Conflict and Cooperation in National Competition for High-Technology Industry (1996)

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.

206 CONFLICT AND COOPERATION Approaches to Conflict Resolution: Agenda for Action SUBSIDIES AND PUBLIC PROCUREMENT Toward an Open Subsidy Club36 Subsidies are one of the most important instruments for the promotion of high-technology industries. The share of public funding in private R&D expenditures varies between 11 and 33 percent in leading OECD countries, with the remarkable exception of Japan, where public subsidies cover only 1.5 percent of private R&D outlays (Bletschacher & Klodt, 1992). What is more, R&D subsidies are directed toward very few high-technology indus- tries—above all to the electronics and aerospace industries—and often go along with investment and production subsidies, as the Airbus case demon- strates.37 In the course of the Uruguay Round negotiations, the GATT signatories have made a new attempt to restrict the use of subsidies for industrial policy purposes. The Agreement on Subsidies and Countervailing Duties intro- duced a so-called traffic light approach that divides subsidies into three categories: (1) prohibited; (2) actionable (i.e., these subsidies can be countervailed); and (3) nonactionable. The “red light” group includes ex- port subsidies and subsidies that can be roughly characterized as import substitution subsidies.38 These subsidies are actionable in any case, regard- less of whether they are specific or not. The “green light” (nonactionable) category covers nonspecific subsidies, R&D subsidies, regional subsidies, and envi- ronmental subsidies. A subsidy is considered specific if it is granted to only a limited number of well-defined industries or enterprises. Although most R&D subsidies are designed to promote specific industries, R&D subsidies are non-actionable as long as public funds cover not more than 75 percent of the costs of industrial research and not more than 50 percent of precompetitive development research. To benefit from non-actionability, “green light” subsidies 36 This chapter is based on J. Stehn, Subsidies, Countervailing Duties, and the WTO: Towards an Open Subsidy Club, Kiel Discussion Paper, Kiel Institute of World Economics, Kiel, June 1996. 37 It has been calculated that almost one-third of all subsidies to Airbus Industries can be characterized as production subsidies. See G. Bletschacher, and H. Klodt, Strategische Handels- und Industriepolitik, Kiel Studies 244, Tübingen, 1992, pp. 64–66. 38 “... subsidies contingent, whether solely or as one of several other conditions, upon the use of domestic over imported goods.” (WTO Subsidies Agreement, Article 3.1.) 

APPROACHES TO CONFLICT RESOLUTION 207 are to be notified to and approved by the WTO Committee on Subsidies and Countervailing Measures. All subsidies that fall into neither the “red light” nor into the “green light” category are defined as actionable (“yellow light” category).39 Although the WTO Subsidies Agreement clarified some of the issues that caused frequent disputes (Schott, 1995),40 it suffers from at least two major weaknesses that might give rise to further conflicts in international trade, especially with a view to high-technology competition. First, countervailing duties imposed by third countries are the only enforcement mechanism of both the WTO Subsidies Agreement and the Tokyo Round Subsidy Code. To initiate a countervailing duty investigation a signatory has to prove that subsidies of an offended trading partner violate existing WTO regulations and that the domestic industry is “experiencing injury” as a result of these subsidies. Thus, industries and enterprises suffering from subsidies abroad are the real supervisors of the WTO subsidy regulations. This “decentral- ized” supervision system has proved to be rather ineffective on several grounds:41 • Many kinds of subsidies, especially investment subsidies and R&D subsidies, do not cause measurable competition distortion effects for several years. By the time the effects become obvious, the relevant subsi- dies have often been phased out, and it may become almost impossible to prove that the subsidies caused material injury. • The current subsidy regulations cannot prevent governments from granting subsidies that distort international competition because a proof of material injury is a first and indispensable step in a countervailing duty investiga- tion. Hence, countervailing duties can mitigate the competition distortions only after they have occurred. • The countervailing duties approach poses the danger of retaliation, especially in high-technology industries, where the time lag between subsi- dies and the resulting competition distortion effects is usually quite long. 39 According to the WTO Subsidies Agreement, both “yellow-light” and “red-light” subsi- dies are actionable. The main difference between these two groups is that “red-light” subsidies can—as a general rule—be countervailed by a foreign government without a proof of material injury. 40 J.J. Schott, Dispute Settlement in the Multilateral Trading System and High Technology Trade. Paper presented at the conference “Towards a New Global Framework for High- Technology Competition,” Kiel Institute of World Economics, Kiel, Germany, 30–31 August 1995. 41 See L.D. Tyson, Who’s Bashing Whom? Trade Conflict in High-Technology Industries, Washington, D.C., 1992, pp. 280–286, for a detailed evaluation of the Tokyo Round subsidy regulations. 

208 CONFLICT AND COOPERATION Under these conditions, the imposition of a countervailing duty by a trading partner might be judged arbitrary and unfair by the offended country. • Countervailing measures may be abused for protectionist purposes and— because of their price effects—may further distort international competition. In view of these shortcomings, the WTO Subsidies Agreement should be reformed (1) by introducing a notification system aiming at an assessment of all planned subsidies prior to their implementation and (2) by defining quantitative thresholds for the provision of subsidies. (1) With respect to the implementation of a multilateral notification system, the aid supervision procedure of the European Union can serve as a reference system.42 A multilateral subsidy supervision should stipulate that all plans to grant new or to alter existing subsidies are to be notified to and approved by the WTO Committee on Subsidies and Countervailing Measures (CSCM).43 The CSCM should be entitled to examine the notified plans and to decide whether they are compatible with the WTO Subsidies Agreement. In the course of the investigation, the CSCM should take into account written comments by third signatories that might be affected by the notified subsidy. After the CSCM has made its decision, any signatory concerned should have the opportunity to initiate a panel procedure against the CSCM ruling in accordance with the WTO Dispute Settlement Mecha- nism. Given that a signatory grants a subsidy in violation of a final CSCM or WTO panel ruling, the CSCM should be empowered to require a repay- ment of the subsidy to the respective national government. The potential threat of a repayment may give an incentive to recipient firms and indus- tries to ask their respective national governments to present an approval of the CSCM before granting the subsidy, and may, therefore, lead to some sort of self-restraint. Only if a signatory does not react to any of the CSCM or WTO panel rulings within an appropriate period (e.g., two months after the final decision), third parties would be entitled—as a last resort—to initiate a countervailing duty procedure according to the regulations of the current WTO Subsidies Agreement. To facilitate a multilateral aid supervision, the current traffic-light ap- proach should be reformed by categorizing all subsidies as either prohibited subsidies or subsidies that are allowed under certain conditions. This can easily be done by just skipping the “yellow light” category. Although this category is defined by default, it is obvious from the definitions of the “red light” and “green light” subsidies that it covers all specific subsidies not explicitly mentioned in the “green light” group. As a consequence, all 42 See J. Stehn, “Wettbewerbsverfälschungen im Binnenmarkt: Ungelöste Probleme nach Maastricht,” Die Weltwirtschaft (1993), pp. 43–60, for an evaluation of the European aid supervision system. 43 Currently, only “green light” subsidies have to be notified to the CSCM. 

APPROACHES TO CONFLICT RESOLUTION 209 non-specific subsidies except for the current “red light” group, i.e., export subsidies and import substitution subsidies, would be admissible. Thus, a workable and clear-cut definition of specificity is an important prerequisite for an effective subsidy supervision. (2) The WTO Subsidies Agreement fails to provide unequivocal guide- lines for measuring specificity. According to the agreement (Art. 2.1(c)), the following factors should be considered in determining whether a sub- sidy program is specific or not: • use by a certain number of enterprises; • predominant use by certain enterprises; • the grant of disproportionately large amounts of subsidy to certain enterprises; and • the manner in which discretion is exercised by administering authorities. Besides being obviously rather vague and ambiguous and thus giving rise to disputes in interpretation, these guidelines do not take sufficient account of the economic effects resulting from firm-specific or industry-specific subsidies. From an economic point of view, the main objective of specific- ity rules is to limit the competitive distortions due to subsidies that are mainly directed to a single industry or enterprise. It can be realistically assumed that the extent of competition distortions depends on the share of production, investment, or R&D costs that is covered by public funding. Hence, the current rules could be considerably improved by defining quan- titative thresholds that limit the provision of subsidies to a certain frac- tion—say, 5 percent—of the respective subsidy base. With a view to R&D subsidies, it is certainly by far too optimistic to expect that governments will agree on a 5 percent threshold in the near future. The “safe harbour provision” (Schott, 1995)44 for R&D subsidies which was agreed upon in the Uruguay Round negotiations rather points in the opposite direction. It is obvious that the current maximum, i.e., a public funding of up to 75 percent of R&D costs, will give a recipient firm a considerable competitive edge and thus might lead to major conflicts in high-technology trade and competition. Under these conditions, the level- ling-down of current thresholds seems to be a necessary and highly desir- able policy goal. As empirical research indicates, high-technology R&D can be expected to generate cross-border externalities.45 Thus, new technological knowl- edge can no longer be assumed to be totally exclusive. If national govern- 44 J.J. Schott, Dispute Settlement in the Multilateral Trading System and High Technology Trade. Paper presented at the conference “Towards a New Global Framework for High- Technology Competition,” Kiel Institute of World Economics, Kiel, Germany, 30–31 August 1995. 45 See above (Analytical Benchmarks). 

