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.
COperation and Effects of U.S. Export Licensing for National Security Purposes STEPHEN A. MERRILL Senior Sta;f{Consultant INTRODUCTION Determining the economic effects of U.S. national security export controls requires a detailed understanding of how technology trade is conducted, how U.S. controls operate, how they compare with the control systems of other countries, and how controls interact with relative prices, productivity, product quality, and other factors to affect interna- tional competitiveness. In administering controls, government officials and private practition- ers acquire knowledge of or an intuitive feel for only certain pieces of this complex puzzle. For purposes of analysis, large pieces are missing altogether and must be assembled from many sources public and private, domestic and foreign. The magnitude of the task is illustrated by the lack of data on essential elements of the analysis. Volume and Structure of Affected Trade The U.S. control system not only affects direct exports from the United States but also reaches sales of U.S. affiliates and foreign firms where these involve resale of U.S. products or original sales of foreign products incorporating U.S. components or technology. Apart from aggregate figures on the number and value of individual export license applications and approvals, detailed information on the amount and composition of business affected is not readily available. Trade data on exports and 221
222 APPENDIX C foreign sales are reported for industrial categories that correspond only roughly to those on the list of controlled commodities. Operation of U.S. Controls An elaborate set of procedures is required of all companies that export controlled products and data; but the scope and mechanics of corporate compliance vary with the commodities being exported, their origins and destinations, and, especially, the type of validated license employed. In addition to licenses for individual exports and reexports (individual validated licenses, or IVLs), the Department of Commerce issues bulk licenses (distribution, service supply, and project licenses) permitting multiple trans- actions in controlled products and services with approved customers in Free World countries over a limited period of time. Substantially more informa- tion is available on the processing and use of IVLs, although even those data are incomplete and in some respects misleading because they do not include actual shipments and relate only to the government's handling of license applications. Virtually no information is available on distribution licenses, the most widely used bulk export authorization. U.S. Versus Foreign Control Systems If U.S. export controls were identical to those of other Western countries with competitive suppliers, their economic costs would be confined to the costs of compliance and of proscribed trade. The effects on relative competitiveness would be negligible. In fact, U.S. controls exceed those of other members of the Coordinating Committee on Multilateral Export Controls (CoCom) in their complexity, product cov- erage, and extraterritorial reach. Moreover, the U.S. export licensing process appears to be less efficient and predictable. Together, these differences make it likely that U.S. firms bear competitive as well as administrative costs in complying with export controls; but there are large areas of uncertainty. One uncertainty is precisely how national control systems vary, not only in formal requirements but, more importantly, in practice. Information even on the former is spotty. Another major uncertainty is how foreign purchasers perceive, weigh, and act on the differences in control systems when choosing among suppliers. Evidence on this score has been strictly anecdotal. Export Controls Vis-a-Vis Other Competitive Factors U.S. export controls may have a net negative competitive effect, but their impact may be relatively slight if U.S. firms are able to offer more
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 223 advanced technology, lower prices, or better service support than their foreign competitors. The effects of the control system cannot be viewed in isolation from exchange rates, relative productivity, comparative quality, and other competitive factors. No analysis has attempted the difficult task of disentangling these erects at a particular point in time, let alone attempted to forecast their variation and interaction over time. To begin to fill in these gaps, the National Academy complex's Panel on the Impact of National Security Controls on International Technology Transfer commissioned several studies, of which two are summarized in this appendix. They are (1) Quick, Finan & Associates, Inc., "Analysis of the Effects of U.S. National Security Controls on U.S.-Headquartered Industrial Firms,"* and (2) Stephen A. Merrill, "International Business Under the Distribution License."t These studies deal in varying degrees with the issues described above. First, an effort was made to determine the value and composition of U.S. foreign sales of manufactured goods affected by national security export controls. This analysis included direct exports and sales by U.S. affiliates. With the exception of the relatively few reexports for which they seek separate U.S. authorization, the analysis did not include independent foreign companies' sales of products with some U.S.-controlled content, on which data are not available. Second, the administration of export controls, from government and private perspectives, received considerable attention. Most of it focused on the processing of individual validated licenses and individual reexport authorizations because delays and uncertainty in the handling of IVLs have long been considered to be significant problems, especially for smaller U.S. exporters. Third, the studies attempted to ascertain how U.S. companies and, indirectly, foreign firms view the relative coverage, stnngency, and effi- ciency of the control systems of the major Western industrialized countries. A separate consultant study,:" which appears in the companion volume of this report, represents the first published comparison of the major features of the control systems of five CoCom and two non-CoCom countries. The findings generally support the private sector perceptions described below. *A report in two volumes submitted to the panel August 25, 1986, by Quick, Finan & Associates, 1020 Nineteenth Street, N.W., Suite 340, Washington, DC 20036. Principal investigators were William F. Finan and Karen M. Sandberg. tA report submitted to the panel September 1986 by Stephen A. Merrill, independent consultant, 148 Eleventh Street, S.E., Washington, DC 20003. Both reports are available for a nominal fee from the National Academy of Sciences. 1:"A Study of Foreign Country Export Control Systems," prepared for the National Academy of Sciences by International Business-Government Counsellors, Inc., October 1986.
224 APPENDIX C Finally, the two studies attempted to determine the administrative costs to U.S. businesses of complying with U.S. export controls and, in two separate contexts, to estimate the magnitude of the competitive effects of these controls. In both cases, discrete actions by the U.S. government to relax existing controls or to impose new controls permit quantitative analysis of the effects of controls on the operations of U.S. companies. In one instance, it was possible to link regulatory changes to changes in the level of U.S. exports in a particular product group after accounting for the effects of fluctuations in exchange rates, foreign industrial production levels, and prices. A preliminary estimate of the aggregate economic costs associated with certain features of the U.S. export control system is described in Appendix D. DATA SOURCES AND LIMITATIONS The Quick, Finan & Associates and Merrill studies rely on data from four principal sources U.S. government trade and foreign investment data, the Commerce Department export licensing data bank,* surveys of U.S.-based companies affected by national security export controls, and interviews with corporate officers responsible for compliance. The major components of the studies are summarized below. Commerce Department Data Bank 1. A sample of 500 cases was drawn at random from 3,613 individual validated licenses approved between April 15 and April 30, 1986. The approvals represented 85 percent of the 4,259 cases for which processing was completed in the 2-week period. Data collected on each case in the sample included processing time, destination country, U.S. Control List number (ECCN), value of shipment, and company size (based on number of employees, determined independently). 2. A random sample of 200 cases returned without action (RWAd) was drawn from the total of 20,675 RWAs during calendar year 1985. Repeat RWAs are included in the sample, which was taken prior to a change in Commerce Department procedures intended to reduce the number of and turnaround time for RWAs. The data collected were the same as those for the sample of approved licenses. *The panel obtained access to the data bank for its consultants under a "national interest" exception to the nondisclosure provision, Section 12(c), of the Export Adminis- tration Act of 1979. Protection of the confidentiality of business information is a strict condition of such access.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 225 3. A sample of 761 reexport authorization applications was drawn from all requests originating in five Western countries during the first quarter of 1986 and in a sixth Western country during 2 months of that quarter a total of 1,894 cases. In addition to value, ECCN, and country of destination, the data collected included the country of origin and whether the applicant was a U.S. or foreign company. 4. A sample of 1,617 cases (approvals, denials, and RWAs) was drawn from the total number of IVL cases (approximately 3,200) for which processing was completed during the week of June 2, 1986. Commerce Department licensing officers examined each case in the sample to determine, regardless of the actual destination, whether the proposed export was (1) within the CoCom administrative exception note (AEN) limits and thus eligible for export to the Soviet bloc without referral to or approval by CoCom; (2) above the AEN limits but within the China "green zone" limits and thus eligible for export to the People's Republic of China without CoCom concurrence; (3) above the green zone limits but eligible for export to approved Western affiliates, distributors, and end users under a distribution license; or (4) ineligible for export under a distribution license, thus requiring an IVL to all destinations. As stated, these categories are mutually exclusive; but they represent progressively higher levels of military sensitivity or criticality, as determined for administrative purposes in U.S. government deliberations and CoCom negotiations.* Additional information collected on each case included the ECCN, destination country, value, and processing time. Survey Questionnaires 1. With the cooperation of 10 industrial trade associations, a question- naire prepared by Quick, Finan & Associates was mailed to U.S.-based member firms in the aerospace, machine tool, electronics, medical equipment, robotics and automated manufacturing, and instrumentation industries. Because of multiple association memberships, many firms received two or more questionnaires. The 170 respondents were esti- mated to account for about one-third to one-half of U.S. aerospace, electronics, instrument, and machine tool exports in 1985. The question- naire focused primarily on experience in the use of IVLs. 2. With the assistance of the Commerce Department, a separate questionnaire prepared by Stephen Merrill was mailed to all recent and current holders of distribution licenses. One hundred seven (107) compa *In theory, items of greater military sensitivity are subjected to closer scrutiny by one or more agencies of the U.S. government and, if destined for the Eastern bloc or China, by CoCom.
