reliability of foreign materials suppliers, the materials requirements and the inventory of the stockpile have changed considerably over the decades. However, there have been significant lags at a number of points: between changes in military planning and the scenarios used for modeling stockpile requirements, between stockpile requirements and legislated stockpiles goals, and between goals and NDS inventory levels.

After the end of the Cold War, stockpile materials requirements dropped precipitously. By 2005, the inventory of goal materials¹ had declined to $90 million for only three materials, identified by a materials requirements decision process. Understanding that process has been a central pillar of the committee’s assessment of the relevance of the configuration the NDS and its assessment of what general principles might be applied to the operation of some future stockpile-like activity.

The NDS operates under the authority of the Strategic and Critical Stock Piling Act.² This act provides that strategic and critical materials be stockpiled by the U.S. government to decrease and preclude, when possible, a dangerous and costly dependence by the United States upon foreign sources for supplies of such materials in times of a national emergency. Under the law, the Department of Defense (DoD) is required every other year to recommend requirements for materials already in the stockpile and others it believes should be in the stockpile. Each biennial report includes assumptions used in making the recommendation. The act directs that those assumptions be based on a military conflict scenario consistent with the assumptions used for budgeting and defense planning purposes. Essentially the law mandates that the process for setting materials requirements include a conflict scenario defined by the following:

• The length and intensity of the assumed conflict;

• The structure of the military force to be mobilized;

• The losses anticipated from enemy action;

• The military, industrial, and essential civilian requirements to support the national emergency;

• The availability of strategic and critical materials from both foreign and domestic sources during the mobilization period, the military conflict itself, and the subsequent period of replenishment, taking into consideration possible shipping losses; and

• Civilian austerity measures required during the mobilization and conflict periods.

According to the law, stockpile requirements are to be set for those strategic and critical materials the United States needs to replenish or replace within 3 years of the end of a military conflict scenario, based on the principles outlined above.

They are to be based on replenishing all munitions, combat support items, and weapons systems that would be required after such a military conflict.

At the time management of the NDS was moved to DoD, a change was made: A more detailed analytical economic modeling of materials supply and demand was developed as the foundation of the requirements identification process. The modeling was and continues to be coordinated and executed by the Institute for Defense Analyses (IDA) under contract to the Defense Logistics Agency (DLA), the parent agency of the National Defense Stockpile Center. The Office of the Secretary of Defense reviews the results of the IDA modeling and makes recommendations to the Congress.

The legislation governing the NDS says

The strategic materials in the NDS inventory as of May 2007 are listed in Table 6-1. The distribution by commodity type and by value, as reported in the most recent Strategic and Critical Minerals Report to the Congress, is shown in Figure 6-1. Of all these materials, only two (beryllium and quartz crystals) were being actively retained at the time of writing this report (June 2007). All the other materials were determined to be in excess and are being sold off. But how is the determination made about which materials to hold given the revolutionary changes in the armed forces over the last half century and, by extension, the material needs of those forces.

There are three broad groups of strategic materials for which stockpile requirements are determined: standard materials, specialty materials, and nonmodel materials (DoD, 2005). In this most recent analysis, requirements were determined for 36 standard materials based on econometric modeling of the supply and demand for strategic and essential civilian materials needs under specified conflict scenarios (Table 6-2). Requirements were also estimated for 17 specialty materials (Table 6-3). The requirements for the three nonmodel materials—that is, beryllium, mica, and quartz—were based on interagency consultations chaired by the DLA.

The modeling process estimates the demands on the economy (industry) for essential civilian and defense goods and services; the resultant demand for strategic and critical materials (SCMs); and the shortfalls in those SCMs by comparing the useable SCM supplies with the SCM demands. Any materials that fall short of these estimates become candidates for stockpiling.

There are three pillars in the modeling process that IDA uses: military planning, large-scale econometric models to forecast materials needs, and forecasts of domestic and foreign supplies.