210 CONFLICT AND COOPERATION ments were to take these international spillovers of domestic innovations into account, they would have strong incentives to cooperate internation- ally.46 An efficient way of international cooperation might be the mutual opening up of national subsidy funds for high-technology development. This approach would allow firms located in third countries open access to na- tional subsidy funds if—and only if—third countries are ready to offer open access to their subsidy budgets on a reciprocal basis. To provide an incentive to open up national subsidy funds, one could also consider a provision that, as a general rule, lowers the threshold for high-technology subsidies to 30 and 15 percent of the costs of basic and applied R&D, respectively, but allows higher public funding—up to the current limits—if a national research program provides open access for firms located in third countries. For practical purpose, this approach may require the implementation of the following general rules: • An R&D subsidy programme should be regarded as open only if at least two foreign firms are participating in the program; • the objectives of the open fund are formulated by the respective na- tional government and must be met by both domestic and foreign firms; • domestic and foreign firms should be treated equally with respect to patents and copyrights that emerge from the funded research; and • open access to national funds should not be linked to cooperation with a domestic firm. To facilitate a multilateral subsidy supervision and international coopera- tion in publicly funded R&D, the WTO should publish annual reports on recent developments in national subsidy schemes of leading OECD coun- tries. Multilateral monitoring of current subsidization practices could also help to prevent international conflicts arising from a misinterpretation of the objectives and potential effects of national subsidy programs (Ostry, 1995).47 Public Procurement: Conflict Resolution by National Courts Public procurement in advanced industrial economies covers a signifi- cant part of overall market demand. Non-defense public procurement in the member states of the European Union is estimated to represent about 7 to 46 In an extreme case, this could lead to a free-rider problem in promoting private R&D. However, with a view to the fact that the number of players in the high-technology game is rather small, this seems to be very unlikely. 47 S. Ostry, Technology Issues in the International Trading System, September, 1995, unpublished paper. 

APPROACHES TO CONFLICT RESOLUTION 211 10 percent of gross domestic product.48 It is therefore tempting for govern- ments to use their demand for goods and services to achieve aims of tech- nology and industrial policy. The GATT procurement code of the Tokyo Round (in force since 1981) established for the first time internationally binding rules that secured for- eign competitors open and undiscriminatory access to bidding procedures for public procurement of goods (though not yet of services).49 It is based on the conditioned most-favored-nation clause, i.e., it applies only to the relationships between those contracting parties of GATT that have actually signed the code. The Tokyo Code, at that time widely regarded as one of the most far- reaching achievements of the Tokyo Round Agreements,50 committed only central governments and directly related entities to internationally open ten- ders (i.e., those surpassing certain threshold levels of tender value). The rather limited range of the Tokyo Code—covering less than 10 percent of nondefense public procurement in the United States and the European Com- munity51—clearly constrained its impact from the outset and goes a long way toward explaining its actually very limited economic consequences. As a part of the Uruguay Round accord, the new Government Procurement Agreement (GPA)—in force from 1 January 1996—extends the reach of the Tokyo code to potentially all kinds of nondefense procurement, i.e., to all nondefense goods and services. The GPA is one of the three major agreements of the Uruguay Round that are not included in the so-called Single-Undertaking procedure. This means that these agreements apply only to their signatories, not to all con- tracting parties of the World Trade Organization at large.52 The potential 48 See Patrick A. Messerlin, “Agreement on Public Procurement,” in OECD, The New World Trading System: Reading, OECD Documents, OECD, Paris, 1994, pp. 65–71. Figures of this kind of course depend on the underlying definition of what is considered public sector procurement. This concerns in particular the procurement policy of government-owned firms and the procurement policy of firms operating in markets that are heavily regulated by govern- ment (such as utilities). 49 Government procurement had been previously exempted from GATT regulations (Art. III (8) GATT). 50 See Robert Stern, and Bernard M. Hoekman, “The Codes Approach,” in J. Michael Finger and Andrzej Olechowski (eds.), The Uruguay Round: A Handbook on the Multilateral Trade Negotiations, The World Bank, Washington, D.C., 1987. 51 See Patrick A. Messerlin, “Agreement on Public Procurement,” OECD, The New World Trading System: Reading, OECD Documents, OECD, Paris, 1994, pp. 65–71. 52 The other two agreements are the Arrangement Regarding Bovine Meat and the Interna- tional Dairy Arrangement. A fourth plurilateral agreement under the umbrella of the WTO, the agreement on Trade in Civil Aircraft, was not changed at the end of the Uruguay Round and remained open to signature only in its already existing form. Parties of the GPA are at present Canada, the fifteen member states of the European Union, Norway, Switzerland, Japan, the United States, Israel, and South Korea. Singapore was a signatory of the Tokyo Code, but opted out this time. In turn, South Korea entered the club as a new member. 

212 CONFLICT AND COOPERATION value of contracts that will be covered by the GPA in signatory states is roughly estimated to be about US $ 400 billion annually in current prices.53 There are two main areas in which the GPA breaks new ground by the standard of its Tokyo Code predecessor. These concern (1) extensions of the coverage provided by the GPA and (2) disciplines imposed by the GPA on signatories. (1) Coverage of the GPA: The principle of exchange of “trade conces- sions” that has governed all multilateral trade negotiations under the aus- pices of GATT means in the context of GPA that contracting parties con- cede “items” (i.e., certain groups of goods and services) to be opened to foreign competition in public tenders, and “entities” (public or semi-public bodies) designated to open their tenders for bids by foreign suppliers. The new GPA extends the range with respect to both “entities” and “items” concerned. Under the new agreement (Article I), all tenders regard- ing rentals or leases of goods, as well as tenders regarding the procurement of services, shall be included. These general extensions are, however, limited by exemptions listed in the Annex of the Agreement for each signatory state.54 Limits concerning goods are listed as exemptions; limits concerning services apply to all items that are not explicitly enumerated. This distinction is in accordance with the different philosophies of regulation in the GATT (for goods) and the GATS (for services). The institutional solutions chosen in service sectors are likely to favor bilateral deals of sectoral reprocity, thus undermining attempts to arrive at a multilateral framework. A similar tendency prevails in the so-called exempted areas of public procurement of goods, notably telecommunications equipment. As far as public or semi-public “entities” are concerned, the new GPA has in principle been extended to “sub-central” entities, i.e., basically regional and local government entities. However, those sub-central entities that are in fact obliged to open their tenders to foreign competition are enumerated in the annex to the GPA.55 There are two major problems involved in the definition of “public enti- ties.” First, the power of the contracting party—the central government—to 53 See Jeffrey Schott (assisted by Johanna W. Buurman), The Uruguay Round: An Assess- ment. Institute for International Economics, Washington, D.C., November. 1994. OECD, Trade and Competition Policies: Comparing Objectives and Methods, Trade Policy Issues, 4, OECD, Paris, 1994. Sylvia Ostry and Richard R. Nelson, Techno-Nationalism and Techno-Globalism: Conflict and Cooperation, Brookings Institution, Washington, D.C., 1994. 54 Note that the public procurement regulation of the European Union, which in many instances seems to have served as an example for the WTO procurement regulation, does not contain exemptions by “public entities” or “items” except the general exemption of defense procurement. 55 For instance, in the case of the United States, many of the obligations under this treaty are limited to a rather small number of states. 