226 APPENDIX C nies or corporate divisions holding a total of 116 licenses responded with general information on their distribution license activities and detailed information on one or more licenses. The responses represent 18 percent of the estimated 650 distribution licenses outstanding between the first quarter of 1985 and the second quarter of 1986.* Only 10 companies responded to both surveys. The characteristics of distribution license holders in both samples were similar except that respondents to the Merrill survey included relatively fewer very large firms and none in the aerospace industry. Of course, neither survey sample can be considered to be representative of the total population of U.S. exporting firms. Interviews Follow-up interviews were conducted with several survey respondents who indicated a willingness to confirm, clarify, or elaborate on their written answers. The companies participating in the interviews were of various sizes and in diverse industrial sectors. In reviewing the findings described below, the reader should bear in mind several limitations of the analysis. First, no attempt was made to assess directly either the effectiveness or the benefits of export controls in preventing or delaying technology transfers to the Soviet bloc. Indeed, it is not possible to determine from the data collected the extent to which U.S. firms are complying with controls, although it is possible to reach some judgments about the relative degree of compliance with U.S. reexport controls by U.S. and foreign firms.T Second, because all but a few survey respondents were active or recent users of validated export licenses, the data derived do not yield any estimate of the extent to which the costs, complexities, or uncertainties of the licensing system deter companies from exporting controlled items.! *In its FY1984 Export Administration Annual Report to Congress, the Department of Commerce reported 780 outstanding distribution licenses. As late as May 1986, when the questionnaire was mailed, the department's mailing list contained a number of duplications, companies that were no longer active license users, and one-time applicants that had not received a distribution license. The estimate of 650 active distribution license holders is that of Commerce Department licensing officials. fin interviews, several firms suggested that stepped-up enforcement activities, especially the Customs Service's Operation Exodus, account for a large share of the recent increase in individual license applications about 70 percent since 1981. Compliance with reexport controls is discussed below. fInterviews were conducted with a few small company survey respondents that did not report any licensing activity. They indicated that the system's complexity has a deterrent effect but that the magnitude is uncertain. For many small and medium-sized firms that deal in controlled products within the Free World, the control system apparently discourages marketing to Eastern bloc countries even of products that do not require validated licenses.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 227 On the other hand, the surveys were designed to elicit information on the experiences of companies of different sizes.* Third, with the exception of a few U.S.-based affiliates of foreign- owned companies, foreign firms were neither surveyed nor interviewed. Evidence of their views and behavior with respect to U.S. export controls was obtained indirectly, from the U.S. companies with which they have done business. Fourth, the data collected relate primarily to U.S. exports and affiliate sales of manufactured goods. Controlled transactions involving services or transfers of technical data not directly related to product sales were not examined in detail. Fifth, an effort was made to distinguish national security export controls from controls for foreign policy, nuclear nonproliferation, short supply, and other purposes; but it is not always possible to isolate data on national security controls. The IVL samples obtained from the Depart- ment of Commerce, for example, contain a small proportion of foreign policy cases. Similarly, survey respondents may in some cases have included information on transactions controlled for other than national security reasons. The distribution license, for example, is available for items to countries subject to nuclear nonproliferation controls. Finally, the data were collected for periods ranging from a few weeks to several months between 1984 and 1986 during which administrative and regulatory changes were being implemented. From one point of view the analysis represents a snapshot of conditions in flux. From another perspective, however, the studies capture many of the regulatory changes instituted since the late 1970s as part of a general policy of strengthening export controls and improving some aspects of their administration and enforcement. Unfortunately, there are no historical data with which to compare current and past experience of the operation and effects of the control system. COVERAGE OF NATIONAL SECURITY EXPORT CONTROLS The segment of U.S. trade covered by national security export controls must be broken down by type of transaction (i.e., export or other foreign sale), type of license, commodity group, and destination to give an accurate picture of the scope and incidence of controls. The narrow objective of preventing or delaying Soviet acquisitions of products and know-how with significant military value tends to obscure the fact that *Depending on available data, various measures of firm size were used in the analysis- number of domestic employees, annual domestic sales, and annual foreign sales. Where data are reported below by size of firm, the measure is described.