The prevailing military scenario of the time is at the heart of the modeling process that leads to a determination of NDS requirements. The NDS requirement process is, therefore, strongly linked to U.S. defense planning through the National Military Strategy or the National Defense Strategy, which are based on periodic reviews of the prevailing threats to the United States and U.S. interests. The strategic construct underlying DoD’s force planning has changed over time and can be expected to change (at least marginally) every 4 years, when a quadrennial defense review (QDR) is mandated. Chapter 4 discusses the recent evolution of military planning in some detail, but essentially the approach to force planning has been largely based on the need to be able to fight two nearly simultaneous conflicts and, possibly, another smaller conflict. Though the process tries to anticipate future conflicts, by its nature force planning has historically reacted to rather than anticipated them.

Extraordinary military demands for a given conflict scenario are estimated using a force mobilization model known as FORCEMOB. A multitude of variables need to be considered for a given scenario, including the duration of the conflict, the number and types of forces, the rate of mobilization, the consumption rates of expendables, transportation needs, and so on. FORCEMOB is used by DoD to model the effect on the U.S. industrial base of an extraordinary military demand during conflict and reconstitution. It is one component of the Joint Industrial Mobilization Planning Process, an analytic process that links warfighting needs with industrial capabilities. FORCEMOB considers demands from the armed services and takes into account the capacities of an economy-spanning set of industries, including industries that would be affected, although only secondarily, by a military conflict.

The two quantitative models that have been developed to calculate detailed industry demands given an economic scenario are also pillars of the materials requirements process undertaken by IDA. Demands for goods and services are

estimated for a set of 320 industry sectors that span the entire U.S. economy using the Long-term Interindustry Forecasting Tool (LIFT) model developed by the Interindustry Forum at the University of Maryland.³ LIFT is a large-scale model of the U.S. economy that builds up macroeconomic forecasts, and it is used to understand industry behavior in response to market conditions by calculating industry sectoral outputs based on econometric forecasts of demand for each good, as well as the dynamically changing structure of the economy. The LIFT model also calculates prices for each industry, based on unit intermediate costs and value added.

Also used is the Interindustry Large-scale Integrated and Dynamic Model (ILIAD) model that translates the forecast from LIFT to a finer level of industrial detail. ILIAD maintains detail for 320 industrial sectors and calculates outputs, imports, prices, and employment (Inforum, 2007).

The quantities of the strategic and critical materials needed to produce the

forecast goods are then estimated using materials consumption ratios (MCRs), which are estimates of material consumed per dollar of industrial output in a given sector and are used to convert an estimate of industry output to demand for a specific material in that sector. MCRs are based on available data on materials consumption in manufacturing sectors across the economy and are developed with assistance of the Department of Commerce, the U.S. Geological Survey (USGS), and the Census Bureau.

The third pillar of the process is the supply of strategic and critical materials from domestic sources and reliable sources of strategic and critical materials. Supply forecasts are compared, on a time-phased basis, to the forecasted materials demands of the United States over the duration of the particular conflict scenario, which can include a reconstitution phase. Factors affecting materials supply forecasts include assessments of the threat posed by potential enemies and other hostile countries and estimates of the reliability of foreign infrastructure, foreign excess capacity, and the risk of domestic port damage. An important set of variables is those for country reliability which express the U.S. vulnerability to supply disruption by nations that may be unfriendly or uncooperative to the United States and its interests during a time of crisis—all these factors depend on the nature of the conflict scenario. Generally, reductions in the imports of strategic materials are assumed to occur at the start of a military conflict and allowed to run for the entire duration of the scenario being modeled.

At the final point in the modeling process, integrating the military planning scenarios, the econometric models, and the supply scenarios results in a detailed time-phased picture of the supply and demand for each modeled strategic and critical material over the duration of the conflicts. It is from this analysis that potential shortfalls in the supply of strategic and critical materials are identified and stockpile requirements established.

