Finding Common Ground: U.S. Export Controls in a Changed Global Environment (1991)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

APPENDIX A 199 APPENDIX A Report of the Subpanel on Advanced Industrial Materials* EXECUTIVE SUMMARY Nature of the Problem • Although U.S. national security export controls apply only to a limited portion of worldwide trade in advanced materials, their estimated impact on U.S. competitiveness is significant. • In general, it is not the advanced materials that are militarily critical, but rather the application of design and fabrication technology for weapons systems. • The basic contents of advanced materials are generally made public in U.S. patents. The application technology and processing know-how, however, are closely guarded as trade secrets. * The Subpanel on Advanced Industrial Materials was appointed by the Committee on Science, Engineering, and Public Policy to work in conjunction with the main panel to examine the impact of both current policy and alternative future policies on its specific industrial sector. The subpanel was not asked to consider the full range of issues addressed by the main panel; rather, it was given a specific set of tasks to undertake. The subpanel met less frequently than—and independently of—the main panel, and it had considerable latitude in conducting its discussions. Thus, it should be noted that the conclusions and recommendations of this subpanel report, while providing valuable input to the deliberations of the main panel, do not necessarily reflect the main panel's views and, therefore, should not be considered to be a part of its findings. 

APPENDIX A 200 • Although military funding has driven research and development (R&D) in advanced materials in the past, the majority of applications for advanced materials today are commercial, not military. • A significant part of U.S. industrial know-how in advanced materials is being "exported" through the sale of small U.S. companies to larger, multinational firms. • A Department of Commerce study indicates that the United States no longer leads Japan in advanced materials or component technologies that are highly dependent on advanced materials. Consequently, the ability of the United States to control global diffusion of these technologies is limited. Further, there is a growing shortage of domestic suppliers of specialty materials for defense-related purposes. Findings and Conclusions • Advanced materials should be grouped and defined differently than they currently are for control purposes. The physical or chemical properties of materials do not necessarily indicate criticality. Design code and fabrication technology generally lead to military use. • A number of materials currently controlled were developed under Department of Defense (DoD) contract, but they have not yet been incorporated into weapons prototypes or systems. • Advanced materials should be controlled on the basis of their demonstrated ability to enhance significantly the performance of weapons systems. • Based on a selective review of the U.S. Commodity Control List, a number of advanced materials currently controlled for national security purposes should be decontrolled (see Annex A3). THE U.S. ADVANCED MATERIALS INDUSTRY AND U.S. EXPORT CONTROLS Defense-critical materials technologies figure prominently in those emerging technologies identified by a Department of Commerce study as potentially having a multitude of civilian applications and substantially advancing production and quality levels.1 The same study also concluded that the United States is currently behind Japan, and likely to continue to lose ground, in advanced materials and technologies that are highly dependent on advanced materials, such as semiconductor devices, optical electronics, and high-density data storage media. The ability of the United States to compete in the advanced materials market is being further weakened by the sale to large, multinational firms of small U.S. companies that specialize in fabricating advanced materials. Foreign ownership of U.S. materials suppliers also is increasing. Foreign 

APPENDIX A 201 multinational firms active in acquiring U.S. firms include, among others, Badische Anilin Soda Fabrik (BASF), Hoechst Celanese, Imperial Chemical Industries, CIBA-Geigy, Rhone-Poulenc, Morgan Crucible, Bayer, and Kyocera Corporation. The most active buyer has been Courtalds (Great Britain), which acquired 27 U.S.-based materials suppliers and fabricators in the late 1980s.2 Foreign materials suppliers gain further competitive advantage because the U.S. materials industry is stratified. There is limited integration between suppliers and fabricators, which allows for more flexibility in the purchase of raw materials from foreign sources. Many of the foreign raw materials suppliers are increasingly moving up the value-added chain by investing capital in processing and fabrication industries, thus displacing similar U.S. industries (e.g., ferro compounds, smelter and mill products, and tool and die blanks). For example, Malaysia has recently emerged as a leading contender in the supply of aircraft subsystems for such customers as Airbus Industrie, Fokker, Boeing, Donnjer, Mitsubishi, Fuji, and Kawasaki Heavy Industries.3 The result of these factors is a significant diffusion of processing technology in the industry and, in some specialty materials (e.g., specialty steel, functional ceramics, high-purity compounds), a shortage of domestic suppliers for U.S. defense-related purposes. Thus, the ability of the United States to control access to these technologies unilaterally is limited and eroding. In fact, continued controls may have a negative impact on U.S. defense capabilities. Although there are a number of critical military applications of advanced materials, and military funding typically drives R&D in advanced materials, the majority of potential applications for advanced materials are commercial, not military. One result of defense-driven R&D is that much of the technical data related to the capabilities and performance of advanced materials is classified. The lack of publicly available data on the operational performance of advanced materials reduces interest in potential applications and forces firms that may wish to use certain advanced materials in commercial applications to duplicate work that has already been done. In addition, advanced materials can involve a relatively long lead time between development and application, which discourages commercial materials R&D in the United States, where the cost of capital is relatively high. In foreign countries in which the cost of capital is lower, private investment in the commercial development of advanced materials outpaces that in the United States.4 Despite the fact that defense funding is responsible for many developments in advanced materials, advanced materials themselves are not inherently critical. The physical and chemical parameters used to identify materials for control do not necessarily reflect critical use, and military performance 