228 APPENDIX C controls affect a large share of U.S. international business, primarily with Western countries. Exports The Quick, Finan & Associates and Merrill studies first attempted to determine the aggregate value of direct U.S. exports of manufactured goods under validated licenses. (Commerce Department data in combina- tion with survey data indicate that the total in 1985 was on the order of $62 billion, or nearly 40 percent of all U.S. exports of manufactures. The total comprised the following: 1. Exports under individual validated licenses. In FY1985,* the Com- merce Department issued licenses for approximately $50 billion of man- ufactured goods. Included in this figure was approximately $6.4 billion in reexport authorizations. In briefing the panel, Commerce Department officials estimated that about 85 percent of the value of approved individual licenses is actually shipped. The Quick, Finan survey elicited an identical estimate, which did not vary by firm size. Furthermore, although the $50 billion of approved licenses does not include military equipment licensed under the International Traffic in Arms Regulations (ITAR), it does include a small percentage-probably as little as 1 percent-of items controlled for foreign policy reasons. Thus, the value of national security controlled, dual use manufactures exported from the United States under IVLs in FY1985 was approximately $36 billion. 2. Exports under distribution licenses. Respondents to the distribution license survey reported that in calendar year 1985 they exported nearly $3.7 billion in manufactured goods under 109 licenses, representing 17 percent of the estimated 650 distribution licenses outstanding in 1985. Large companies holding distribution licenses may have been under- represented in the sample. In that case the estimate of $22 billion in exports under such licenses in 1985 is conservative. This figure is significantly higher than a recent Commerce Department estimate (of $12 to $15 billion) that was derived from a sample of 1985 shipper's export declarations submitted to the Bureau of the Census. The latter sample excluded export documen- tation filed electronically, typically by large exporters. The distribution license is not available for the most sensitive dual use products, for munitions, or for items restricted to particular countries for foreign policy reasons; but it is available for items controlled for nuclear nonprolifera- tion reasons, some of which may be included in the estimate. *Department of Commerce licensing data are reported on a fiscal year basis. Survey responses were on a calendar year basis, as are U.S. government trade data.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 229 TABLE C-1 U.S. Exports of Manufactures Under Validated National Security Export Licenses in 1985 (in billions of dollars) 36 22 Exports under individual licenses Exports under distribution licenses Exports under other bulk licenses (project, service supply) 4 TOTAL 62 SOURCES: Commerce Department license data; Quick, Finan and Merrill surveys of U.S.-based firms. 3. Exports under other bulk licenses. Survey respondents reported that their 1985 shipments of goods under service supply and project licenses did not exceed 1 to 2 percent of their total exports. The value of all 1985 exports of manufactures under these types of bulk licenses was approx- imately $4 billion. (See Table C-1.) U.S. Affiliate Sales U.S. business activity affected by national security export controls is not limited to direct exports from the United States but extends to sales of licensable commodities by U.S. corporate affiliates abroad. It is impossi- ble to determine precisely the value of affiliate sales under validated U.S. export licenses, but it is possible to derive rough approximations. In 1982 affiliate sales were nearly 30 percent of the worldwide sales of U.S. companies; and in five industries heavily affected by export con- trols, affiliate sales were about one-fourth of parent company sales ($77 billion of $309 billion).* Survey respondents who fell mainly into these five industrial categories reported that 60 percent of their 1985 foreign sales (exports and affiliate sales) were under validated licenses. Because of the reexport authority the distribution license accords approved foreign consignees, U.S. multinational companies rely heavily on the distribution license to cover the activities of affiliates. The Merrill survey obtained estimates of the value of affiliates' 1985 sales under this type of bulk license. For a sample of 112 licenses, the figure was approximately $3.8 billion and for all distribution license holders was probably on the order of $22 billion-or roughly equal to the value of direct U.S. exports under distribution licenses. - *See U.S. Department of Commerce, Bureau of Economic Analysis, U.S. Direct Investment Abroad: 1982 Benchmark Survey Data (Washington, D.C.: Government Printing Office, December 1985), p. 16. The industry categories were office equipment, other machinery, electrical equipment, other transportation equipment, and instruments. , .
230 APPENDIX C TABLE C-2 Destination of U.S. High-Technologya Exports (in billions of dollars) Destination1980 Percent1985 Percent CoCom less Canada21.3 3926.7 39 Canada5.5 108.1 12 PRC<1.0 <11.7 2 Bloc<1.0 <1<1.0 <1 All other countries27.0 4930.9 46 TOTAL54.7 10068.4 100 aDoC-3 definition; see the note in Chapter 3. SOURCE: U.S. Department of Commerce. Between 30 percent and 40 percent of U.S. parent companies' exports are to foreign affiliates for their use in manufacturing facilities, incorpo- ration in finished products, or resale in their original form.* It may be supposed that U.S. export controls would not affect this large intrafirm component of export trade; but this assumption ignores the fact that affiliates' sales of controlled products are themselves subject to licensing and accompanying restrictions on reexporting. For multinational firms that market primarily through foreign affiliates, any adverse competitive effects of controls initially will show up abroad in affiliates' sales perfor- mance. Eventually, U.S. export trade will be affected; but the linkages back to U.S. corporate operations may be difficult to measure directly. Destinations of Controlled Exports The vast bulk of controlled U.S. trade is with Western countries, nearly one-half of it with other CoCom members. In the sample of IVL applications on which final action was taken in the first week of June 1986, Eastern bloc applications accounted for 3 percent of the total number, the PRC for 6.5 percent, CoCom countries for 46.2 percent, and other Western countries for 44.3 percent. The value of license applications in the sample by destination shows roughly the same distribution. And both distributions are similar to the breakdown of all U.S. high-technology exports by destination, excluding Canada for which validated licenses are not required. (See Tables C-2 and C-3.) High-technology trade with China - *Shares of exports to affiliated and unaffiliated parties vary widely among high-technology industries affected by controls. Intrafirm trade exceeds exports from the United States to independent firms in computing/office equipment, electronic components, and instruments. In machinery, communications equipment, and transportation equipment the pattern is reversed.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 231 TABLE C-3 Distribution of Sample of Processed IVL Applications by Destination Average Total Percentage Value Percentage Destination Applications of Sample ($000) of Value Bloc 48 3.0 128.10 < 1.0 PRC 105 6.5 1,830.20a 2.9 15 Western countriesb 296 18.3 319.00 9.9 Other Western countries 421 26.0 410.70 19.6 CoCom less Canada 747 46.2 602.30 66.5 aExaggerated by inclusion in the sample of a very large aircraft sale. bNon-CoCom countries subject to a presidential directive authorizing Defense Depart- ment review of certain applications. SOURCE: Quick, Finan analysis of Commerce Department export license data. expanded at a rapid rate between 1980 and 1985, but the level remains relatively small. U.S. high-technology trade with bloc countries has been insignificant in recent years. Product Composition of Affected Exports COMPOSITION BY INDUSTRY CATEGORY Notwithstanding the large volume of trade and broad range of products affected, export licensing is concentrated in a relatively few industrial categories that account for a large share of U.S. exports of manufac- tures-electronic components and computers, aircraft and aircraft en- gines and parts, instrumentation, and manufacturing and communications equipment. Table C-4 shows the 10 largest U.S. Control List categories, ranked by value, of approved applications for manufactured goods in FY1985. These 10 categories alone accounted for 92 percent of the value of all approved IVLs. Ranking Control List categories by the number of applications shows a similar high degree of concentration; six of the largest categories by value also appear on this list. (See Table C-5) Use of distribution licenses shows a similar pattern.* The items for which survey respondents had authority to export under their licenses fell into 65 U.S. Control List categories. But by far the largest number of *An exception is the absence from the distribution license sample of aircraft, aircraft engine, and related equipment producers who account for nearly one-quarter of the value of approved IVLs as well as a large number of IVL applications.