IDA completes the modeling process, first for what is called the base case, which involves a variety of scenarios coupled with the U.S. macroeconomic outlook and reliability factors for foreign suppliers, as discussed above. The elements of the conflict scenarios are based first of all on the guidelines in the legislation that governs the NDS. The scenarios also make certain assumptions about the nature of the conflict—including the length and intensity, the size and types of military forces to be mobilized, and the potential losses that could be incurred. The modeling process is then also run for various what are called stressing excursion

scenarios, both less stressed and more stressed, whereby other disrupting events that could occur during the conflict are introduced to determine the sensitivity to base assumptions and to estimate the full range of possible SCM supply and demand. Possible disruptive events include a partial disruption of oil supplies from the Middle East, uncertainties about the reliability of countries that supply SCMs, alternative economic scenarios in the United States, including a sudden fall in the value of the dollar (“dollar shock”), and variations in demand in the civilian economy during a conflict.

The committee was somewhat surprised, given the complexity of the IDA process for modeling materials requirements, that since 1999 the NDS requirements as reported to Congress have remained largely unchanged despite the considerable changes that have occurred, as discussed throughout this report, in globalization, U.S. industrial capacity, the status of U.S. mining operations, and military planning.⁴ All this change notwithstanding, in the past four requirements reports—with the single exception of the 2001 report, which identified a requirement for antimony—the DoD has consistently reported to Congress the same three materials requirements: mica, quartz, and beryllium hot-pressed powder metal. The committee decided, therefore, that the details require closer inspection.

The committee had trouble comparing DoD-recommended materials requirements with the materials goals set by Congress since DoD’s requirements reports are more frequent than the legislation establishing specific goals. Although the goals are established in authorization legislation that is enacted annually, they have been set not annually but only every few years.

Before the 1990s, the Congress tended to accept the DoD’s recommendations and the relevant goals were set. But as the stockpile moved into a sales mode in the early 1990s, while Congress continued to request that DoD provide it with a report on stockpile requirements, Congress did not automatically accept the department’s recommendations. As a consequence, the legislated goals were at times higher than DoD requirements, and sales of materials considered excess by DoD were delayed while awaiting legislation authorizing their sale. It is noteworthy, however, that in recent years, there has once again been little difference between DoD requirements and legislated goals. At the time of writing, the last authorization was in FY2002 and was effective December 12, 2001.

Table 6-4 shows the NDS requirements from 1989 to the present and Table 6-5 shows the materials goals from 1999 to the present. After the end of the Cold War,

the committee found that DoD requirements declined more rapidly than the legislatively authorized goals. This may be partly because requirements reports are issued more frequently than the goals are changed in law. While excesses or shortfalls in the inventory have generally become disposition or acquisition goals, respectively, there has not been one-to-one correspondence between the goals for the stockpile and the requirements based on modeling; the reason for this discrepancy is not altogether clear.

A review of DoD’s reports to Congress on the NDS requirements from 1989 through 2005 shows how these modeling efforts have evolved over time. The underlying structure of the model has remained remarkably constant. The Office of the Under Secretary (Policy) gives the Joint Staff a war planning scenario, and based on this scenario the Joint Staff develops detailed time-phased production requirements for weapons systems and other materiel. Current inventories and assumptions about consumption, attrition, and other variables are used in developing these requirements. A “translator” is used to aggregate wartime procurement by DoD into a set of demands for output by U.S. industry. An input-output model then determines the total demands on all U.S. industries associated with wartime defense procurement. To defense total demand on U.S. industries (that is, taking into account intermediate goods purchased from other industries that are used in the production of defense goods) is added nondefense total demand (that is, civilian demand) for goods purchased from these same industries. The resulting defense and nondefense demands are then compared to supply available from U.S. industrial capacity and imports. Any apparent shortfalls stimulate additional investment in industrial capacity.

The resulting total industry demands by sector are then multiplied by a set of detailed material input coefficients that are developed through a separate process. These material input coefficients are based on historical data and have not included any direct impact of higher prices in stimulating substitution or reducing consumption. (Civilian austerity assumptions and the definition of “essential civilian demand,” it might be argued, implicitly reflect the effects of some unspecified price-driven substitution reducing demand for nonessential civilian goods.) The resulting demands for modeled materials are compared to estimates of available imports and U.S. production of those same materials, derived primarily from data supplied in early years by the U.S. Bureau of Mines and in later years by the U.S. Geological Survey. Available imports reflect assumptions about disruption and the reliability of supplies developed through another process. Any deficits in net supply of the studied materials generate requirements for inventories to be held by the NDS.