APPROACHES TO CONFLICT RESOLUTION 213 control the conduct of sub-central entities may be rather limited, e.g., in the case of federal member states. Second, previously “public firms,” once included in international arrangements on public procurement, may become privatized, thus leaving other contracting parties with a loss of “trade con- cessions” formerly granted to them. (2) Disciplines Imposed by the GPA:56 According to the two basic principles embedded in the GPA, suppliers from other signatory states should benefit from the conditional MFN clause, and thus tendering procedures should not entail any discrimination between domestic suppliers and suppliers from other signatory states. In turn, discrimination against suppliers from third countries (i.e., nonmembers of the GPA) is still allowed, which is consistent with the philosophy of the WTO as an open club. A core achievement of the new GPA is that foreign suppliers discrimi- nated against in a national tendering process can use the so-called challenge procedure, i.e., they can submit their appeal directly to a ruling by the courts of the country that issued the respective tender, and these courts are then obliged to provide reasonably rapid proceedings.57 Moreover, “offsets” (i.e., deviations from the GPA) are explicitly prohib- ited by Art. XVI, meaning that additional requirements attached to a bid on a national public tender (e.g., local content, countertrade, and the like) are to be considered illegal. Furthermore, the GPA does not contain any safe- guard clause that could allow signatory states to refrain from fulfilling, or to circumvent, their obligations under this agreement. On the other hand, the GPA does not contain any provision against collusion among bidding domestic firms, which is left to the competence of domestic competition policy. The regulations of the GPA, which by themselves appear to be quite strict,58 may yet be undermined by regulations of the TRIPs Agreement 56 Cf. Bernard M. Hoekman and Petros C. Mavroidis, “The WTO’s Agreement on Govern- ment Procurement: Expanding Disciplines, Declining Membership?” Discussion Paper 1112, CEPR, London, 1995. 57 In addition, the WTO dispute settlement procedures are open to such cases and can also be used. There are, however, significant differences between these two routes of appeal against discriminatory treatment by particular national “public entities.” To be successful in a WTO dispute settlement procedure, the claimant (the government of the affected firm) must provide evidence that its previously granted trade concessions were impeded in the case in question. 58 Here are selected articles of the GPA with some relevance to the issue of maintaining club discipline: Art. VI defines rules for technical specifications of the items included in a tender; Art. VII regulates the choice among possible tendering procedures; Art. VIII is con- cerned with qualification requirements for tenderers; the invitation procedures to participate in a tender are circumscribed in Art. IX; the selection procedure is outlined in Art. X; prescrip- tions on time schedules are given in Art. XI and on necessary documentations in Art. XII; submission procedures are regulated in Art. XIII; Arts. XIV through XVIII are primarily 

214 CONFLICT AND COOPERATION (Agreement on Trade-Related Intellectual Property) of the Uruguay Round. This means in particular that a national government will still be allowed to specify the conditions of a tender, if “there is no sufficiently precise or intelligible way of describing the procurement requirements and provided that words such as “or equivalent” are included in the tender documenta- tion” (Article VI.3 GPA).59 This provision may give national governments enough leeway to limit international competition whenever they think it to be appropriate. In practice, foreign firms affected by discrimination in public tendering procedures will weigh the possible advantages of a court ruling in their favor against the disadvantages of possibly foregoing the good will of the government concerned. In the longer term, however, the opportunities es- tablished for foreign suppliers’ access to domestic court procedures should work in the direction that national governments will increasingly behave more strictly according to the rules laid down in the GPA. Although clearly the GPA is a major step in the right direction, major problems remain unsolved: • At present, WTO regulations still allow legal subsidies to R&D activi- ties and subsidies on regional policy grounds. This opens the gate for abuse in national procurement: e.g., a national government may grant R&D subsi- dies and then define the terms of a later procurement tender on the basis of those specific R&D requirements that only a domestic firm can possibly meet. Bidders from the outside would obviously have little chance to suc- ceed in such a procurement tender. • Decisions of governments on matters of national security are exempted from WTO regulations. Obviously, clauses of this kind can also be mis- used, in particular with respect to the treatment of dual-use goods and ser- vices (i.e., goods and services that can be used for both military (security) and civilian purposes). • The many national exemptions in the GPA with respect to public entities and items (or sectors) are likely to encourage bilateral bargaining, which is obviously not in the spirit of an international and multilateral framework of trade regulations. • The question of what has to be considered a public enterprise or an enterprise under significant influence of government, can hardly be an- concerned with other technicalities of the tendering process; and finally, Art. XIX requires that the parties concerned collect annual statistics and provide these statistics to the Committee on Government Procurement at the WTO—which by the way can be considered the “nucleus” of a future supervision board in this area, the creation of which is considered necessary by many experts (see Sylvia Ostry, “Technology Issues in the International Trading System,” unpub- lished paper, OECD Trade Committee, September, 1995.) 59 See also Jean-Jacques Laffont and Jean Tirole, “Auction Design and Favoritism,” International Journal of Industrial Organization, vol. 9, 1991, no. 1, pp. 9–42. 

APPROACHES TO CONFLICT RESOLUTION 215 swered on the basis of common criteria across countries. As this definition is critical for the assessment of trade concessions in international trade negotiations, there is a strong need for further international agreements on this matter. As the new GPA code is entering into force by 1 January 1996, it re- mains to be seen how effective this code will prove to be, and whether the problems enumerated above will in fact call into question the overall spirit of the whole venture. MARKET ACCESS AND STRUCTURAL IMPEDIMENTS In addition to subsidies and public procurement practices, international competition in high-technology goods is distorted by a variety of other barriers to market access, including “structural” impediments to trade and investment. Trade and investment are in fact two major channels for the international diffusion of technology. In exporting countries, the foreign trade outlet enables technology enterprises to exploit economies of scale in production and facilitates their recovery of R&D expenditures. In import- ing countries, user industries benefit from access to the most advanced technology-intensive capital goods, components, and services at competi- tive prices, often a precondition for their own international competitiveness. Consumer welfare, too, is enhanced by the supply of high-technology con- sumer goods and services, again at competitive prices. From this one might infer that governments should be interested in open markets for high-tech- nology goods and services. In practice, however, barriers to market access prevail. Some barriers, like tariffs, certain antidumping practices, voluntary export restraint agree- ments, and safeguards, are deliberate attempts to shield domestic producers against competition from technologically superior, or simply less expen- sive, foreign suppliers, often with the intention of providing them a respite for catching up. Others, like government regulatory measures and technical standards, are often the by-product of differences in national cultures, policy objectives, and technology practices which may be difficult to reconcile internationally. And still others, like barriers to access to private technol- ogy “clubs” or to national dealer networks, are the result of restrictive, though not necessarily illicit, business practices. All these practices are potential sources of economic and political friction. Tariffs and Non-Tariff Barriers Tariffs, traditionally the preferred instrument to restrict market access, have greatly lost importance in the course of repeated GATT rounds. The 

216 CONFLICT AND COOPERATION Uruguay Round resulted in a (further) lowering of import duties for indus- trial goods by 38 percent in developed countries, from 6.3 to 3.9 percent on average. Forty-three percent (up from 20 percent) of the total value of industrial products imported by developed countries are now entering duty- free, and only 5 percent (down from 7 percent) are subject to peak rates.60 Equally important as the reduction of tariffs is that developed economies have now accepted a binding of virtually all their tariff lines on industrial products, thereby greatly improving the security of market access.61 Tech- nology products, among them electrical and nonelectrical machinery, chemicals, and pharmaceuticals, were also subject to the elimination or significant reduction of tariffs; on important electronics items, such as semiconductors, semiconductor manufacturing equipment and computer parts, tariff cuts of 50 percent and above were achieved. Yet, for certain items, especially in the field of consumer electronics, substantial tariff barriers persist. Their reduction should be made part of a concerted multilateral effort to improve market access in which the European Union, the United States, and the major East Asian economies should take the lead. With MFN tariffs rendered increasingly obsolete as measures of protec- tion, at least in developed economies, non-tariff barriers to trade and market access have received growing attention. Among border measures, anti- dumping duties, countervailing duties, and safeguards are increasingly ap- plied in high-technology trade. They are treated in the following chapter. The much criticized Voluntary Export Restraints (VER), a frequently ap- plied substitute for these measures, shall be phased out as a result of the Uruguay Round. No less significant than border measures are the already mentioned structural impediments inside the borders: government regula- tory practices, technical standards, and restrictive business practices. They may be deliberately designed in such a way as to discriminate against im- ported products: luxury taxes on certain classes of automobiles, the U.S. corporate average fuel economy (CAFE) law,62 costly and time-consuming testing, certification, and conformity assessment procedures for technology- intensive products, and limitations to foreign ownership in business sectors deemed “strategic.” Discriminatory measures are an offense against Article III of the GATT, which establishes the principle of national treatment (non- discrimination) of imported products, a guiding principle of the multilateral trading system. Such measures are therefore challengeable under the GATT 60 Cf. GATT, News of the Uruguay Round, April 1994. 61 The percentage of tariff lines bound has been increased from 78 to 99 percent in developed economies, and from 22 to 72 percent in developing economies. Ibid. 62 Cf. G. Kleinfeld, “Taxation of Automobiles and GATT National Treatment Obligations: Where to Draw the Line?” World Competition, vol. 19 (1995) no. 1, pp. 77–90. 