232 APPENDIX C TABLE C-4 Approved Individual License Applications for Manufactures to All Countries in FY1985 (ranked by dollar value) Percent No. of Value of Total Rank ECCNApprovals ($ billions) Value Category 1. 156542,110 26.1 52.1 Electronic computing equipment 2. 1460536 7.8 15.6 Nonmilitary aircraft/ helicop ters/engines 3. 44601,100 3.8 7.6 Nonmilitary aircraft/helicopters and engines/equipment 4. 156412,137 3.6 7.3 Electronic assemblies and inte grated circuits 5. 1091513 1.1 2.3 Numerical control equipment 6. 15722,205 1.1 2.2 Recording/reproducing equipment 7. 13552,631 1.0 2.0 Electronic device manufacturing equipment 8. 15191,505 0.6 1.2 Single/multichannel transmission equipment 9. 63993,917 0.5 1.0 General industrial equipment 10. 1485920 0.5 1.0 Compasses/gyroscopes/ acceler ometers TOTAL 67,574 46.1 92.3 SOURCE: Quick, Finan analysis of Department of Commerce export license data. licenses of llWincluded or were limited to electronic computing equipment (ECCN 1565), followed by electronic assemblies and integrated circuits (ECCN1564 41 licenses), electronic device manufacturing equip- ment (ECCN 1355- 16 licenses), single- and multichannel transmission equipment (ECCN 1519), electronic testing and measuring equipment (ECCN 1529), and recording and reproducing equipment (ECCN 15721. It should be noted that in some cases there is little correspondence between U.S. Control List categories and the more familiar Standard Industrial Classifications under which most economic data are collected and reported. For control purposes, for example, a great many products qualify as computing equipment (ECCN 1565) solely because they contain microprocessors. In FY1985 the Commerce Department approved a total of $26.1 billion of IVL applications alone under ECCN 1565; but according to trade data the United States exported only $15 billion of computers in calendar year 1985. COMPOSITION BY LEVEE OF MILITARY SENSITIVITY The distribution of licensed exports by level of military sensitivity shows a heavy concentration at the lower end of the spectrum. In the
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 233 TABLE C-5 Approved Individual License Applications for Manufactures to All Countries in FY1985 (ranked by number of approvals) No. of Percent Rank ECCN Approvals of Total Category 1 a 1565 42~110 48 Electronic computing equipment 2.a 1564 12,137 14 Electronic assemblies 3.a 6399 3,917 5 General industrial equipment 4. 1529 3,325 4 Measuring/ testing equipment 5 a 1355 2,631 3 Electronic device manufacturing equipment 6. 1537 2,350 3 Microwave equipment 7 a 1572 2,205 3 Recording/ reproduction equipment 8. 4529 2,089 2 Equipment, test, computerized electric 9. 6299 1~615 2 Equipment, electrical power-generating 10 a 1519 1~505 2 Transmission equipment TOTAL 73,884 86 aAlso appears on the list of 10 largest categories by value. SOURCE: Quick, Finan analysis of Commerce Department export license data. sample of processed IVL applications categorized by Commerce Depart- ment examiners, slightly more than one-third of the cases fell within AEN limits;* slightly more than one-third fell between the AEN limit and the PRC green zone limit; 20 percent were above the green zone but still eligible for shipment under a distribution license; and 13 percent were ineligible for a distribution license. When actual exports under distribu- tion licenses are taken into account, it is apparent that only a small proportion of licensed goods is considered extremely sensitive in that individual shipments must be reviewed and approved to all countries. Firms Affected by Export Controls Just as controls heavily affect only a few industries, the licensing system applies to a relatively small subset of U.S.-based manufacturers. In FY1985, approximately 250 firms filed 50 or more individual license *Roughly one-half of these AEN-level cases, 17 percent of the total sample, involved exports to CoCom member countries for which validated license requirements were dropped in December 1985. It is unclear why many firms were continuing to submit applications 6 months after the adoption of the general license G-COM. The Commerce Department usually processes applications that need not have been submitted.
234 APPENDIX C applications, with their total (approximately 33,000) constituting about one-fourth of all IVL applications received by the Department of Com- merce. The top 164 firms accounted for about half of the total value of IVL applications. Survey data in combination with Commerce Depart- ment data suggest that between 2,000 and 3,000 U.S.-based firms are applying for licenses each year. Among the 170 respondents to the Quick, Finan survey, 70 percent were small companies (under $50 million in annual domestic sales), 12 percent were medium-sized ($50 million to $250 million), and 18 percent were large companies (over $250 million). The average small firm reported that 68 percent of its 1985 exports were under validated licenses, the average medium-sized firm, 47 percent, and the average large firm, 56 percent. The distribution license is generally regarded as the preserve of large multinational firms, but the Merrill survey revealed great diversity in the size of licensees. Respondents included 44 small companies (less than $25 million annual domestic sales), 39 medium-sized companies ($25 million to $250 million), and 23 large companies (more than $250 million). Nevertheless, the average value of 1985 distribution license-covered exports by firm size was $2 million, $25 million, and $100 million, respectively. Thus, it is the case that large companies, although in the minority, account for most of the exports and an even higher percentage of other foreign sales under distribution licenses. According to the findings of Quick, Finan, exporters that rely entirely on individual validated licenses are almost exclusively small firms. The number of foreign enterprises U.S. affiliates, distributors, and unrelated firms that have some involvement with the U.S. control system is much larger than the number of U.S.-based exporters that seek and use licenses. In April 1985, before the implementation of new Commerce Department regulations that resulted in the removal of a large number of foreign consignees, more than 20,000 consignees were qualified on U.S. distribution licenses, although many were listed on two or more licenses. In addition, several thousand other foreign businesses are affected by the terms of individual validated licenses. ADMINISTRATION OF NATIONAL SECURITY EXPORT CONTROLS The U.S. government's handling of export license applications affects business operations in several ways. A long-standing concern of the business community, Congress, and licensing agencies has been the time it takes to process licenses. License review delays, especially unpredict- able delays, may impose additional carrying and transportation costs,
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 235 erode customer confidence in the reliability of U.S. suppliers, and in extreme cases result in contract penalties or lost sales. Licensing actions short of outright denial, such as the return of applications with requests for additional information or the approval of licenses with conditions on the configuration or use of the product, may have similar consequences. Delays and uncertainty and their associated costs are more trouble- some for small companies than for large ones. This section examines export control administration from both govern- ment and private sector perspectives and mainly with respect to individ- ual validated licenses and reexport authorizations. According to respon- dents to the Merrill survey, the processing of distribution license appli- cations and amendments (e.g., to add foreign consignees or extend their approved sales territories) often entails delays and uncertainty. Never- theless, a distribution license can substitute for tens, hundreds, and even thousands of individual licenses. Distribution license holders prize its flexibility and believe that they could not be competitive without it. As experienced users of IVLs, many distribution license holders associate those licenses with difficulties in responding to consumers, high admin- istrative costs, and a high risk of losing business to foreign competitors. This section also examines other issues of administrative efficiency and corporate behavior. First, to what extent does the licensing system, in practice rather than by design, treat firms of various sizes differently? That is, does the system discriminate against small firms? Second, does the government's handling of export licenses reflect differences in the military importance of various products and in the risk of diversion associated with various country destinations of U.S. exports? In other words, does the system discriminate among transactions in ways that serve the purpose of export controls? Third, to what extent do foreign companies comply with U.S. reexport controls? That is, is this unique, controversial feature of the U.S. system an effective means of control? License Processing Times OVERALL DISTRIBUTION As reported by the Commerce Department, the average processing times for various types of cases and destinations for most of the first quarter of 1986 are shown in Table C-6. For the department the process- ing time extends from the day receipt of a license application is recorded to the day of license issuance or other final action. The overall average processing time according to this definition is 27 days; but the distribution is very skewed. While 74 percent of cases are completed in less than 25 days, approximately 5 percent of cases extend beyond 100 days.