APPENDIX A 202 characteristics do not necessarily differentiate militarily critical use from commercial use. It would be more useful to control the design of special, military-application materials rather than basic, commodity materials. In some instances, however, controls on the basic material have been maintained while the manufacturing process and end product that use the material have been decontrolled (e.g., polysilicon is controlled but personal computers have been decontrolled). Controlling the export of the material itself does not necessarily control the militarily critical application. For example, canopies for jet aircraft, which can be made from polycarbonate sheet, are controlled as a munitions item. The polycarbonate sheet is controlled for export by the Commerce Department. Although only a certain quality sheet is used, it is the process for forming it into the canopy, not the material itself, that is complicated and protected, even in the United States, for proprietary reasons. In fact, such factors as fabrication and processing techniques and ingredient percentages are closely guarded as trade secrets, but the basic physical properties and contents of advanced materials are revealed in U.S. patents. Given the market implications, materials firms are more likely to reveal specific contents and processing techniques in patents for materials that may be reverse engineered than in patents for materials for which there is little chance of reverse engineering. Thus, export controls on advanced materials may be somewhat redundant in that the most critical aspects of advanced materials fabrication are either closely guarded as trade secrets or published in patent applications. RELATIONSHIP OF ADVANCED MATERIALS AND ASSOCIATED TECHNOLOGY TO MILITARILY CRITICAL WEAPONS SYSTEMS Military funding has been a principal driver in advanced materials R&D since World War II. Significant advances in structural and electronic materials can be traced to DoD funding. Given cuts in military spending, however, defense-related incentives for continued development of an advanced materials technology base are likely to decline. This is particularly important to the U.S. materials industry, because much of the foreign investment in U.S. materials firms in the 1980s reflected an effort to participate in the development of new materials technologies funded by U.S. defense spending. For example, Imperial Chemical Industries purchased the Fiberite and LNP Engineering Plastics divisions from Beatrice in 1985 to gain access to both military and commercial aerospace developments involving advanced composite systems and to gain an avenue through which to introduce its polyetheretherketone (PEEK) thermoplastic resins into U.S. defense programs. 

APPENDIX A 203 The combination of uncertain investment prospects for the U.S. materials industry and the technological lead of the Japanese and some European materials firms is highly significant to the U.S. defense posture because advanced materials figure prominently in DoD's science and technology strategy for the long-term qualitative superiority of U.S. weapons systems. The problem is further compounded by the unspecified lifetimes of military- platform equipment and technologies. Advanced materials that have some application to current military equipment may be controlled, even though they represent relatively old technology and are produced in a number of foreign countries. In fact, many of the avionics suites incorporated in the Boeing 757 and 767, the McDonnell Douglas MD-80 and MD-90, and the Airbus 330 and 340 series aircraft are more advanced and easier to use than those found in the current U.S. inventory of fighters and bombers. On the other hand, advanced materials for which there is no specific advanced development program or weapons procurement plan may be classified or controlled due to their potential applications, and consequently, they languish in the DoD technology base. Examples include superconducting magnets, which are likely to be used in Maglev (magnetic levitation) and people-mover applications before being used in weapons systems, and the National Aerospace Plane (NASP) project, in which the benefits of research in new metal, ceramic, and polymer matrix materials are likely to be applied to high-speed civil transport before they are applied to an aerospace plane to replace the current shuttle fleet. Clearly, technological progress occurs at a pace that outstrips the U.S. ability to incorporate such developments in weapons systems. The subpanel agreed that only materials marked for procurement for a prototype or existing system should be subject to commercial restrictions or export controls. If advanced materials that have high commercial potential continue to be subject to export restrictions, the U.S. technology base will not be able to "breathe," and there will be little foundation for investment risks in research and development. The natural counterpart to allowing freer circulation of U.S. advanced materials would be to source foreign materials technologies for integration in U.S. defense systems if long-term access can be assured. Based on these considerations, the subpanel identified the following as examples of materials that are militarily critical: • chemical weapons precursors • high-temperature, nonablating structural materials for hypersonic aircraft and missiles • silicon carbide, fiber-reinforced titanium aluminides for high-performance military jet engines 