APPROACHES TO CONFLICT RESOLUTION 217 dispute settlement mechanism. Often, however, structural impediments simply reflect differences in national value systems, historic institutional settings, and grown systems of business culture (like the systems of corporate gover- nance). While they may form effective barriers to foreign entry, the WTO recognizes the diversity of national business cultures, institutional settings, and policy approaches which are at the heart of comparative advantage63 and competition among which is indeed a condition for advancement. It is no wonder, then, that the Uruguay Round has made only limited progress in this difficult area. To be sure, the Agreement on Technical Barriers to Trade, concluded under the Tokyo Round, was revised and strengthened, and Article XXIII of GATT was clarified. What is still missing is, among others, an agreed catalogue of government practices (including omissions) considered to be (challengeable) structural impediments, as well as “objec- tive” criteria for determining whether and to what extent a market is indeed “contestable,”64 and if not, how this situation can be remedied. In the field of high technology, market presence through foreign direct investment is no less important than market access through trade. Domestic presence is often critical to success in following market trends, gaining access to the R&D infrastructure, marketing research-intensive products with short product cycles and high sunk costs, and responding to marketing needs. At the same time, foreign investment is an important channel for the inter- national diffusion of technology, including skills. In spite of these microeconomic and macroeconomic benefits, the notion of national treatment, while ac- cepted in principle, is not yet fully applied to investment. Governments seek to attract greenfield investment from foreign high- technology enterprises through a great variety of financial incentives and/or through trade pressure. At the same time that they restrict market presence in business sectors deemed “strategic,” they oppose the acquisition of do- mestic technology firms by foreign investors, or they restrain access to national technology programs. True, an opening of the respective sectors to foreign investment often goes hand in hand. E.g., in telecommunications, a high-technology service sector, deregulation in Europe has promoted for- eign entry as well as the formation of international strategic alliances with non-European partners. Still, in all countries of the triad, and even more in emerging and developing economies, obstacles and market distortions per- sist, as foreign investors are denied effective market presence. The Uruguay Round Agreement on Trade-Related Investment Measures (TRIMs) is but a first step to a multilateral investment agreement. The 63 Cf. S. Ostry, “Technology Issues in the International Trading System,” in OECD, Market Access after the Uruguay Round: Investment, Competition and Technology Perspec- tives, Paris, 1996, p. 159. 64 Ibid. 

218 CONFLICT AND COOPERATION agreement is narrow in its focus, being limited to the trade-related aspects of investment, and here in particular to local content and trade balancing requirements, which are expressly prohibited. This prohibition applies whether the measures are mandatory or are required in return for an incentive or advantage. The TRIMs Agreement had a poor start: under the U.S.-Japan Auto Agreement of 1995, the Japanese car manufacturers entered into a “voluntary” commitment to raise the U.S. content of their transplants in the United States and to meet NAFTA local content standards by 1998 “in order to boost U.S. and North American content in their vehicles.”65 This commitment, like the other commitments under the agreement, will be monitored by “a new, effective monitoring system jointly developed by the U.S. indus- try working closely with the U.S. government.”66 It remains to be seen whether the auto agreement will serve as a model for circumventing the TRIMs Agreement by “voluntary” commitments or whether it remains a one-time transgression. What is needed beyond the TRIMs Agreement is a General Agreement on Investment (GAI), the scope of which is not limited to trade-distorting measures. This agreement should guarantee the right of establishment and full national treatment,67 subject only to exceptions which are clearly and narrowly defined (e.g., national security) and open to judicial review.68 Among the issues to be addressed under a GAI should be: access to govern- ment technology-support programs; state sanctioned monopolies and other sector-specific reservations; most-favored nation (MFN) treatment; trans- parency; barriers to foreign takeovers; performance requirements; invest- ment incentives; restrictive business practices and competition policy; in- vestment protection; access to technology; payments and transfers; movement of key personnel and data; and a dispute settlement mechanism for resolv- ing conflicts not only between governments, but between governments and investors.69 65 USIS, “U.S. government fact sheet: U.S.-Japan Auto and Auto Parts Agreements,” U.S. Information and Texts, 30 June 1995, p. 11. 66 Ibid. 67 Cf. R.Z. Lawrence, “Towards Globally Contestable Markets,” in OECD, Market Access after the Uruguay Round: Investment, Competition and Technology Perspectives, Paris, 1996, p. 29. 68 Cf. E.M. Graham, “Investment and the New Multilateral Trade Context,” in OECD, Market Access after the Uruguay Round: Investment, Competition and Technology Perspec- tives, Paris, 1996, pp. 4f. The authority granted to the U.S. president under the Exon-Florio amendment to block the takeover of a U.S. enterprise by a foreign investor in the case of an “impairment of national security” is much too general, and it is not subject to court challenges. 69 Cf. Ibid., pp. 35–62; D.M. Price, “Investment Rules and High Technology: Towards a Multilateral Agreement on Investment,” in OECD, Market Access after the Uruguay Round: Investment, Competition and Technology Perspectives, Paris, 1996, pp. 171–186; D.L. Aaron, “After the GATT, U.S. Pushes Direct Investment,” The Wall Street Journal, 2 February 1995. 

APPROACHES TO CONFLICT RESOLUTION 219 Work for a Multilateral Agreement on Investment (MAI) is currently under way in the OECD. This takes account of the fact that most of the issues at stake are less controversial among industrialized countries than with emerging and developing economies. Nevertheless, the aim must be for an agreement under the WTO to complement the GATT and GATS Agreements and covering these countries as well, though initially a WTO agreement may be less far-reaching in scope and commitments than a MAI under the OECD. In the face of what has been perceived as “inadequate” access to markets due to structural impediments, effective market access has become the catchword for triggering (and/or rationalizing) unilateral trade measures aimed at forc- ing open foreign markets and securing targeted market shares for imported (technology) products, thus circumventing the laborious procedure of nego- tiating away specific obstacles. This was especially the U.S. approach toward Japan under (the threat of) Section 301 in the cases of semiconduc- tors, flat glass, telecom equipment, medical equipment and, recently, autos and auto parts. The European Community has as yet made little use of its New Commercial Policy Instrument (NCPI) introduced in 1984 or its suc- cessor, the Trade Barriers Regulation (TBR). Between 1984 and 1994 only four examination procedures were opened. In one case the EU obtained a favorable GATT decision; in the three other cases the disputed practices were discontinued without the EU taking retaliatory measures.70 With ac- cess of firms to the new TBR, “liberalized” relative to the old NCPI, “it is likely that the use of the instrument will be more frequent in the future.”71 In the face of effectively closed markets, with no clearly identifiable—and yet quite effective—barriers to entry, taking recourse to market-opening instruments such as Section 301 or the Trade Barriers Regulation appears at first sight an appropriate way of approaching the obstacles. The temptation of a unilateral approach is all the greater since experience in the semicon- ductor and automobile cases, and in cases which were resolved with less publicity under the threat of these instruments, may well lead to the conclu- sion that this is indeed a successful means of enforcing market access. Yet there are serious arguments against a “managed trade” approach, especially if applied unilaterally. First, unilateral action undermines the multilateral trading system which is an international public good. Sec- ondly, a unilateral approach lends itself only to countries or trading blocs with a high bargaining leverage, such as the United States and the European Union, with the rest of the world put at a clear disadvantage. It is therefore 70 Cf. H. Beekmann, “The 1994 Revised Commercial Policy Instrument of the European Union,” World Competition, vol. 19 (1995) no 1, pp. 59f. 71 Ibid., p. 75. 

220 CONFLICT AND COOPERATION liable to misuse as an instrument of power politics. Thirdly, because of that there is a strong presumption that any concession achieved will be at the expense of third parties. This is even true if the claimant pretends not to seek preferred treatment over other foreign supplier countries, since the defendant country may not be prepared to open its market for imports gen- erally and may well curtail existing market shares of third parties in favour of the economic superpower involved. Moreover, follow-up measures are discussed bilaterally without the participation of third parties.72 Fourthly, lack of market access is often due to one’s own insufficient marketing effort. Third-country suppliers which undertook such effort at high cost will find their investment depreciated by the easy access offered to their foreign competitors as a result of government pressure.73 Fifthly, more often than not quantitative import targets can only be met through active involvement and commitment of domestic suppliers in the country con- cerned. This encourages collusive agreements among these firms and fur- thers the cartelization of the respective market, a result which is counterpro- ductive to economic efficiency. With the dispute settlement procedure greatly improved as a result of the Uruguay Round, there is therefore a strong case for taking resort to multi- lateral rather than unilateral action. Article XXIII:1(b), while not yet tested in practice, may offer access to the GATT dispute settlement procedure in cases where a contracting party finds that “any benefit accruing to it di- rectly or indirectly under this Agreement is being nullified or impaired or that the attainment of any objective of the Agreement is being impeded as the result of the application by another contracting party of any measure, whether or not it conflicts with the provisions of this Agreement”74 (“non- violation clause”). The Articles of Agreement of GATT thus seem to offer a multilateral route to dispute settlement even in cases of “structural” im- pediments, including private restraints of trade.75 This route should be tested and, if found inadequate, be improved.76 72 G. Zeidler, “Distribution Systems and Vertical Integration: Protectionism or ‘Normal’ Business Practice?” Discussion paper prepared for the conference “Towards a New Global Framework for High-Technology Competition,” held in Kiel, Germany, on 30–31 August 1995. 73 For instance, U.S. car manufacturers “estimate” that under the U.S.-Japan Auto Agree- ment of 1995, Japan will increase its imports of Big Three U.S.-made vehicles from 45,000 in 1994 to 160,000 by 1998. This “estimate” is the basis for the monitoring of the agreement. A similar “goal” is the establishment of 200 new dealerships in Japan by 1996 and 1,000 by the year 2000 which is based on access to dealership networks of Japanese auto manufacturers. Cf. “U.S. Government Fact Sheet: U.S.-Japan Auto and Auto Parts Agreement,” U.S. Informa- tion and Texts, 30 June 1995, p. 11. 74 Italics added. 75 Cf. S. Ostry, “Technology Issues in the International Trading System,” in OECD, Market Access after the Uruguay Round: Investment, Competition and Technology Perspec- tives, Paris, 1996, pp. 160f. 76 As Ostry remarks, “The 1995 auto dispute between the United States and Japan was a lost opportunity to launch that process.” Ibid., p. 161. 