Civilian demands are based on macroeconomic forecasts of overall economic

growth coupled to a module translating an overall economy-wide forecast of growth (a macroeconomic forecast) into growth rates in civilian demand by industry. The resulting profile over time for civilian demand by sector is then adjusted for reductions in “nonessential civilian demand,” which reflect assumptions about austerity measures imposed during wartime, formulated by DoD in collaboration with civilian agency advisors.

Noteworthy developments in the evolution of this modeling framework over time include the following:

All scenarios historically have been provided by the Under Secretary of Defense for Policy and do not reflect independent judgments by the Under Secretary of Defense for Acquisition, Technology and Logistics or by IDA. From 1989 to 1993, DoD requirements reports assumed a conflict of indefinite duration and modeled materials requirements for the first 3 years of the conflict. The 1989 report assumed 1 year of warning, in which materials stocks could be built up; the 1992 report dropped the assumption of a warning year, and the 1993 report added a 3-year mobilization period coupled to a reduced force structure. The 1993 report assumed that the indefinite-duration conflict would be a nonnuclear, conventional war, so the NDS requirement for beryllium was dropped.

The 1995 report hypothesized 7 to 9 years of warning, 2 to 4 years of mobilization, and an approximately 3-year-long conflict—3 to 4 months of intense conflict, followed by a 2-year stalemate, followed by another 3 to 4 months of concluding conflict. There continued to be no beryllium requirement.

From 1997 through 2001, the assumptions were two major theater wars (MTWs) with little warning. Beryllium was added as an “off-line,” special-study-based requirement in 1997.⁵ The 1999 scenario specified a 1-year duration for the MTWs, followed by a 3-year regeneration period. The 2001 scenario continued these assumptions.

In 2003, the MTWs were renamed “major conflict operations” (MCOs) but otherwise continued the overall 2001 assumptions. In 2005, the assumptions were altered to include, within a single year, a catastrophic attack on a U.S. city, two nearly concurrent MCOs, and a smaller scale contingency. As before, the 1 year of conflict followed by a 3-year regeneration period drove requirements models.

From 1989 to 1992, the input-output model and macroeconomic growth forecasts for the civilian economy were provided by the Defense Research Institute. From 1993 through 2005, the University of Maryland’s Inforum group provided these models.

The 1989 report specified that civilian motor vehicle demand would fall by 50 percent in the first year of a 3-year conflict, followed by a 75 percent reduction in year 2 and a 100 percent reduction in year 3. Residential construction was assumed to drop 50 percent in year 1, 67 percent in year 2, and 75 percent in year 3. Other structures dropped 25 percent in year 1 and 50 percent in years 2 and 3. In later reports, these assumptions grew into more finely specified reductions in civilian spending: In 2005, the assumptions included reductions in autos, boats, aircraft, and recreational vehicles, jewelry, foreign travel, auto accessories, other consumer durables, gasoline and oil, and both residential and nonresidential construction. Consumer spending reduced by austerity was shifted into other categories of consumption.

The 1995 report assumed that emission controls on motor vehicles would be relaxed in wartime, so no stockpile requirement for platinum or palladium was recommended. The high dollar cost of a platinum stockpile was mentioned as a justification for this assumption.⁶ Reports after 1995 returned to the assumption that emission controls would not be relaxed; apparently other changes in assumptions were sufficient to ensure that there would be no requirement for high-value platinum.

Reports from 1989 through 1993 were based on static materials consumption coefficients derived from data from the early 1980s. These were not updated until the 1995 report. Reports from 1997 on updated at least some of these coefficients as part of the modeling process.

As previously remarked, requirements for indium, rayon, rhodium, and ruthenium were specified by DoD in 1989. These “new stockpile materials” became the basis for a new category of “advanced materials,” with requirements determined by special off-line studies. In 1992, this category included five materials (the above four plus germanium). By 2003, it had grown to 19 advanced materials (including, as the committee has noted, beryllium). It is noteworthy that the econometric modeling methodology was not made use of in determining requirements for these newer materials.