APPENDIX A 204 • Certain optically switched coatings and laminates for spacecraft threat resistance • ultraclean, nickel-based alloys, powders, processes, and products for high- performance jet engine parts • initiators for ultrahigh energy materials • high-performance armor and penetrator materials REVIEW OF THE CONTROL/DECONTROL OF ADVANCED MATERIALS* Current Methodology The subpanel reviewed advanced materials controlled under various export commodity control numbers (ECCNs) and made the following observations concerning the methodology by which the materials are determined to warrant control. A specific analysis of several entries on the Commodity Control List is attached as Annex Al. • The ECCN entries are defined too broadly and the rationale for control is not clearly stated in either the Commerce Department's Commodity Control List or in the Defense Department's Militarily Critical Technologies List. • Although the Commerce Department's technical advisory committees (TACs) are sometimes consulted on the foreign availability of materials, they are not consulted in determining the critical nature of materials. Moreover, the TACs do not adequately interact with the State Department's technical working groups. • The list construction/management process does not take into account the dynamic nature of technology transfer from military to commercial applications, or vice versa. No attempt is made to assess the market opportunity or economic impact of restricting trade in the materials. • Some of the advanced materials on the Commodity Control List have no direct relation to a current DoD mission area (e.g., superconductor magnets, NASP technology). • Foreign countries often have superior capabilities in producing some of the controlled materials (e.g., Soviet Union in energetic materials, Japan in silicon chips). * This review was undertaken prior to the development of a core list of CoCom- controlled items that was begun in the latter half of 1990. Although the analysis of the subpanel remains valid, the categorization and control status of many materials will change when the core list exercise is completed. 

APPENDIX A 205 New Methodology* The subpanel recommends a new methodology based on a series of filters, or stages of review. For example, the first filter might be R/O, where R represents the risk associated with decontrol and O represents the economic opportunity of decontrol. The subpanel identified five categories of risk: V. supercritical and unilaterally controllable IV. supercritical and multilaterally controllable III. supercritical, not multilaterally controllable, or critical and multilaterally controllable II. critical, not multilaterally controllable I. not critical A chart separating certain selected ECCN materials entries into these five categories is attached as Annex A2. Definitions of supercritical and critical (see Annex A2) address whether the material is directly related to a primary mission capability and the effect that an adversary's acquisition and exploitation of the material would have on the balance of power. Opportunity is defined for the purpose of this example as the product of the unit price (dollars/pound) of a material and world market volume. Five categories are then established using appropriate bounds. For example, V. > 200 million IV. 50–200 million III. 20–50 million II. 5–20 million I. < 5 million Numerical values are assigned to the categories of risk and opportunity by using the well-known mathematical relationship 2n, where n is the number of the category. Category V = 32. Category IV = 16. Category III = 8. Category II = 4 and category I = 2. The ratio of risk (R) to opportunity (O) can then be calculated. Bounds are placed on the resulting numbers to indicate items for continued control, items for further consideration, and items for decontrol. For example, R/O > .5, and R > 8, continue control. .5 % R/O % 1 and R % 8, consider further ("middle ground"). R/O < .5, decontrol. Note that this formula favors the risk factor. A chart of this formula applied to certain selected ECCN materials entries is attached as Annex A3. * The methodology described herein was tested on advanced materials only. A more generic methodology applicable to all items is described in Chapter 10. 

APPENDIX A 206 Illustrations An item in risk category V is 3604A, zirconium. Because this material has a nuclear proliferation possibility, it is assigned an R of 32 (n = 5) and an O of 8 (n = 3). R/O then calculates to 4, which indicates continued control. An item in risk category III is 1702A, hydraulic fluid. Because of foreign availability, it is assigned an R of 8 (n = 3) and an O of 8 (n = 3). R/O then calculates to 1, which indicates middle ground, or consider further. An item in risk category I is 1631A, magnets. Because of extensive foreign availability, the item is assigned an R of 4 (n = 2) and an O of 32 (n = 5). R/O calculates to .0625, which indicates decontrol. The second filter or stage of review would then apply to those items in the category that calls for further consideration. Factors for consideration in this stage include the following: • the learning curve and technology diffusion rate associated with the item, • the cost of efficiently controlling the item, • a productivity index of sales per employee multiplied by a skill index, • the ratio of value added to labor costs (capital intensity), and • economic incentives for trade. This stage favors the opportunity factor since most, if not all, supercritical items would be recommended for continued control in the first stage. The third stage of review should include an analysis of the foreign policy objective in controlling the item and foreign commitments to continue control. The results of the third stage of review should be compared with the results of the second-stage analysis to determine the eventual control or decontrol of middle-ground items. A further elaboration on the possible series of filters or stages for list construction is contained in Annex A4. RECOMMENDATIONS Based on its review and analysis, the subpanel makes the following recommendations: • The U.S. government should adopt a multistage control methodology, similar to that outlined above, that takes into account both military and commercial aspects of trade in advanced materials. • Commercial restrictions or controls should be applied only to those advanced materials marked for procurement for a prototype or existing weapons system. 