APPROACHES TO CONFLICT RESOLUTION 221 Technical Norms and Standards Major issues concerning market access are technical norms and stan- dards. Generally speaking, standards, whether aimed at social objectives like human health and safety (regulatory standards) or at reducing transac- tion and information costs (compatibility standards), can be considered as welfare enhancing.77 Yet, they may also serve as instruments to promote the international competitiveness of national industries, and they may act as barriers to trade and market access. In both cases they give rise to friction. In consideration thereof, in the Tokyo Round the Agreement on Technical Barriers to Trade (TBT) was concluded. In the Uruguay Round this Agree- ment was revised and strengthened in several respects. The Agreement now applies to all member countries of the WTO rather than to a limited number of signatories only, and it now covers not only the central government bodies but, though with major reservations, local government and non-gov- ernmental bodies, too. In the latter cases the member countries “shall take such reasonable measures as may be available to them to ensure compli- ance” with the provisions of the Agreement. The Agreement establishes national treatment and MFN status as the guiding principles. It obliges the member countries to ensure that technical regulations as well as conformity assessment procedures are not prepared, adopted or applied with a view to or with the effect of creating unnecessary obstacles to international trade. The key concept of unnecessary obstacles has been clarified to the effect that regulations and procedures shall not be more trade-restrictive than necessary to fulfil a legitimate objective, taking account of the risks non-fulfillment would create. Among such legitimate objectives are national security requirements, the prevention of deceptive practices, protection of human health or safety; animal or plant life or health; or the environment. Where relevant international standards exist they shall be used as a basis for national technical regulations, except if this would be ineffective or inappropriate because of “fundamental” climatic, geographi- cal or technological factors. Whereever appropriate, member countries shall specify technical regulations in terms of performance rather than design requirements. The Agreement sets forth provisions for the ex ante and ex post information on technical regulations (“transparency”); equitable proce- dures for conformity assessment, testing, and certification including infor- mation requirements, the respectance of confidentiality of information about products; equitable fees; recognition of conformity assessment in other member countries; and consultation and dispute settlement according to the rules of the WTO Understanding. 77 Cf. B.M. Hoekman and P.C. Mavroidis, International Antitrust Policies for High-Tech Industries? Paper presented at the conference “Towards a New Global Framework for High- Technology Competition,” held in Kiel, Germany, 30–31 August 1995, pp. 10f. 

222 CONFLICT AND COOPERATION Standards are set by the government, private standards organizations, and individual enterprises or alliances. There are no clear-cut boundary lines between the three: the—now dropped—European HDTV norm was agreed between the European Community and leading European manufac- turers. As major, and sometimes sole, buyers of specific technology prod- ucts, large state enterprises or monopolies, such as railroads, post and tele- communications, and other public utilities, have a decisive influence on the relevant technical standards. Most of the standards are set by the industrial standards organizations. In Europe, eighteen national organizations, such as DIN in Germany (with about 25,700 standards in 1986), Afnor in France (with 13,400), and BSI in the United Kingdom (with 9,400),78 work to- gether in the Comité Européen de Normalisation, (CEN), the Comité Européen de Normalisation Electrotechnique (CENELEC), and the European Tele- communications Standards Institute (ETSI), reaching even beyond the Eu- ropean Union. Standards decisions are taken by qualified majority under a system of weighted votes.79 In the United States, the process is more complicated. Approximately 400 organizations, working independently of one another and only part of them coordinating their activities in the Ameri- can National Standards Institute (ANSI), are involved in standards develop- ment.80 This puts obvious limits to any attempt to arrive at a binding plurilateral standards-setting procedure following the European model. On the international level, only seventeen out of 89,000 standards used by the United States in 1989 were adopted from ISO standards, a shortcoming which has been criticized as violating the GATT Standards Code.81 In standards enforcement, there are as many as 44,000 jurisdictions in the United States, resulting in “overlapping responsibility and redundant standards and regulations. In some cases, the products are regulated di- rectly through inspection or testing programs, or both. In other cases, an approval body may have to certify that products meet standards set by a particular state or municipal government. This becomes a technical barrier in cases where states and municipalities have regulations that apply differ- ent standards, or where certification requirements differ.”82 When testing 78 U.S. Congress, Office of Technology Assessment, Global Standards: Building Blocks for the Future, Washington, D.C., 1992, p. 63, based on: F. Nicolas and J. Repussard, Common Standards for Enterprises, Luxembourg, 1988, p. 26. 79 U.S. Congress, op. cit., p. 70. 80 Ibid., pp. 49ff. 81 B. Dhar, “The Decline of Free Trade and U.S. Trade Policy Today,” Journal of World Trade, vol. 26 (1992), no. 6, pp. 149ff., based on Commission of the European Communities, Report on United States Trade Barriers and Unfair Trade Practices, Brussels, 1991. 82 Department of Foreign Affairs and International Trade, Register of United States Barri- ers to Trade, 1994, Ottawa, 1994, pp. 11f., quoted in S. Ostry and R.N. Nelson, Techno- Nationalism and Techno-Globalism: Conflict and Cooperation, Washington, D.C., Brookings Institution, 1995, p. 94, footnote 15. 

APPROACHES TO CONFLICT RESOLUTION 223 organizations even within the United States need multiple state and local accreditation,83 is it conceivable that the principle of home country control be applied in trans-Atlantic (and/or trans-Pacific) trade? In the European Community, this principle (“country-of-origin” principle—mutual recogni- tion of standards and procedures) is a cornerstone of the so-called New Approach to Standardization adopted as part of the Single Market program. It is coupled, though, with minimal harmonization of performance require- ments and even, where necessary to ensure compatibility and interconnec- tion, of design requirements as well as of testing and certification proce- dures. Under the New Transatlantic Marketplace, a key component of the New Transatlantic Agenda endorsed at the U.S.-EU Summit in Madrid on 3 December 1995, the United States and the European Union undertake to work toward agreements recognizing each other’s testing and certification of products in a number of sectors, among them computers, telecommunica- tions, and business equipment.84 This is a first step in the same direction. In new areas of high-technology, the problem of an international harmo- nization of compatibility standards is increasingly being “solved” by stan- dards-setting by leading firms or strategic international alliances which seek to promote their own standards internationally, often in competition with rival groups, by-passing the traditional standards organizations. Among the examples are DOS vs. Apple and OS/2, VHS vs. Betamax and Video 2000, Hi-8 vs. S-VHS, and GSM vs. CDMA and PDC (the respective standards for mobile phones in Europe, the United States, and Japan). While the problem of international compatibility is solved to the extent that a firm standard is able to succeed as a new industry standard, issues of market power and collusive behavior, including “fair” access to the “club” for competing firms, may arise. This poses problems for international competi- tion policy. DUMPING, ANTIDUMPING AND COMPETITION POLICIES Private versus Public Conduct in Trade and Competition Apart from public obstacles to, and distortions of, high-technology trade and competition, trade-related barriers to market entry raised by private com- panies become increasingly important. This also means a greater role for competition policies and their international coordination. Existing interna- tional trading rules cover private business conduct to the extent that dumping practices are involved. An issue of particular interest in this context is the 83 Ibid. 84 USIS, “U.S.-EU Economic Relationship: the NTA Marketplace,” U.S. Information and Texts, 7 December 1995, p. 9; USIS, “Joint U.S.-EU Action Plan,” U.S. Information and Texts, 7 December 1995, p. 16. 