236 APPENDIX C TABLE C-6 Profile of Commerce Department Processing Timesa for Individual Licenses from January 5, 1986, to April 5, 1986 Destination Countries All Free 15 Total Average World Cocom Westernb Bloc PRC All Cases 21 14 36 75 64 27 Nonreferred Cases 16 13 24 29 18 16 Referred Cases 49 31 57 132 156 73 Other Agency Processing Times State Defense Energy CoCom CIA SNECC Other 30 64 6 48 8 o 41 126 o All Agency Average 9 18 o 41 12 o o 0106 76 41 63 40 31 28 114 40 6 41 7 6 41 63 6 41 331 34 71 66 59 44 0 126 16 16 31 8 9 68 aAverage number of days from recorded entry into the system through license issuance. bNon-CoCom countries subject to the presidential directive authorizing Defense Depart- ment review of certain applications because of perceived diversion risk. CSubgroup on Nuclear Export Coordination. SOURCE: Department of Commerce export license data. From the perspective of the firm applying for a license, the processing time extends from the mailing or delivery of an application to receipt of a notice of action. This is a better measure of the system's performance because it governs the timing of transactions and shipments. For the cases approved in April 1986, the average total processing time was 54 days, twice as long as the Commerce Department processing time. One-third of all cases took more than 30 days. Figure C-1 compares the distribution of cases completed in April 1986 according to the two definitions of processing time, and Figure C-2 shows the same comparison on a cumulative basis.* Under the Commerce Department definition, 80 per- cent of the cases are completed in less than 30 days, whereas only one-half of the cases are disposed of in that period under the total processing time definition. *In both figures, cases returned without action (RWAd) are included in the total processing time distributions but not in the Commerce processing time distributions. This extends the right-hand tails of the total processing time distributions but presents a more accurate picture because at the time the sample was taken, RWAs were occurring in about one out of six license cases.
237 i~//////~ ////// i////////// 1~////~ ///// 1 5 _ 1 1 n 5 to 1 rot ,. ~ 1 U 44 o o I £ -1 ~ _ ~9 1 ,4 1 ~ e' - Z to 1 - MU 1 1 - 1 1 1 1 5 5 (9 44~ on Cot o Go ;> o Ct Cot - o s o o ID ra U) o ID o - ID o ID C) it a) to c c) - Q a) Q - ._ C ~ a, C) ~ ~ it: _ o A - C] o _ ~ o E _ o ~ U} o _ ~ E ~c o' LL - C) - Q C' Q .. o ~ C) O Ct ._ ._ C~ U) ._ ~_ ._ C~ U) ao o Ct C~ o o U. o ._ ._ - ~:
238 'A ~//~ a, fin I_ A/////// ~ ~ __~ V ~ O O fin C I////// ~ ~ ////////// ~ 1 V V 1 V C" O 1 V £ at 1 ~ V U M 1 1 1 1 1 1 Us ~Rae to ED ' oo Cot .c ._ o U. U. C) ._ - v 1 P] - , , , m ~M ~ LOG ~ Cal _1~ Ut~ - it on . o x o ~ E Q a) C] a E E o o U) a ~_ ._ C, . . ~: o ~n - Ct ._ :~ ._ .= C~ U. ._ ~4 ~: ._ Cq C) s~ - Ct ~o C~ o C~ o o U. .c Ct o C) :> ._ E C: V
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 239 BY DESTINATION If the sample of April 1986 approvals is divided between West-West and West-East (including PRC) transactions, two very different distributions of total processing time result. More than one-half of West-East cases take more than 50 days to process, but only 15 percent of West-West cases take that long. (See Figure C-3.) Processing times vary not only between West and East but also among Free World destinations and between China and the Eastern bloc in ways that are consistent with U.S. government assessments of the risk of diversion to Soviet military uses or to undesirable PRC uses. Average processing times by the Commerce Department definition were lowest for CoCom cases (14 days), more than double that time (36 days) for the 15 non-CoCom Western countries covered by the President's directive authorizing Defense Department review, almost twice as long again for China cases (64 days),* and highest for bloc cases (75 days). (See Table C-6.) BY LEVEL OF MILITARY SENSITIVITY Although processing times vary with the diversion risk associated with different country group destinations, they do not vary significantly with the military sensitivity of the items proposed for export. A summary of average processing times by the Commerce Department definition for each combination of destination and level of sensitivity is presented in Table C-7. BY FIRM SIZE It was hypothesized that small firms experience longer processing times than large firms because they have fewer resources to devote to coping with the licensing system. Under the Commerce Department definition the distribution of average processing times does not support the hypoth- esis; but a significant difference emerges when total processing times are examined and if destination is also taken into account. On average, small firm applications take 14 percent longer to process. In West-West trade the small firm average is 46 days and the large firm average is 35 days. Almost one-half of large company West-East licenses are approved within *The analysis of processing times was undertaken at a time of transition in licensing procedures for China cases. Among other changes, CoCom agreed to raise the levels of technology requiring CoCom review and approval. Presumably, these changes have reduced average processing times for licenses to the PRC.
240 ~4 3 W · e, 1 1 1 1 ~M tip _I ~/////////////////////~ //////////////// //// _ In 1 -I ~L4 . . 4 - Ct 1 CO 1 v ~^ ;~ I_ ~.. -I 1 ~ V _ .c ¢ 1 1 44 O 1 1 P. 1 £ z ._ a, ° to & C =, C. ^ - U. . o x _ . _ ~.= Q ~ ~._ c) ·- ~ ~0 O ~ ~._ O ~n ,C' U:} ~U. a' C O 3 & IL _ Y ~ ~ o 3 ~ a ~ .. 0 1 1 0 ~: o ~n £ lr, v o' U) 1 1 o~ C~ p4 o E" U' o . _ .c 4 - ._ 1 C: - v
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 241 TABLE C-7 Average Commerce Department Processing Times (in days) by Level of Military Sensitivity and Destination Multilateral Control Unilateral Control Destination AEN PRC DL >DL DL >DL Bloca 46 103 b b b b (27)C (9) PRC 23 81 144 119 b b (39) (41) (8) (11) 15 Western 39 41 36 43 31.9 b countries (89) (110) (43) (36) (10) Other Western 27 35 33 31 35 b countries (158) (129) (59) (38) (21) CoCom 19 26 26 23 21.7 b (270) ~(196) (133) (87) (38) aNo licenses in the sample were to USSR destinations. bInsufficient sample size. C( ) = number of cases. These cases need not have been filed because the items were eligible for general license G-COM as of December 1985. SOURCE: Quick, Finan analysis of sample of individual cases completed in the first week of June 1986. 40 days versus only 13 percent of small company applications. (See Table C-8.* License Actions The Commerce Department denies very few license applications- between 1 and 2 percent. This is not a significant measure of the restrictiveness of the system because companies generally avoid submit- ting applications that are likely to be rejected or that will require . inordinate time and effort to get approved. Moreover, companies occa- sionally withdraw applications (i.e., request an RWA) to modify them or to avoid an adverse decision. More commonly, RWAs occur when the Commerce Department re- quests additional information, suspects errors on the application, or *In the sample a number of extremely protracted Soviet bloc cases submitted by large firms raised their average processing time well above that for small firms. The share of cases completed within 40 days is more representative of companies' experience. The possibility that the firm size effect Is a function of differences in the product composition of small and large firm exports was examined at least with respect to major U.S. Control List categories. The results suggest that line of business does not explain differences in processing time.