APPENDIX A 207 • Controls should focus on the design code and fabrication technology necessary for military application, rather than the basic material. • A number of advanced materials should be decontrolled, including those listed under export commodity control numbers (ECCNs) 1648A, 1587A, 1760A, 1749A, 1675A, 1746A and 1631A, 1110A, 1129A, 1145A, 1203A, 1301A, 1561A, and 1635A. NOTES 1. U.S. Department of Commerce, Technology Administration, Emerging Technologies: A Survey of Technology and Economic Opportunities (Washington, D.C.: U.S. Government Printing Office, 1990). 2. Personal Communication, Robert D. Wilson, November 2, 1900. 3. Personal Communication, Robert D. Wilson, November 2, 1990. 4. U.S. Department of Defense, Defense Logistics Agency, Strategic and Critical Materials (Report to Congress) (Washington. D.C., 1990). ANNEX A1 ANALYSIS OF U.S. COMMODITY CONTROL LIST ENTRIES BASE MATERIALS* This is the control entry for the most advanced ceramic and ceramic composite materials. The entry covers high-purity fine powders that are crucial to making high-tech fine ceramics and whose control is of strategic concern. However, the entry also covers many compounds, precursors, and composites and is in effect a ''catchall." Some of these materials, particularly the composites, are becoming significant for military (especially aircraft) applications. For these materials, the technology of synthesis and fabrication is more important than the materials themselves. Many new organic precursors have recently been developed. but others have been known for about 10 years. The best are still being made by the Japanese, despite the investment of many U.S. research dollars in this area. U.S. work on composites, however, is as good as any. Reverse engineering is very difficult. The problem with this control entry is that it contains too many materials, a number of which are not critical. For example, the reason for controlling either silicon carbide or boron carbide powders, as described in the entry, * ECCN 1733A. Contributed by Neil Ault. 

APPENDIX A 208 is not apparent. Some of the precursors should be controlled, but controlling the fibers, the methods for making them, and the composites from them is far more important. Many of the materials in 1733A belong in category II, critical but not controllable. QUARTZ CRYSTALS* I do not know the capabilities of the Soviet Union or East European countries in regard to these crystals, which are used in military equipment. These controls are not a hardship in terms of lost business for U.S. companies, however, and the controls seem to be justified. I am not familiar with the capabilities that exist within the Coordinating Committee for Multilateral Export Controls (CoCom). I would place 1587A in either category III, critical and multilaterally controllable, or II, critical but not controllable. POLYMERIC SUBSTANCES AND MANUFACTURE THEREOF† Nature of Criticality The 12 classes (a through 1) of polymers indicated in this entry can be categorized as follows: 1. Have been commercially available for > 15 years—a, c, h, j, k, l. 2. Have been commercially available for 7 to 10 years—b. 3. Are not commercially available—d, e, f, i. 4. Uncertain—g. None of these materials qualifies as being supercritical or critical. On the other hand, specific formulations, fabricated forms, and recipes for end-use applications may very well be supercritical. Examples of use: Large manufacturing facilities ranging up to 20 X 106 lbs/ yr exist for items listed under category 1 above. Hence, the number of applications is very large. Nomex (c) is used in flame-resistant textiles and honeycomb structures; Kevlar (c) is used in bulletproof vests and as reinforcement for secondary structures in aircraft. Item (b) is proposed for use in fabric filter systems. In terms of product life cycle, materials such as those in (a) and (c) have been available for 20 to 30 years, and there does not appear to be any diminution of interest in these systems. * ECCN 1587A. Contributed by Neil Ault. † ECCN 1746A. Contributed by James Economy. 