224 CONFLICT AND COOPERATION question of whether antidumping policies should—and could—in future be replaced by genuine competition policies. Competition policy will also have to deal with private business practices which so far are not subject to interna- tional disciplines but may equally disrupt international trade and competition. These include various forms of international technological cooperation. The dumping practices of companies and the antidumping policies of governments both lead to conflicts between countries as well as inside the borders. The international conflicts concern mainly producer interests. Foreign dumping causes friction if it hurts domestic industries. However, the poli- cies applied in response to dumping are no less contentious among trading partners if weak domestic industries are protected against efficient foreign competitors. In the national context, friction arises if antidumping policies work to the detriment of consumers and user industries. Defined as selling abroad at prices below “normal value,” dumping is assumed to be unfair per se, regardless of its overall welfare impact. For protection measures to ensue, it is normally sufficient that dumping is shown to cause “injury” (e.g., losses to sales, employment, or profit) or to threaten to do so, to a domestic industry or to retard its establishment. Antidumping therefore has become the instrument of choice for industries seeking to reduce competition from imports. In many cases, it is “just ordinary protec- tion with a good public relations program.”85 In contrast to its actual “capture” by special interests, the economic ra- tionale of antidumping policy (and indeed its original design in the United States in 1916 as an international extension of competition law to deal with cross-border violations of antitrust standards) is based on the existence of protected monopolistic market structures in foreign countries, facilitating “predatory” pricing strategies which finally lead to a dominant position of exporting countries on the importer’s market. For predatory dumping to take place, the foreign monopolist (or cartel) must not only eliminate exist- ing competition in the importing country, but must also be able to prohibit entry by new (domestic and foreign) competitors. It must either establish a global dominance or convince the host government to pursue a policy that tolerates or supports entry restrictions.86 However, even the logical founda- tion of the argument appears weak. It seems more reasonable to argue that both the predator and the would-be victim would be better served by colluding with one another or by the takeover of the latter by the former than by going through the evidently expensive procedure of dumping.87 85 J.M. Finger, Legalized Backsliding: Safeguard Provisions in the GATT [mimeo], Washington, D.C., 26 January 1995, p. 29. 86 See B.M. Hoekman and P.C. Mavroidis, Antitrust-Based Remedies and Dumping in International Trade, August 1994. 87 See P.K.M. Tharakan, “Anti-Dumping Policy and Practice of the European Union: An Overview,” Economisch en Sociaal Tijdschrift, 1994, no. 4. 

APPROACHES TO CONFLICT RESOLUTION 225 In actual fact, cases of successful predatory dumping remain largely un- documented. A recent study for the OECD concludes that “in the over- whelming majority of antidumping cases, there was little or no threat to competition in the domestic market.”88 This was true in more than 90 percent of the cases in which the United States and the EU imposed anti- dumping duties in the 1980s, and especially so with high-technology prod- ucts such as semiconductors, compact disc players, and plain-paper copiers. In particular, no clear evidence was found that Japanese producers of elec- tronic products applied “monopolizing” dumping designed to cripple for- eign competitors. Therefore, international dumping, as defined in antidumping regulations, largely appears to be normal competitive business behavior aimed at ex- panding foreign market shares, offsetting a sudden fall in demand, or ex- ploiting economies of scale. It is international price differentiation or “pen- etration” pricing below average cost (and even below short-run “static” marginal cost, which can be significantly higher than long-run “dynamic” marginal cost) that is consistent with free trade, as it boosts competition to the benefit of the domestic economy as a whole. In consequence, the poli- cies to curb “dumping” frequently harm the whole economy, as they pit the interests of a few, if powerful, domestic producers against those of the domestic “many” and of foreign competitors. A recent study by the U.S. International Trade Commission counts the net welfare loss for the United States of the 279 antidumping (and countervailing) duties that were in place in 1991 at roughly 1.6 billion dollars.89 Apart from the standard static welfare losses, antidumping measures have important dynamic consequences. Other than patent policies, which lower static welfare, but do so by increasing the long-run rate of growth, anti- dumping action, by isolating national markets, tends to slow the pace of innovation (and with it, growth). It may induce industrial users of the protected product to relocate plants outside the domestic territory, in order to avoid the duty, and in effect widen rather than close existing efficiency gaps between foreign and domestic companies. By indirectly increasing state interference with markets also in the exporting countries, policies may effectively bring about what they are supposed to fight: public protection of a dumper’s home market. Antidumping policies are also closely intertwined with other trade-pro- tective measures such as export-restraint arrangements, which impair eco- nomic welfare as well. Actually, the two forms of intervention have proved 88 OECD, Competition Policy and Antidumping: Summary and Conclusions [mimeo] Paris, October 1995, p. 7. 89 U.S. International Trade Commission, The Economic Effects of Antidumping and Countervailing Duty Orders and Suspension Agreements, Washington, D.C., June 1995. 

226 CONFLICT AND COOPERATION to be effective complements, as many antidumping cases were superseded by “voluntary” export restraints (VER). The threat of formal action under the antidumping law was the lever to force an exporter to accept a VER. It has also been shown, e.g., in the U.S.-Japan DRAMs case, that antidumping measures originally targeted at a limited number of product variants finally extended into an agreement covering the whole range of the respective product group.90 Dumping and Antidumping in High Technology In high technology, the two basic forms of “dumping,” i.e., regional price differentiation and (temporary) pricing below cost, seem to be wide- spread business practices reflecting central characteristics of the sector such as market segmentation, high fixed costs (especially up-front R&D expendi- tures), decreasing variable costs over time due to learning by doing, and first-mover advantages. All this should make high-technology producers particularly vulnerable to antidumping measures, and especially so in the early stages of production. Competition in high-technology markets has also been described as de- mand-driven product differentiation, where the ability of suppliers to accu- rately forecast consumers’ tastes (thereby creating temporary market access barriers subject to quick erosion) counts more than lasting entry deterrence based on an exclusive mastery of the technology at hand.91 This proposi- tion is consistent with the empirical finding of relatively low degrees of supplier concentration in many high-technology markets in the exporting countries, pointing to little leeway for predatory pricing. On balance, there are even more possibilities—and stronger incentives—for a protectionist use of the antidumping instrument in high technology than in “ordinary” business. In many cases, calculated dumping margins, and actual duties imposed, are indeed extraordinarily high for high-technology products.92 Even though, in terms of numbers of cases, high-technology products have 90 In fact, the DRAMs under antidumping measures represented only 7 percent of the whole semiconductor sector. See P.A. Messerlin, Reforming the Rules of Antidumping Poli- cies. Paper presented at the conference “Towards a New Global Framework for High-Technol- ogy Competition,” Kiel, Germany, 30–31 August 1995, pp. 7–8. 91 Messerlin points to the general availability of basic technologies, which is in most cases also suggested by the fact that many complainants in U.S. and EC antidumping cases involving high-technology products were the technological leaders of the product a couple of years before lodging the complaints. See P.A. Messerlin, Reforming the Rules of Antidumping Policies, p. 4. 92 For instance, in the U.S. EPROMs case, the calculated dumping margins ranged from 60 to 188 percent, in the EC DRAMs case from 8 to 378 per cent. See P.A. Messerlin, Reforming the Rules of Antidumping Policies, p. 9. 

APPROACHES TO CONFLICT RESOLUTION 227 not been the main target of antidumping policies, their incidence is high if measured by product (wide coverage of affected product groups), market (high volumes of sales and trade), and supplier (broad representation of major producers) characteristics. The largest case in the whole EC anti- dumping history, for instance, concerns plain-paper copiers, with imports of US $3.5 billion in 1992. In this case, 20 percent duties on Japanese photo- copiers, originally introduced in 1987, were reimposed (and extended to larger models) in October 1995, for three more years. The decision was reached by the Council of Ministers with a close (8–7), majority which was sufficient only after voting rules were changed (from qualified to simple majority) in March 1994. European imports of Japanese photocopiers have dropped sharply since the introduction of antidumping duties, but the mar- ket share of Japanese producers has nonetheless increased, as many of them have moved production inside the European Union.93 Empirical analysis for a number of electronic products (color television sets, compact disc players, and plain paper photocopiers) from Japan on which antidumping duties were imposed by the United States and the Euro- pean Community reveals a strong tendency of antidumping policies in this sector to protect domestic competitors rather than preserve competition in the long run.94 The case studies investigate certain structural characteristics of the respective markets and industries—such as entry barriers, relative home-market size, concentration ratios, market shares, and static and dy- namic economies of scale—that are potentially conducive to predatory or strategic dumping.95 Three results of the studies stand out: First, declining government protection by tariff and non-tariff barriers as well as subsidies contrasts with persisting low import penetration of the examined Japanese markets. Available evidence on private restrictions against import competition—such as incompatible product standards agreed by do- mestic companies, entry-deterring horizontal collusion among suppliers, and vertical restraints through exclusive distribution channels—is unable to ex- 93 See Financial Times, 19 October 1995 (G. de Jonquières and E. Tucker, “Pressure Grows for EU to Overhaul Dumping Policy”) and 15 September 1995 (E. Tucker, “EU Split over Dumping Duties on Japanese Copiers”). 94 See P.A. Messerlin and Y. Noguchi, Antidumping Policies and International Trade of Electronic Products [mimeo], Paris, OECD, August 1995. 95 Predatory dumping refers to low-priced exporting with the intention of driving rivals out of business in order to obtain monopoly power in the importing market, whereas strategic dumping describes exporting that injures domestic rivals through an overall strategy or general circumstances of the exporting nation, encompassing both the pricing of the exports and re- straints protecting the exporter’s home market. See R.D. Willig, The Economic Effects of Antidumping Policy [mimeo], Paris, OECD, October 1995, p. 6f. 