The treatment of investment in new capacity appears to have been refined substantially over time. The 1995 requirements report excluded partially finished weapons and platforms produced during a conflict period from materials requirements; prior and later reports continued to count such work in progress toward materials requirements. Finally, the methodology for calculating availability of imports, based on supplier reliability and risks of disruption, has also been much refined over time.

Of course, over the history of the stockpile there were times when Congress and the administration had different policy perspectives. There were also instances when DoD did not propose in its annual budget to fund the acquisition of materials required to meet projected shortfalls. The reason for this is also unclear. A case in point is the most recent materials requirements report from DoD, which identified requirements for only three materials even though the IDA process had identified a potential shortfall in a number of materials.

The IDA report in question, also the most recent at the time of writing, was issued in 2005 following IDA’s analysis of 55 materials. Five materials were identified that might experience supply shortfalls in the base case, 15 were flagged in the more stressful scenario, and 3 in the less stressful case. The 15 materials in the more stressful case were antimony, bauxite, beryllium, bismuth, boron nitride, cobalt, fluorspar, mica, palladium, quartz crystal, natural rubber, tin, titanium sponge, tungsten, and yttrium. The most influential variables in modeling the shortfalls in the more stressful case were foreign excess capacity and foreign infrastructure reliability on the supply side and nonessential civilian demand and “dollar shock” on the demand side. The largest single requirement, which is an ongoing requirement,

was 171 short tons of hot pressed powder (HPP) beryllium, valued at $89.6 million as of March 31, 2005.

It is noteworthy that despite the different outcomes listed above, the 2005 DoD report to Congress recommended the stockpiling of only three materials: HPP beryllium metal, mica, and quartz crystal. The committee was also struck by the fact that these three materials requirements resulted not from the IDA analysis but from the interagency process, which involved a more subjective analysis of the potential for a disruption in the supply chain. Specific materials currently in the inventory, along with the recommended requirements as of May 2007, are shown in Table 6-1. Mica is no longer a required material. DoD has indicated to the committee that the requirement for stockpiling mica was dropped after one of the services indicated that it no longer needed that mineral. The committee notes also that that both quartz and beryllium are being held to guarantee a supply in peacetime and not just for use in a national emergency. Although the peacetime supply is important for defense, it seems to be inconsistent with the currently stated purpose of the stockpile. A case study for beryllium is contained in Appendix F.

At the center of the operation of the NDS, there has been a continual refinement of the models and of the inputs to the models, as well as refinement of the interagency process for vetting assumptions. This is a considerable effort that has resulted in the accumulation of significant expertise and the ability to identify materials requirements on the basis of a high-caliber input-output economic approach. But is the econometric approach yielding results that reflect actual military needs? What of the legal framework within which the NDS operates? Because both concerns affect the configuration of the NDS, the committee believes a comprehensive assessment of the NDS is warranted. Clearly, since the end of the Cold War the need for the NDS appears to be waning. The changes in the IDA modeling process—for example, changing supplier reliability estimates, military planning, and so on—have resulted in outcomes that predict minimal materials requirements. There have been no upgrades or acquisitions since 1997, and NDS inventories have fallen in value from over $15 billion in the mid-1980s to about $1 billion in 2007, with further declines anticipated. The predicted materials requirements have also fallen in value from a high of some $20 billion in 1991 to three materials valued at approximately $100 million in 2005. This quick look at the situation leads one to wonder what the point is of having a stockpile.

Looking back at the history of the stockpile (Chapter 2), the most dramatic swing in stockpile requirements and the concomitant shift in U.S. policy toward the stockpile since World War II occurred in the late 1980s and into the 1990s in the aftermath of the Cold War. Conflict scenarios underwent sharp revision and the reliability of foreign suppliers of materials was reevaluated. While it is to be expected that in a given year the requirements for a given material may be higher or lower than current amount held in the stockpile, the data in Figure 6-2 show not much of a relationship between the value of stockpile goals—that is the materials identified as being required and authorized to be held—and the value of the actual inventory in the stockpile, even over the long term. Stockpile goals have varied considerably depending on the policy for assessing requirements at the time, and inventory, with an exception in the early 1980s, just declines steadily.