242 APPENDIX C TABLE C-8 Total Processing Time for IVL Applications of Large Versus Small Firmsa by Destination Average total processing times All firms Large firms Small firms Large firms 53 days 49 days 56 days Average processing time (days) Percent <40 days Percent >40 days West-West 35 75 25 Percentage of Total 88 12 Small firms Average processing time (days) 46 96 Percent <40 days 70 13 Percent >40 days 30 87 Percentage of Total 91 9 West-Eastb 145 46 54 aSize classification based on number of domestic employees: large firms, >1,000; small firms, <1,000. bIncludes China. SOURCE: Quick, Finan analysis of sample from Commerce Department export license data. suggests modification of the proposed transaction.* Prior to a recent change in the department's handling of RWAs, one of about every six IVL applications was returned for one or more reasons. This rate is somewhat misleading because some cases are RWAd a number of times. Nevertheless, the frequency is sufficiently high to represent a significant share of the licensing load and a significant source of uncertainty for exporters. LICENSE DENIALS Table C-9 summarizes adverse licensing actions for different combina- tions of destination and level of military sensitivity. It shows that no AEN-level items were denied. Although the denial rate was higher for extremely sensitive items requiring an individual license regardless of destination than for items eligible for export under a distribution license, only a handful of cases in each category were rejected. Three were to countries covered by the presidential directive and one each to bloc, PRC, *The consultants did not examine cases in detail to determine either what proportion of RWAs are on applicants' versus licensing officials' initiative or the reasons in either case.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 243 and CoCom destinations. In the Quick, Finan survey, small companies reported a slightly higher denial rate than large companies. (See Table C-10.) LICENSE APPLICATIONS RETURNED WITHOUT ACTION In the sample of RWAs drawn from the Commerce Department data bank, the small firm share was not significantly larger than the small firm share of licenses filed. But small company survey respondents reported an RWA rate more than double that of large and medium-sized firms as well as a higher rate of license approvals with conditions. (See Table C-10.) In the sample of licenses categorized by destination and level of military sensitivity, RWAs were most frequent for CoCom cases. No AEN-level case was RWAd. (See Table C-9.) Reexports Reexports are often authorized in advance in connection with IVL applications for exports from the United States or in qualifying a foreign consignee to resell U.S.-origin goods under a distribution license. Resell- ers and users of U.S. components may also seek a separate individual reexport authorization from the Department of Commerce. It is possible TABLE C-9 Distribution of Adverse Actions on IVL Applications by Destination and Level of Sensitivity Multilateral Control Unilateral Control Destination AEN PRC DL >DL DL >DL Bloc Denied RWAd PRC Denied RWAd 15 Western countries Denied RWAd Other Western countries Denied RWAd CoCom Denied RWAd SOURCE: Quick, Finan analysis of sample from Commerce Department license files. O 0 1 0 O O O O 'O O O 0 3 0 2 o o l O O o o o O O O O 1 0 2 o 4 2 0 2 0 0 O O O o 1 0 0 0 0 9 5 2 1
244 APPENDIX C only to approximate the value of reexports under the first two authorities and only for U.S. affiliates. Depending on firm size, between 15 percent (in the case of small companies) and 82 percent (in the case of large companies) of distribution licenses are used for the reexport authority accorded foreign consignees. Sales by foreign affiliates under distribution licenses in 1985 were on the order of $22 billion. Requests for individual reexport authorizations totalled approxi- mately $6.4 billion and constituted about 10 percent of the Commerce Department licensing load in 1985. Analysis of the sample of these applications (summarized in Tables C-11 and C-12) not only shows their country origin and destination of the reexports but also gives some indication of foreign companies' compliance with this unique, contro- versial feature of the U.S. control system. The overwhelming majority (about 90 percent by value) of reexport applications are from U.S.- headquartered companies and their foreign affiliates. Unrelated foreign firms initiate only 10 percent of applications. The disparity is greatest in the case of applications from CoCom member countries, which are 80 percent of the total number. Between 87 percent and 98 percent of the submissions originating in three major CoCom countries were traced to U.S. affiliates. (See Table C-12.) TABLE C-10 Action on Individual Validated Licenses and Processing Times by Size of Exportera Average Large Medium Small 1. Action on License Applications -Percent approved 91 -Percent approved with 6 modification -Percent denied -Percent RWAd 93 92 88 8 3 7 7 <1 <! 4 2 11 2. Average processing time (in days) -Free World Average 45 38 40 48 Longest 5%b 97 46 68 119 -Communist countries 136 132 174 132 -China 119 107 136 127 aRespondents grouped as follows: large firms, exports greater than $250 million; medium- sized firms, exports between $25 to $250 million; small firms, exports under $25 million. bThese cases were the 5 percent of total applications that took more than the reported number of days to process. SOURCE: Quick, Finan survey of U.S.-based companies.
OPERATION AND EFFECTS OFU.S.EXPORT LICENSING 245 TABLE C-11 Profile of Actions on Reexport Applications Year Approved Denied RWAd Cancelled Total Value ($ bit) FY83 7,041 111 688 8 7,848 2.5 FY84 9,699 148 1,575 8 11,430 3.6 FY85 12,055 164 1,858 11 14,088 7.4 Reexport Applications in FY1985 by Destination Denied % Total To CoCom 11 0.08 To non-CoComa 156 1.11 Reexport Applications in FY1985 by Source and Destination From To Denied % Total RWAd % Total 366 2.60 1,512 10.73 RWAd % Total CoCom CoCom 8 0.06 271 1.92 Non-CoCom CoCom 3 0.02 95 0.67 CoCom Non-CoCom 61 0.43 576 4.09 Non-CoCom Non-CoCom 95 0.67 936 6.64 aIncludes bloc destinations. SOURCE: Department of Commerce export license data. Because U.S. parent companies' shipments to their foreign affiliates are less than one-half of their exports, it is reasonable to conclude that non-U.S. firms based in other CoCom countries frequently ignore or are unaware of U.S. reexport authorization requirements. This is not surpris- ing in view of the allies' hostility to U.S. extraterritorial controls. A few non-CoCom European governments have agreed to require their export- ers to show evidence of compliance with the rules of the country of origin when reselling foreign-controlled goods or components. EFFECTS OF CONTROLS ON BUSINESS Administrative Costs The Quick, Finan and Merrill surveys requested detailed information on the administrative costs firms incur in complying with national security export controls, taking into account all personnel time, overhead, and expenses devoted to preparing and filing applications, training corporate personnel in required procedures, hiring outside consultants and legal assistance, and recordkeeping, reporting, and auditing. The accuracy of the responses depends on how such expenses are allocated in corporate accounts. In most companies, export administration is not a budget line item.