228 CONFLICT AND COOPERATION plain the phenomenon. In fact, low degrees of supplier concentration in the respective markets in Japan indicate rather that private entry barriers are moderate. Moreover, a limited number of domestic newcomers and foreign investors have successfully entered the markets considered in the analysis, the existence of interlocking “keiretsu” systems notwithstanding. Second, factual exclusion of American and European exporters from the investigated Japanese markets has not evidently disadvantaged these com- panies in international competition. In particular, it has not prevented them from exploiting scale economies in production, nor discouraged R&D ef- forts which for their profitability hinge on the number of units produced. Reliance on OEM (Original Equipment Manufacturing) agreements, for in- stance, by which firms are allowed to sell products of other companies under their own brand name, permits segment-specific economies of scale to be utilized without constraining the product range of each supplier. Finally, the case studies demonstrate that the recoupment of initial losses (and harvesting of some additional gain) on which the success of any preda- tory scheme crucially depends, is unlikely to happen. An important factor here is (actual or potential) market entry by third-country suppliers, e.g., European firms in the United States, challenging possible dominant posi- tions of potential predators. It has also been shown that access to the basic technologies needed is relatively easy, which increases the contestability of the markets in question. In sum, little evidence on the possible existence of predatory or strategic dumping practices has been found in empirical analysis. The Uruguay Round Approach to Antidumping The degeneration of antidumping measures into protective, selective, anticompetitive, and strategic trade policy devices has not been effectively contained in the Uruguay Round. The agreed changes will hardly transform antidumping action into competition policy. An attempt was made to cor- rect some of the biases embodied in national anti-dumping regulations and to remove inconsistencies in the use of procedures where rules often changed between cases and even methods within cases. However, no standards of competition law were adopted, and a number of highly questionable prac- tices were allowed to continue as long as they are “appropriately” explained. Of particular relevance for high-technology products in this context are the modifications regarding averaging methods in price comparisons and the treatment of start-up costs. In certain circumstances, and if appropriately explained, the new anti- dumping agreement allows the national authorities to deviate from the gen- eral method of comparing average home prices with average export prices, or individual domestic transactions with corresponding individual transac- 

APPROACHES TO CONFLICT RESOLUTION 229 tions abroad. Instead, dumping investigators may compare a weighted av- erage of domestic prices (omitting, if necessary, low-priced sales outside the “ordinary course of trade”) with individual export prices if the latter vary significantly among different purchasers, regions, or time periods. As this allows high-priced exports to be disregarded in the calculations, it is a sure way of finding dumping in almost every case, and particularly in cases where prices (and models) are as diverse and change as quickly as in high- technology products. It may be an easy task for policymakers to offer plausible explanations for the chosen calculation method, which withstands any possible challenge by the affected trading partners with relative ease. Past dispute settlement panels in antidumping cases in general used to treat lack of explanation as reason to recommend that an antidumping duty be revised, not as reason for it to be removed. With the new general (“integrated”) dispute settlement mechanism in place, this could possibly change in the future, as panel deci- sions no longer need the consent of the defending country. However, the specific dispute settlement section of the antidumping agreement gives the national determination of dumping margins the benefit of the doubt. It is explicitly recognized that the provisions of the agreement admit of more than one permissible interpretation. The panel is only to ask whether the antidumping authorities reached their conclusion without explicit error of fact or reasoning. This will make it very difficult for it to reject an explana- tion that has even the slightest plausibility.96 The provisions of the agreement concerning start-up costs or sales below total unit costs (including a “reasonable” profit margin) are equally ambigu- ous. On the one hand, below-cost sales in the exporter’s market have to be eliminated in the determination of the “normal value” if certain conditions are given.97 This raises the “normal value” and so tends to make the finding of dumping even more likely than before (when elimination was a mere possibility). On the other hand, the agreement for the first time re- quires nonrecurring items of cost (like R&D expenditures), which benefit future production, to be allocated over a longer period of time. In the case of start-up operations the costs at the end of the start-up period are relevant. This should lower the calculated normal value and hence constrain the finding of dumping. It could be particularly important in cases of low home-country sales (less than 5 percent of the disputed export sales) where 96 See B. Hindley, “Two Cheers for the Uruguay Round,” in: Trade Policy Review 1994, London, September 1994, p. 27. 97 Below-cost sales must occur over an extended period of time (normally one year), in substantial quantities (at least 20 percent of the total sales under consideration), and “at prices which do not provide for the recovery of all costs within a reasonable period of time” (Article 2 of the Agreement on Implementation of Article VI of GATT 1994). 

230 CONFLICT AND COOPERATION constructed values instead of direct price comparisons are used in the deter- mination of dumping margins. Other modifications of antidumping rules agreed in the Uruguay Round are a mixed blessing as well. The five-year sunset clause, according to which antidumping measures automatically expire after five years, was adapted from existing EU legislation. However, its value appears to be limited: (1) not only is it possible to extend the limit (and this is actually happening, as recently shown in the EU-Japan copier case noted above) but (2) the limit has also apparently not prevented European antidumping enforcement from being as biased as U.S. practices. For similar reasons, the new de minimis provision—which requires the immediate termination of an antidumping investigation when the margin of dumping is less than 2 percent of the export price, or the volume of dumped imports less than 3 percent of the respective total imports—seems unlikely to contain harassment of exporters in an effective way. At the same time, the more elaborate requirements for an antidumping investigation to be initiated after application may effectively lengthen the period of uncertainty about trading conditions, and thus discourage imports, since there is no time limit in the pre-initiation phase as compared with the subsequent stages. With regard to the determination of injury and the causal link between dumping and injury, the new agreement provides little change from the preceding antidumping code of the Tokyo Round, and so allows considerable scope for arbitrary decisionmaking (e.g., in the definition of “injured” domestic industries, which may include or exclude foreign subsid- iaries) to continue. By and large, the substantive and procedural conditions imposed by the new antidumping agreement on policies in this area appear unable in them- selves to effect substantial change. Some of the improvements reached have also been eroded in national implementation legislation reflecting the political power of protection-seeking interests. Apparently, even the small- est content of ambiguity of any word has been intensely exploited. Amend- ing antidumping procedures has been depicted as an activity with a produc- tivity close to Sisyphus’s; new provisions seem to be deviated from their initial purpose at a more rapid pace than they were introduced.98 However, antidumping rules and practices remain far removed from the standards of sound competition policy which they were originally supposed to supple- ment at the international level. Disciplining antidumping policies should therefore rank high among the market access issues on the post-Uruguay Round trade agenda. 98 See P.A. Messerlin, Reforming the Rules of Antidumping Policies. Paper presented at the conference “Towards a New Global Framework for High-Technology Competition,” Kiel, Germany, 30–31 August 1995, p. 12. 