The modeling methodology used to estimate stockpile requirements has undergone significant refinement since 1988. While limitations exist for all econometric models, and the committee has not undertaken an in-depth audit of the models

FIGURE 6-2 Stockpile goals and inventory (billion dollars). SOURCE: Data from U.S. census reports and presentations to the committee.

used, the current model-based process to set materials requirements appears to be robust and well developed and to properly investigate a broad range of scenarios that generate materials requirements. However, the committee is concerned that the econometric models lack the specificity to identify actual military materials needs.

The committee is struck by the lack of coordination across the DoD and the military services to identify specific individual and shared materials needs. The IDA modeling to estimate stockpile requirements is done on the basis of economic models and not on specific DoD requirements. There may be specific high-priority needs for materials that are not being addressed owing to a lack of information on material usage. Indeed the fact that recent requirements for mica and quartz came from a process other than the modeling just underlines the weakness of the outcomes of a modeling process no matter how robust it is. These two materials requirements came not from the requirements model but from off-line policy considerations.

It appears that DoD itself has been concerned about such specific materials requirements for some time. The committee notes that a DoD instruction, DoDI 4210.8, issued on April 15, 1972, states as follows: DoD Components will obtain Bills of Materials from contractors for all major Class A items of procurement, and will keep their Bill of Materials files current. Class A items were defined as “usually specially-designed military end items or components” containing any controlled materials. Controlled materials were defined as certain shapes and forms of steel, copper, aluminum, and nickel alloys. To the extent that DoD is concerned about certain materials in the future, the instruction could be updated to obtain information on those materials. That information could be integrated into a summary of critical materials needs and the results used to determine potential materials problems during peacetime as well as during a national emergency. It is not clear that this instruction has ever been implemented.

The NDS operates within a defined legislative framework. However, the committee found a number of instances in which this operational framework could do with some clarification.

The first instance needing clarification arose from the committee’s comparison of the results of the requirements analysis and what is reported by DoD as requirements in its biennial report to Congress. In preparing the report, the Secretary of Defense is directed by law to use the following guidance: “The stockpile

requirements shall be based on those strategic and critical materials necessary for the United States to replenish or replace, within three years of the end of the military conflict scenario required under subsection (b), all munitions, combat support items, and weapons systems that would be required after such a military conflict.”⁷ However, the DoD requirements for beryllium HPP, mica, and quartz in recent reports came not from an analytical analysis but from an interagency process considering the supply chain needs for those materials. While this process is not inconsistent with the stated purpose of the law, which is “to provide for the acquisition and retention of stocks of certain strategic and critical materials and to encourage the conservation and development of sources of such materials within the United States and thereby to decrease and to preclude, when possible, a dangerous and costly dependence by the United States upon foreign sources for supplies of such materials in times of national emergency,”⁸ from a practical standpoint, the direction in law on how the DoD is to determine requirements for the report may be overly restrictive and may not reflect actual national security needs.

The second instance needing clarification concerns the limitation on operation of the stockpile, by which the Congress must approve all acquisitions and disposals in nonemergency periods unless the action is directed by the President of the United States. The committee notes that requirement may keep the NDS from taking advantage of short-term changes in market prices and demand when disposing of material. If there were a requirement to acquire a material, the congressional approval process could slow down the acquisition even if the stockpile Transaction Fund had sufficient excess cash to carry it out.