APPENDIX A 209 Foreign Availability Polyimide materials are manufactured in Japan and in Germany. Generally speaking, the preparation of polyimides is well known throughout the world. Surprisingly, the polyimide film is excluded from control, even though that system, to my knowledge, is manufactured only in the United States (by Du Pont). Item (b) is also manufactured solely in the United States (by Hoechst Celanese), but I know of little commercial utility for it. Items (d), (e), (f) and (i) are not, to my knowledge, manufactured anywhere. Item (c) is manufactured in the United States, the Netherlands, and Japan. Pilot quantities of Kevlar fibers were prepared in the People's Republic of China in the late 1980s. Item (h) is manufactured in the United States, Japan, England, and Italy. The knowledge for making these polymers is readily available throughout the world. Item (j) is primarily available from England, and the knowledge for fabricating composites from polyetheretherketone (PEEK) (j) remains proprietary with Imperial Chemicals Industries in England. Many of the systems listed under butadiene polymers are manufactured throughout the world, and the specific knowledge to prepare any of these systems is readily available. Rather than enumerate specific items available from foreign countries that are comparable or identical to those manufactured in the United States, it is safer to state that the sophistication of polymer synthesis and scaleup is at a level that almost any country could set up a capability to produce such items. Relative Importance of Design Technology vs. Materials There is little question that polymeric substances afford considerable opportunity for control. Polyimide film is still manufactured only in the United States, primarily because of the large capital investment required. Fibroids of Nomex that are used in paper manufacture remain proprietary knowledge to Du Pont. Typically, the following factors have resulted in one company being the sole manufacturer of a polymer: 1. High cost of capital investment 2. Limited market 3. Strong patents The following factors promote entry into manufacturing a polymer developed by another company: 

APPENDIX A 210 1. The tradition that the second company makes far more money than the first because of its much smaller R&D cost and its ability to exploit developed markets. 2. The ability to skirt existing patents and still produce a competitive material. 3. The availability of patent literature that provides the details of the invention. Substitutability The field of specialty polymers has reached a level of sophistication that one skilled in synthesis and processing can generally design alternative systems to achieve a given end. Further, many of the major manufacturers of polymers in the United States look on themselves as global companies without any particular allegiance to the country of origin. Hence, the site for manufacture of key materials is often determined primarily by economic considerations. Conclusions I would assign all the materials listed under 1746A to category II, critical but not controllable. I would also eliminate controls on those items that are not manufactured today, such as items (d), (e), (f) and (g), although I am not sure about item (g). In fact, the only areas I would consider appropriate for control are those concerned with processing of polymers into finished articles. Even here, only a very few items might warrant control. FIBROUS AND FILAMENTARY MATERIALS THAT MAY BE USED IN ORGANIC "MATRIX," METALLIC "MATRIX" OR CARBON MATRIX* Nature of Criticality • The parameter of 1 X 108 in. restricts the export of all carbon fiber, Kevlar, boron, and silicon carbide composites. • These fibers are used in commercial and military aircraft in many parts of the world. U.S. military aircraft would include the F-16, F-15. F-14, F-18, F-111, B2, AV8-B, and others. • Estimated length of product life cycle for commercial applications is 30,000 flight hours; for military applications, approximately 8,000 flight hours. * ECCN 1763A. Contributed by William Yee. 

APPENDIX A 211 Foreign Availability The fibers in this entry are available from Taiwan, Japan, South Korea, Israel, Germany, England, France, the Netherlands, the Soviet Union, and China. T300/epoxy and H46/epoxy are available from Japan; G30/epoxy is available from Germany. Graft carbon composites are available from Great Britain. Relative Importance of Design Technology vs. Materials The design technology is distinct from the material. It is not likely that design technology could be understood or reverse engineered from the composite material. Substitutability In most cases, other products could not be used to accomplish the same military or critical objective because of the structural tailoring and dielectric properties involved in structural applications. Conclusions These fibers belong in category II, critical but not multilaterally controllable. Perhaps a fiber that has the stiffness of pitch carbon fiber, which has a specific modulus of 1 x 109 in., should be restricted. Another consideration is to restrict the fibrous material by some measure of the dielectric properties as well. POLYCARBONATE SHEET* Nature of Criticality The polycarbonate sheet with the optical property and strength described in the entry is not militarily critical because many of the new military aircraft do not use it exclusively. However, commercial applications, such as window material for office buildings, are gaining rapidly. In any event, this item is not considered critical and is not multilaterally controllable. It belongs in category I, not critical. Foreign Availability The polycarbonate sheet described within this control list entry is considered to be a standard polycarbonate sheet of optical quality and defined * ECCN 1749A. Contributed by William Yee. 