246 APPENDIX C TABLE C-12 Distribution of a Sample of Reexport Applications by Source Percent of applications from CoCom countries Percent of applications from non-CoCom countries Percent of value from CoCom countries Percent of value from non-CoCom countries Percent of value from U.S. firms/affiliates Percent of value from foreign firms/affiliates Percent of firms exporting to CoCom countries Percent of firms exporting to non-CoCom countries 81 19 93.3 6.7 89.1 10.9 41.2 58.7 Reexport Applications from Selected Countriesa by National Ownership of Originating Firms CoCom Countries Percent U.S. firms Percent foreign firms A 98.2 1.6 B 91.5 8.5 C 86.6 13.4 Non-CoCom Countriesb D 66.8 33.2 E 61.3 38.7 F 18.3 81.7 aNames of countries are not identified at the request of the Department of Commerce. bCountries with which the United States has bilateral control arrangements. SOURCE: Quick, Finan analysis of sample from Commerce Department export license data. With these caveats, direct compliance costs do not appear to be a significant burden for most exporters. Based on the data obtained from the 170 respondents to the Quick, Finan survey, it is estimated that U.S. firms are currently spending approximately $500 million on export admin- istration. A small share of this amount is for outside service providers. Current expenditures nevertheless represent a sharp increase over pre-1985 expenditures, largely as a result of May 1985 regulations requiring distribution license holders and their foreign consignees to ensure against the diversion of controlled items to the Soviet bloc by establishing internal control and recordkeeping systems subject to onsite inspection by agents of the license holder and the U.S. government. Respondents to the Merrill survey reported that distribution license compliance costs increased more than five times, to approximately $100 million, as a consequence of the change in regulations.* Currently, average *In the late 1970s the Commerce Department all but ceased to audit ongoing activities under distribution licenses and to enforce the conditions on their use. This may have encouraged a minimal compliance effort on the part of license holders.
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 247 expenditures are $21,000 by small companies, $76,000 by medium-sized companies, and $649,000 by large companies, although within each group there is extremely wide variation. Although it might be assumed that installing an internal control program is more expensive than maintaining it, that is not an assumption shared by many license holders. Three-fourths of the respondents expect their 1986-1987 compliance costs to be higher than current expenditures. The 107 license holders in the sample employ 747 people, 20 percent or more of whose time is devoted to distribution license compliance. Sixty percent of these employees are foreign based. Perceived Competitive Erects Overwhelmingly, survey respondents are of the view that U.S. national security export controls have a negative effect on their ability to compete in international markets. This is perceived to be a function of: (1) the greater complexity, coverage, and stringency of U.S. controls relative to those of other industrialized and newly industrializing countries; and (2) the increasing availability outside the United States of competitive products and services subject to fewer or no restrictions. More than one-half of respondents reported that they have lost Free World as well as bloc sales primarily as a consequence of controls. Forty percent have had existing customers express a preference for or an intention to shift to non-U.S. sources of supply to avoid entanglement in U.S. controls. One-fourth indicated that the system is causing an erosion of the international distribution and marketing networks of U.S. companies with a consequent loss of business. A majority of the respondents expects these problems to become more severe in the next few years. Two unrelated recent changes in the Export Administration Regula- tions, imposing new controls or relaxing existing controls, permit tests of these claims and partial estimates of actual competitive costs. The Case of Analytic Instruments Normally, it is extremely difficult to quantify the relationship between changes in export regulations and changes in export sales. The case of analytic instruments, however, provides a unique opportunity to isolate the effects of discrete regulatory changes on a particular category of exports. As far back as 1979, there was industry concern about U.S. unilateral export controls on products that, although not militarily sensitive them- selves, contained embedded microprocessors with potential military applications. In April 1984 the Commerce Department announced decon- trol of roughly half of the categories of unilaterally controlled instru
r 248 APPENDIX C meets. Following completion of the multilateral review of the CoCom International List late in 1984, however, the department issued new regulations that had the effect of reinstituting validated license require- ments for most of the same instruments.* To test the effects of these changes on instruments trade, ECCN 4529B was cross-referenced to commodity 711 (excluding 711.8001) on Schedule E, U.S. Exports, and export data were obtained for the period 1978 through 1985. The analysis was limited to exports to CoCom countries, excluding Canada, for which validated licenses are not required. Ex- change rates, level of foreign industrial production, and changes in price levels were accounted for along with two terms to capture the ejects of changes in the regulations.! The results indicate that the regulatory changes had a statistically significant effect on exports. When controls were relaxed early in 1984, U.S. analytic instrument exports increased (by the third quarter of the *Until 1984, analytic instruments containing microprocessors were covered by ECCN 4529B (the letter indicating a unilateral U.S. control). Recontrol resulted from changes in the International List specifications for computing equipment, changes that were subsequently incorporated in ECCN 1565A on the U.S. Control List. Many industry observers believe that implementation of the CoCom agreement has not been uniform, but no thorough effort has been made to determine which products are being controlled multilaterally versus unilaterally. tThe following equation was specified: RAI = a + TRIP + b2*XR + binds + b4*D2 The variables of the regression are defined to be: RAI = value of real U.S. exports of analytic instruments to CoCom (less Canada) IP = weighted, aggregated industrial production indexes for CoCom countries (excluding the United States and Canada) XR = four-quarter moving average of weighted exchange rates for CoCom countries (excluding the United States and Canada) Do = time dummy for 1984.2 to 1985.1 (represents loosening of export controls) values: 84.2 = .25, 84.3 = 1.0, 84.4 = .5, 85.1 = .25 D2 = time dummy for 1985.2 to 1985.3 (represents tightening of export controls) values: 85.2 = .5, 85.3 = 1.0 where: coefficient (t-statistic) a = -20.80 bt= 0.92 b2=- 0.27 b3= 3.34 _ b4=- 4.93 R = 0.87 F= 42.20 Time period = 1978.2 to 1985.3 Significant at the 0.10 level. Significant at the 0.05 level. (- 1.41)a (- 12.01)b ( 6.39)b ( 1.33)a (- 1.82)b
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 249 year) roughly 7 percent over what they would have been without the change. When the relaxation was reversed late in 1984, exports (by the third quarter of 1985) were 12 percent below what they would have been if licensing requirements had not been reimposed. These fluctuations in trade reflect only the short-run observable effects probably attributable to export controls. In the long term the regulatory changes may erode demand for U.S. products. Also not reflected in the analysis are the effects these restrictions may have had on foreign transactions in similar instrumentation produced abroad with U.S. tech- nology or containing U.S. components. The Case of Foreign Consignees Under Distribution Licenses Under the Export Administration Regulations (section 373.3), as re- vised in May 1985, foreign consignees under U.S. distribution licenses must for the first time also establish internal control programs. The required features vary with the nature of consignees' activities (e.g., end use or reselling) but in general parallel those of the license holders' internal controls screening transactions against the U.S. Table of Denial Orders, a diversion risk profile, and criteria of "sensitive" nuclear uses; training personnel; and maintaining records subject to audit by the license holder and the U.S. government. The combination of increased financial costs, foreign sensitivities to the extraterritorial application of U.S. law, and, in the case of firms located in other CoCom countries, the duplica- tion of effort entailed in complying with domestic as well as U.S. controls raises a concern that the new requirements discourage companies from doing business with U.S. suppliers. Surveyed in May 1986, only 1 month after the regulations became fully effective, U.S. license holders responding (accounting for about 18 percent of the total number of licenses) reported the loss or removal of 32 percent of all their consignees 1,175 out of 3,686 in the sample-in the 12 months since the regulations were issued. Business changes unrelated to the regulations, inactive sales, and product decontrol actions were reported to account for one-half of the drop-outs for which respondents gave explanations; but the expense of compliance and consignees' refusal to comply accounted for 40 percent of the cases. (See Table C-13.) As expected, among the drop-outs, independent foreign firms far outnumbered affiliates of U.S. license holders by ratios of 11 to 1 among small firms, 17 to 1 among medium-sized firms, and 16 to 1 among large companies. Furthermore, almost all of the independent former foreign consignees were engaged in either reselling U.S. products in the form received or in selling foreign-made products with attached or incorpo- rated U.S. components.