The third matter needing clarification is foreign sources. As the U.S. economy has become more globalized, defense production has also become more globalized. DoD defines a strategic material as “material required for essential uses in a war emergency, the procurement of which in adequate quantity, quality, or time, is sufficiently uncertain, for any reason, to require prior provision of the supply thereof.” This would cover all materials regardless of where they are obtained. In contrast, the legislation governing the NDS defines strategic and critical materials in a more restrictive way in that such materials “are (B) are not found or produced in the United States in sufficient quantities to meet such need.” As pointed out in

previous chapters, most materials are not found or produced in the United States in sufficient quantities to meet all U.S. needs. The purpose of the stockpile, as stated in the law, is “to decrease and to preclude, when possible, a dangerous and costly dependence by the United States upon foreign sources for supplies of such materials in times of national emergency.” Given the fact that defense production has become more globalized in recent years, some clarification in defining strategic and critical materials that takes into account current supply chain operations, would help to identify the materials that should be considered for stockpiling.

The committee concludes as follows:

• The list of specific materials that have been considered for the stockpile over the last 20 years or so has largely been static. It is narrow as well, especially in light of emerging and future materials needs.

• There have been significant time lags between changes in military planning and their reflection in the scenarios used for modeling stockpile requirements; between DoD identifying its stockpile requirements and Congress legislating the stockpile’s goals; and between legislating the goals and achieving the inventory levels in the NDS.

• The goal materials that are being held and are not for sale by the NDS were not identified by detailed econometric modeling methods and are not based on changes in military scenarios.

• The NDS goals have not changed in response to changes in military planning scenarios.

The committee believes that the key to any analysis of the NDS operation today is that in the 1990s the materials requirements reported to Congress shifted such that they now equal zero, except for three items, bringing into question the need for a stockpile. The committee considers the main cause of this shift was the interpretation of the “dangerous and costly dependence” clause of the law governing the NDS. The committee heard from DoD that in the 1990s, it believed the more globalized supply for defense systems, components, and raw materials (as opposed to reliance on an entirely domestic market) would mitigate the risk of dangerous and costly dependence, especially when there was a stated willingness to pay any price required for defense-related raw materials. Such a policy position seemed to be justified as raw materials became more available in the 1990s. However, the committee became convinced during the course of this study that (1) the increasing demand from large developing economies, (2) the recent decline in the capacity of U.S. industry to supply and process raw materials for defense systems, and (3) the continuing increase in U.S. dependency on foreign sources for materials call for a

fresh assessment of the risk and a new policy response. The committee’s analysis identifies a potential today for disruption in the supply of materials and minerals critical to the U.S. military that economic modeling might not identify. There is the possibility, therefore, that the risk of a more dangerous and costly dependence is not being identified by the current system. It seems that the law, written so long ago, requires updating to better define what a dangerous and costly dependence might be in the twenty-first century. A new system needs to be established to assure the supply of materials critical to defense systems.

Department of Defense (DoD). 1989. 1989 Report to the Congress on National Defense Stockpile Requirements.

DoD. 1992. 1992 Report to the Congress on National Defense Stockpile Requirements.

DoD. 1993. 1993 Report to the Congress on National Defense Stockpile Requirements.

DoD. 1995. 1995 Report to the Congress on National Defense Stockpile Requirements.

DoD. 1997. 1997 Report to the Congress on National Defense Stockpile Requirements.

DoD. 1998. Strategic and Critical Materials Report to the Congress, Operations under the Strategic and Critical Materials Stockpiling Act During the Period October 1996 through September 1997.

DoD. 1999. 1999 Report to the Congress on National Defense Stockpile Requirements.

DoD. 2001. 2001 Report to the Congress on National Defense Stockpile Requirements.

DoD. 2003. 2003 Report to the Congress on National Defense Stockpile Requirements.

DoD. 2004. Strategic and Critical Materials Report to Congress, Operations under the Strategic and Critical Materials Stockpiling Act During the Period October 2002 through September 2003.

DoD. 2005. 2005 Report to the Congress on National Defense Stockpile Requirements.

DoD. 2006. Strategic and Critical Materials Report to Congress, Operations under the Strategic and Critical Materials Stockpiling Act during the period October 2004 through September 2005.

Inforum. 2007. A Review of Inforum’s Iliad Model. University of Maryland. Available from http://www.inforumweb.umd.edu/Iliad.html. Accessed July 2007.

This page intentionally left blank.