APPENDIX A 212 thickness. This material is manufactured by numerous companies both within and outside the United States. The technology to process this material is therefore not specific to the United States. The number of U.S. manufacturers is too large to even begin to list in this review. However, the following is an abbreviated list of companies in foreign countries that possess the technology and capability to process this type of polycarbonate sheet: Imperial Chemicals Industries England Shell The Netherlands Mobay-Bayer Germany Rhone-Poulenc France Mitsui Japan The manufacturing of this polycarbonate sheet is by no means limited to the countries listed, however. The technology necessary to produce the sheet is old, well distributed, and readily available. Export control of this material would simply limit the U.S. ability to compete in the worldwide polycarbonate market and would not be safeguarding sensitive technology. In addition, limiting the use of this material to Defense Department projects would negatively affect the automotive industry and its conversion to plastics technology. TANTALATES AND NIOBATES* Nature of Criticality • No criticality. These salts have little or no military or nuclear applications, and purities cited are low in the industry. • Lithium niobate is used in the military and in civilian applications as a piezoelectric. • Salts are used primarily as a precursor to making the elements of tantalum and niobium. Salts have a zero product life cycle. Elements are alloy additions and have a product life cycle of the alloys from which they are made. Foreign Availability/Design Technology/Substitutability These salts are available worldwide as commodity chemicals. Design technology is not applicable to this item, and there is no substitutability because these salts are the precursors to the metals. * ECCN 1760A. Contributed by Edward Van Reuth, subpanel consultant. 

APPENDIX A 213 Conclusion These salts are not critical and are not controllable. They belong in category I, not critical. TITANIUM-BASED ALLOYS WITH 12 PERCENT ALUMINUM AS TIAL* Nature of Criticality • Critical as primary structure for the NASP (leading edges) and other hypersonic aircraft. • Domestic funding has been entirely by the Department of Defense. • Key parameters are specific strength and low strain rate above 1000°F. • Length of product life cycle is difficult to extrapolate because titanium alloys are 35 years old and show little sign of degradation. Foreign Availability The following countries have supersonic transport designs: France (Hermes), Germany (Sanger), United Kingdom (Hotel), Japan, and the Soviet Union. Relative Importance of Design Technology vs. Materials Design technology is not distinct from the material. Reverse engineering could be determined from the material. Substitutability There are no substitutes for this material. The nearest substitutes would be nickel-based alloys (which are too heavy), ceramics (which are too brittle), or carbon-carbons (which are too prone to oxidation). Conclusion This entry should be assigned to category III, critical and multilaterally controllable. * ECCN 1672A. Contributed by Edward Van Reuth, subpanel consultant. 

APPENDIX A 214 ANNEX A2 SELECTED ECCN MATERIALS ENTRIES CLASSIFIED BY CATEGORY OF RISK NOTES: A supercritical technology is one whose acquisition and exploitation by a potential adversary would negate or impair a primary U.S. mission capability to such an extent that a major commitment of national resources would be needed to offset the loss. Such technologies typically enable primary mission capabilities or provide a qualitative superiority essential to maintaining the balance of power. Primary missions are those having high mission value and leverage, such as strategic deterrence, power projection, air superiority, or sea control. An item is multilaterally controllable if it is not produced outside the United States, United Kingdom, Canada, Germany, France, Japan, Norway, Denmark, Netherlands, Belgium, Luxembourg, Australia, Italy, Spain, Portugal, Greece, Turkey, Austria, Finland, Switzerland, or Sweden. ANNEX A3 ILLUSTRATIVE APPLICATION OF RISK/ OPPORTUNITY FORMULA TO SELECTED ITEMS ON COMMODITY CONTROL LIST ECCN Risk Risk Opportunity Opportunity Risk ÷ Decontrol? Category Value Category value Opportunity ($) 1733A IV 16 >200 32 0.5 No 1672A III 8 <5 2 4 No 1588A IV 16 5–20 4 4 No 1763A III 8 >200 32 0.25 Yes 1573A III 8 20–50 8 1 Maybe 1574A III 8 20–50 8 1 Maybe 1648A I 2 <5 2 1 Yes 1587A II 4 >200 32 0.125 Yes 1760A II 4 <5 2 2 Yes 1749A I 2 50–200 16 0.125 Yes 1675A II 4 5–20 4 1 Yes 1746A II 4 >200 32 0.125 Yes NOTE: ECCN = export commodity control number. 