250 APPENDIX C TABLE C-13 Reasons for Loss or Removal of Foreign Consignees from Distribution Licenses from April 1985 to May 1986 No. of No. of Firms Citing Consignees 1. Directly related to new regulations Expense/burden of compliance given volume of business Consignees declined to assume responsibility Consignees refused to comply License holder could not rely on consignees to comply Country no longer eligible Consignees failed to certify Consignees switched to non- U.S. sources 2. Indirectly related to regulations Consignees not active customers Lack of business Consolidation of licenses Included in affiliate reexport temtory 3. Other reasons Business change without regard to regulatory changes Products decontrolled (now GDEST) 10 8 10 2 2 25 13 s 2 13 2 202 27 21 1 2 138 52 20 43 144 3 SOURCE: Memll survey of distribution license holders. More often than not, business is continuing with former consignees under different licensing arrangements. To simplify compliance, some license holders have consolidated consignees under fewer licenses. Others are using an affiliated consignee to serve independent former consignees, although without the reexport authority the latter previ- ously enjoyed. Finally, as a direct consequence of the regulatory changes, 65 percent of respondents expect to apply for 67 percent more individual licenses and reexport authorizations than they used in 1985. Despite these adjustments, which in most cases entail additional administrative costs and uncertainty, 28 licensees (25 percent of res- pondents) reported that the loss of 164 consignees has meant an immediate loss, albeit small, of business for the foreseeable future. They estimated this loss, over a 3-year period, at $78.6 million, confirming that these consignees, although active customers, were low-volume ones. Extrapolating to all license holders, the total loss is in the range of $450 million. Most license holders consider this an acceptable price to pay to retain their distribution licenses. Nevertheless, the favorable benefit-cost mar- gin has narrowed considerably for smaller companies that are apprehen
OPERA TION AND EFFECTS OF U. S. . EXPOR T LICENSING 25 1 sive about being held accountable for the conduct of foreign customers with whom they have little leverage. Moreover, companies of all sizes reported that it is becoming more difficult to recruit new consignees; some consignees have reduced their orders even though remaining on a license; and in general the United States may be imposing too many restrictions to retain and expand its foreign customer base. These concerns are reinforced by perceptions of widespread foreign availability of products eligible for distribution licenses. Reporting on 114 distribution licenses, respondents claimed in 91 cases that comparable products are available from non-U.S. sources. Major CoCom partners (Japan, Germany, France) have bulk export authorizations for West-West trade, but only the French license has restrictions and procedures comparable to those of the U.S. distribution license. CONCLUSIONS Coverage of the Licensing System National security export controls reach a major portion of U.S. international business activity. In 1985 two-fifths of all U.S. exports of manufactured goods, excluding military equipment, received some form of prior government screening and approval to prevent Soviet acquisition of items of military value. In addition, controls applied to a large share of U.S. affiliates' international sales and to foreign companies' resales of U.S.-origin products and original sales of foreign products incorporat- ing U.S. components or technology. With ' the exception of pharmaceuticals and many chemicals, export controls affect most of the high-technology sector, in which U.S. produc- ers have long enjoyed a strong comparative advantage but are now vulnerable to foreign competition. The $62 billion in 1985 exports under national security controls compares to $68.5 billion in total high-tech- nology exports. It is estimated that the United States registered a deficit of $2 to $3 billion in high-technology goods in 1986, the first such deficit since the category was identified. Only 7 years ago, high-technology exports exceeded imports by $27 billion. Ninety-six percent of licensed exports are to Western countries; roughly half go to the NATO allies and Japan. According to U.S. government and CoCom criteria, many controlled items are of less than critical military value. One-third of licensed transactions involve products that under CoCom rules may be sold to the Soviet Union without multilateral review and approval. Two-thirds involve products that may be sold to the People's Republic of China without CoCom concurrence.
252 APPENDIX C Administration of Controls Efforts to improve the efficiency of the U.S. control system have focused primarily on reducing the average time individual license appli- cations are under review by one or more agencies of the U.S. government. This objective does not address several problematic characteristics of license processing that emerge from analysis of survey and government data: · The formal license review process occupies as little as one-half of the average time from submission of a license application to receipt of an export authorization or notification of other action. · A small but not insignificant number of cases extend beyond 100 days or nearly four times the average processing time. · Small firms experience longer processing times and more uncertainty about licensing outcomes than do large firms. · Processing times, at least for applications to destinations other than the Eastern bloc, do not vary consistently with the degree of military sensitivity associated with different levels of technology, as judged by the United States and its CoCom partners. The treatment of cases involving the least sensitive (i.e., AEN) technology suggests that the controls at this level may be largely a paper exercise for license applications to the Eastern bloc as well as for those to Western countries. A sample of processed applications contained no AEN cases that were either denied or returned without action. On the other hand, for reasons that are unclear but merit further investigation, exporters continued to submit applications for AEN-level items to CoCom countries several months after they became eligible for a general license and no longer needed approval. It is doubtful, too, that the requirement that foreign companies seek U. S. approval to reexport certain U.S.-origin products and incorporated components is an effective or enforceable instrument of control, at least within CoCom. Cost of Controls The amount of trade affected by export controls is so large that even a marginal negative competitive effect is likely to have significant economic consequences. Additional high-technology exports of $3 billion would be sufficient to convert the current U.S. deficit in that category into a trade surplus. Simply by virtue of the geographical distribution of U.S. exports, the costs of export controls fall primarily on West-West rather than on West-East trade. The aggregate economic costs of controls are exceedingly hard to determine. But for one category of products analytic instruments from
OPERATION AND EFFECTS OF U.S. EXPORT LICENSING 253 which validated licensing requirements were first removed but on which they were subsequently reimposed, it has been possible to estimate the short-run trade effects of regulatory actions, independent of changes in exchange rates, production, and prices. Decontrol had a positive effect on U.S. exports to CoCom countries other than Canada of about 7 percent; recontrol reduced exports to those countries by about 12 percent. Similarly, the recent imposition of new accounting and auditing require- ments on foreign customers that receive controlled goods under U.S. distribution licenses has already caused some erosion of the distribution networks of U.S. exporters and a small loss of business.