APPENDIX A 215 ANNEX A4 TOWARD A NEW MECHANISM OF TECHNOLOGY EXPORT CONTROL Peter Cannon Conductus Inc. RATIONALE FOR A NEW MECHANISM One of the principal difficulties with the export control process is the severity of export restrictions, which are, in themselves, unexpected in a free market society. They have become more onerous because the U.S. economy has, once again, become mercantile. The nation has evolved from a "Fortress of Freedoms" to a "Purveyor of Values"; it exports concepts of government and law, along with increasing quantities of goods. Increasingly, the United States should promote truly progressive trade practices, rather than enforcing restrictions. The United States has accumulated a number of contradictory and inhibiting bilateral agreements, which have not kept pace with rapid political changes. In addition, the General Agreement on Tariffs and Trade (GATT) talks are not moving fast enough to provide conditions for the expansion of U.S. trade. The champions of the current export control regime are the "Whigs" of the twentieth century. The appropriate premise for export control has changed from the goal of disadvantaging an adversary to one of global consent that there should not be traffic in arms. In addition to the outdated premise, export controls, which involve case-by- case review by experts, have become cumbersome, nonuniform, and unpredictable in their application. Moreover, engineers who graduated in the 1980s might not recognize some of the obsolete terms used to describe restricted items. It would be desirable to see an end to case-by-case restriction and use of a broad self-administered rule. In such a scheme, export control becomes a system of voluntary questioning by a producer of propriety and a process of exception rather than one of rule. Fixed requirements for analysis are replaced by the conscience of the individual and the language of self-inquiry, which leads to voluntary submission for further examination. A process of self-assessment by those engaged in exports and world trade must be based in the language of trade, that is, in manufacturing, value, time, and usage terms. Since this process would be subjective and imperfect, its utility would be improved by the simple artifice of multiple filters—a cascade of decisions, each perhaps only 80 to 90 percent relevant or accurate, but multiplied through four or five steps to improve performance to a 99.9 

APPENDIX A 216 FIGURE A 4-1 The cascaded filter for export control decision making or 99.99 percent (see Figure A4-1). Such a process, voluntarily certified (within a statute of sanctions for violation) could enable an enormous simplification in the bureaucratic implementation of the policy of export control. At the risk of overemphasis, note that this paper presumes that some control is necessary. The purpose of the paper is to provide notional limits to the requirement of sufficiency. What are the guiding values, the purposes of control? The broad policy principle of U.S. export control is to deny to a potential adversary those articles, and means of manufacture, that could be used in the near term to damage or destroy the property or lives of U.S. citizens (private, public, or corporate), wherever they might be. The general environment for application of export controls is currently one in which invasion or other use of major military force between superpowers is thought to be highly unlikely. The United States should also recognize, from the standpoint of human interdependence, a need for some systematic exclusions from trade controls. Immediately, the export of articles and knowledge intended for the long-term benefit of broad populations everywhere, including U.S. adversaries, should be excluded from control. Thus, the United States should not seek to control the export of the following: • Food • Medical supplies, including pharmaceuticals • Humanitarian aid • Telecommunications and computer apparatus (at a consumer or user level) • The means of education, even at risk to the current generation of new technology The last category raises the issue of timing in controlling the export of the technical-base infrastructure ("6.1 and 6.2," in Defense Department parlance). 

APPENDIX A 217 A great deal of such information is generated in academic institutions, which pride themselves on being noncontrollable and which depend on the enrollment of foreign nationals to sustain low-cost operations. Moreover, the real issue is the time frame within which such information becomes operational. Much of the information is, at best, existence proof of principle and does not achieve operability, for at least a decade. The possibility of control at any point in the development process reduces the incentive for private development of such information into dual use technology. When the political uncertainties of the federal budget process are added, it becomes totally likely that export- controlled, technical-base information will molder in classification. Depending on its intrinsic merit, such information may in fact be rediscovered or arrived at independently in a more aggressive economy. The exclusions listed above still clearly contemplate the control of weapons exports. But even such a list is equivocal given that there are technologies of transportation and communication that permit delivery or projection of belligerent or military intentions, as well as provide useful social functions. Among these key areas are telecommunications encryption, heavy- lift space vehicles, and even fuel-efficient jet transports if built in quantities in excess of prompt civil demand and stockpiled. It is these ''dual use" uncertainties that have created a complex bureaucratic conflict among cabinet departments and that have rendered the purpose of export control, except in the case of obvious weapons, moot. What is needed is a real simplification of purpose and process. This paper advances the idea that goods have clear characteristics that can be described in the language of business and that could enable a voluntary screening of prospective exports in the interest of national security. Those characteristics can be expressed in quantitative microeconomic terms that relate to military sensitivity. An important condition for application of this concept is that the economic interest of producers must frequently coincide with the broad policy interests of the United States, as exemplified by the proposed liberalization of trade with specific members of the former Soviet bloc. The natural wish to protect profitable trade today mimics the wish to maintain military advantage. This is because unlike the situation as recently as 40 years ago, today's U.S. military advantage depends on technology and value- laden means of force multiplication and projection. So it is the value, value density (some quotient of price over unit volume or mass), and value rate of change that should form the basis of the criteria for self-inquiry. In the particular case of advanced materials, the primary means of insertion into trade is through substitution of improved parts during system upgrading. Examples include substituting a gallium arsenide processor or high-density mass storage subsystem for older, less capable components and retrofitting a composite wing for aircraft. Here, the justification and motive are clearly 

APPENDIX A 218 known in economic terms, and thus, the argument for the use of economic criteria, as opposed to implication of military value, is highly relevant. The system use is known, and the cost numbers are available. It should be possible to state clearly which part of the dual use item relates to unacceptably sensitive military technologies without breaching security. THE CASCADES OF FILTERS The first decision filter is specific value density. Here, the selling price (market value) is divided by a dimension, usually weight. This quantity for a number of familiar industrial products is shown plotted against annual production, on a log-log scale, in Figure A4-2. This kind of aggregate treatment has a number of flaws, but the value of such a broad treatment is that it clearly shows regimes of value—commodity products (at the bottom); items that should be in museums, military arsenals, or other noncommercial protected environments (at the top left); and items of dual use, high commercial value as well as military potential (in the middle). It is highly instructive to reflect that the two internal bounds between inexpensive goods and those priced by their technical labor or scarcity on the one hand, and between obviously scarce or costly components and nuclear materials essential for weapons making on the other, can also be identified with the value of bullion silver and the value of bullion gold. It could be concluded from Figure A4-2 that anything worth less than the value density of silver should not be controlled at all; anything over the value density of gold requires case-by-case clearance; and items in the middle ground should be subject to further examination. The second filter could be technical value density. This might be expressed as the professional labor content of the product divided by its market value. In general, high-tech products have high professional labor content, and custom systems exhibit the highest such quotient. The highest values would lead to probable control, the lowest would probably lead to decontrol, and the middle ground would require yet a further look (Figure A4-3). The third filter might involve the rate of change of the technology, as exemplified by the change in the value of the product as experience with the product accumulates (the slope of the so-called price-experience curve). In this well-known business analysis, price (or more rigorously cost) per unit of experience is plotted on a log scale against the log of the cumulative number of units of experience (Figure A4-4). The slope is naturally negative, given competition, and typically is between-.1 and-.2. These values are usually quoted in percentages, 10 percent or 20 percent, and the minus sign is dropped. The larger the value, the more rapid the evolution (downward) of competitive market value, which implies a more vigorous competition and which can imply a more rapidly improving product technology. It is in these 

APPENDIX A 219 FIGURE A 4-2 Value density versus volume latter, fast-moving areas that dual use technologies can usually be found. A zero slope, that is, stable prices regardless of volume, usually implies complete control of the commercial marketplace under protection or monopolistic control— the same consequence that is sought under rigorous export control. So, while the mapping of the evolution of market value onto the evolution of technology is inevitably imprecise, the broad correlative truth appears valid. Finally, one has to think about the broad social consequences of excessive export control. After reviewing the above criteria, it is easy to conclude that an overly rigorous control program cuts a nation off from exporting the 

APPENDIX A 220 FIGURE A 4-3 Screening on professional labor content highest value-added products it is capable of building, that is, export control damages the national capacity for wealth generation. It is important to recognize that this affects more than the high-tech professional. Most manufacturing businesses depend on a ratio of contribution of value to pay and benefits for their profitability. In mass production, the pay and benefits are traceable to planned rates and numbers of direct labor employees. The ratio is usually known and is frequently used as a management control variable. In high-tech or custom engineering businesses, it is essential to include professional labor in this measurement, and the apparent ratios are less than for regular, repetitive production. In the latter case, the notion of the percentage gross prime margin (GPM)—the complement of the direct manufacturing cost proportion of the selling price—is sometimes also used as a general indication of the health of the enterprise or product line. This latter measurement should be 80 percent or higher to ensure the continued ability to employ a professional work force in direct support of production, and it is fairly typical of young instrument or electronics activities. These are also typical areas of sensitivity to military use or purpose. Therefore, the fourth filter would be the percentage GPM. A product family that has higher than 80 percent GPM should be considered for control; between 80 and 55 percent, a "maybe" for control; and anything less than 55 percent GPM (i.e., a 

APPENDIX A 221 FIGURE A 4-4 Rate of technical change screen (identified with price-volume experience curve) benefit of less than about twice), decontrolled. For final filtering, total business volume should be considered. Given the administrative advantages of sequential, quantitative filters using the producer's own statistics, it seems an attractive possibility to reduce the current complexity and uncertainty (almost arbitrariness) of the existing export control process.