APPROACHES TO CONFLICT RESOLUTION 231 Reforming Antidumping Policy The first-best solution, from an intervention-theoretic point of view, would be to substitute competition rules and procedures, either harmonized and coordinated among nation-states or truly globalized, for existing antidump- ing laws. This would allow the removal of distortions, such as abuse of a dominant market position in the exporting country, at the source, rather than merely compensating for them. However, this is a long-term perspec- tive at best, unlikely to materialize in the near future. It sharply contrasts with political realities, which are characterized by a proliferation of anti- dumping regulations and an unwillingness of the most prominent user coun- tries and regions (the United States and the European Union) to dispense with an expedient instrument of trade policy. Protection-seeking interests would have to be accommodated in one form or another anyway, which would leave little net value in abolishing antidumping measures. The prevailing political view is of raising international standards in the area of competition law, thus reducing the need to invoke the antidumping law, i.e., competition laws working effectively alongside the antidumping law without substituting for it. In fact, there have been only a few cases of anti-dumping policies being superseded by competition policies. Apart from the European Union, these are the formation of the European Economic Area and, to a lesser extent, the Australian-New Zealand Closer Economic Relations Trade Agreement. In the face of the widely perceived political need to retain the antidump- ing option in trade policy, reforming policy in this area could proceed on four tracks: • Removal of trade and investment barriers in the exporting (and im- porting) countries • Ad-hoc review of the state of competition (actual and potential) in the exporting-country markets and of existing public barriers to entry (and pres- ence) by foreign competitors • “Importing” competition-policy standards into antidumping legislation, and raising the legal standing of consumer and industrial-user interests in the investigations • Developing international competition rules. The first track involves ongoing liberalization of trade on a reciprocal basis, including the removal of remaining tariff, non-tariff, and regulatory barriers to market access raised by governments, as well as the negotiation of an international investment agreement that guarantees foreign companies freedom of establishment and national treatment in host countries. This will significantly change the environment for predatory behavior and make it even less likely to happen. By the same token, it weakens the legitima- 

232 CONFLICT AND COOPERATION tion bases of antidumping policies. However, as international negotiations on broadly based liberalization and deregulation take time, no quick impact on dumping practices and antidumping measures can be expected. A more immediate impact would have a multilateral agreement, as pro- posed by Hoekman and Mavroidis (1994),99 providing for an examination (preferably by competition authorities) of market characteristics (including private and public entry barriers) in the exporting country to precede any antidumping investigation. The policy objective in this second track would be to establish whether the structural preconditions for economically harm- ful (i.e., reducing the importing country’s welfare) dumping strategies are given or not. An antidumping petition would consequently have to be turned down from the outset, irrespective of the existence of dumping, if (1) the competition agency in the exporting country should find the accused firms neither engaged in anticompetitive practices nor benefiting from gov- ernment-created or supported market access barriers against foreign compa- nies, and (2) the corresponding authorities of the importing country agree with the finding. In the third track competition aspects are considered after an antidump- ing investigation has been launched. The focus now shifts from the export- ing to the importing country market, where the quality of the injury in- flicted on domestic industries through foreign dumping apparently varies depending on the state of competition. In highly concentrated markets with high entry barriers, for instance, “injury” often means a reduction of non- competitive excess returns. “Dumping” would correspondingly work to “disrupt” monopolistic market structures—to the benefit of other domestic agents—and to enhance general welfare.100 The injury test in antidumping laws should accordingly be redesigned and expanded so as to more compre- hensively include the interests of consumers and industrial users and review the state of competition in the domestic markets concerned. It could possi- bly also include an examination of whether antidumping duties would help domestic industries to compete more effectively. A more competition-oriented approach to antidumping cases could in- volve the adoption of competition-policy concepts such as “relevant mar- 99 See B.M. Hoekman, and P.C. Mavroidis, Antitrust-Based Remedies and Dumping in International Trade. 100 For example, Tharakan finds most of the EU antidumping impositions to have taken place in favor of industries which have a high degree of concentration in the European Union. He concludes that the lobbying power of oligopolistic industries is effectively used by them to obtain protection from import competition through antidumping measures. See P.K.M. Tharakan, “AntiDumping Policy and Practice of the European Union: An Overview,” Economisch en Sociaal Tijdschrift, 1994, no. 4, p. 565. 

APPROACHES TO CONFLICT RESOLUTION 233 ket” and “critical degrees of concentration” in antidumping investigations, in order to correct the biases inherent in current antidumping concepts such as “like product” and “major proportion of the domestic industry.” Defin- ing quantified “thresholds” (comparable to the “tariffication” process in the Uruguay Round) would facilitate multilateral negotiations in this area. In addition to these reinterpretations of existing antidumping provisions, the appeals system in antidumping cases could be substantially improved. As proposed by Messerlin,101 an antidumping case could have to be passed from antidumping authorities to antitrust agencies, e.g., from DG I to DG IV in the European Commission or to the cartel offices in individual EU member states, should a court ruling repeal the case for the formers’ neglect of competition aspects. In sum, to the three elements of an antidumping investigation (unfair trade, injury, and causal link between the two) a fourth element—competition analysis—would be added.102 It would be a logical complement to the pre-test of competition on the exporting-country market noted above. The fourth track entails the codification of international competition rules. This would contribute importantly to an eventual elimination of antidump- ing policies. Their final replacement by competition policies is feasible, though, only with a high standard of integration between countries. This would require realization of the “four freedoms” (i.e., complete liberaliza- tion of trade in goods, services, labor, and capital) plus agreement on com- mon disciplines for, or mutual recognition of, certain policies and regula- tions regarding in particular subsidies, government procurement, technical standardization, services, and for that matter competition. Removing the antidumping instrument accordingly seems to be possible in the foreseeable future only for limited numbers of countries and typically in a regional context. This, again, bears chances as well as risks, since increased compe- tition—and related adjustment pressures—inside the “club” might provoke compensatory antidumping action against outsiders. Toward Multilateral Competition Rules International coordination, or harmonization, of competition policies may not only help to contain dumping practices and counter-productive anti- dumping policies, but also prevent anticompetitive business conduct in in- ternational trade in general. This may essentially take three forms: 101 See P.A. Messerlin, Reforming the Rules of Antidumping Policies, p. 13. 102 See also H. Vandenbussche, “How Can Japanese and Central European Exporters to the European Union Avoid Antidumping Duties?” World Competition, March 1995. 

234 CONFLICT AND COOPERATION • Exclusion of foreign suppliers from domestic markets by the exercise of horizontal market power through collusive behavior or vertical restraints of competition such as exclusive dealerships between companies of one country. This may effectively—and asymmetrically—thwart the liberaliza- tion of trade negotiated by governments. • Formation of export cartels among a country’s firms, and similar ac- tion to inhibit foreign competitors in international markets, with the ulti- mate effect of significantly raising international prices. With asymmetric incidence of private trade barriers between countries, trading partners may in this case even suffer net losses from “public” liberalization. • Restraint of competition between companies from different countries in the form of strategic alliances (frequently of a technological nature), cartels, mergers, and aquisitions in particular. Again, this could impair the benefits of liberalization. However, in the majority of cases, international cooperation among firms seems to stimulate rather than reduce competition on national markets, as these become increasingly global in the course of trade liberalization. On this background, international policymaking in the field of competi- tion faces two major challenges: • Adjusting existing rules and procedures in national competition poli- cies—or building them from scratch—and monitoring their implementation, so as to remove—and prevent—trade-restrictive and mercantilistic biases in competition policy. In particular, strict enforcement of the agreed provi- sions would have to be guaranteed and export cartels prohibited if they go beyond a mere sharing of expertise and services in supplying foreign mar- kets. Export cartels may also be conducive to collusive behavior on domes- tic markets. Devising principles—and practical guidelines—for the international co- operation among national agencies responsible for competition policy, in order to resolve conflicts that may arise when private restraints of competi- tion simultaneously affect a number of national markets. Such conflicts are likely to occur also if competition laws—and the weight given to competi- tion as compared with other policy aims—are largely identical between the countries concerned. This is because in many cases the impact of the anticompetitive practices will vary from country to country.103 In the final analysis, international action may require a supranational authority to coor- 103 With markets becoming increasingly global, international antitrust frictions of this kind should, however, tend to diminish. See R. Jungnickel and G. Koopmann, “Globalization of Business—Implications for International Competition and Related Policies,” in E. Kantzenbach, H.-E. Scharrer, and L. Waverman (eds.), Competition Policy in an Interdependent World Economy,. Hamburg, 1993, pp. 44–45. 

APPROACHES TO CONFLICT RESOLUTION 235 dinate the collection of information, and to apply any sanctions to the firm or firms involved.104 In general, greater market interpenetration due to the integration of na- tional markets has resulted in the need for competition policies to adjust to different circumstances of international competition.105 More specifically, and with particular relevance to high technology, an important rationale, denoted by Soete,106 for international competition policy aimed at counter- acting the emergence of worldwide cartels between global firms could be to challenge national strategic policies based on arguments of dependence on “foreign” monopoly pricing. Other reasons for specific (national as well as international) antitrust concerns in high technology refer to (1) network externalities (e.g., in computer software), which could bias industries to- ward monopoly, as the value placed on network membership by a consumer rises with the number of other people on the network; (2) systems leverage, by which a firm which controls one part of a system may spread its mo- nopoly into others through, e.g., cross-subsidization;107 (3) standardization, which may lend individual firms, and groups of companies, considerable market power, as consumers and user firms get “locked” into the system and as entry barriers, through costs of “switching” systems, are raised; and (4) innovation cartels, inhibiting the innovation process proper (by reducing the number of innovative efforts) and possibly “spilling over” into other fields of activities such as production and marketing.108 The actual development of international competition rules is still in its infancy. It could be advanced through “case law” built in dispute settle- ment procedures based on existing GATT provisions. Of particular interest in this context is the non-violation clause of GATT Article XXIII:1(b), designed to address the concern of GATT members relating to a modifica- tion of negotiated concessions through subsequent government action in 104 See D.A. Hay, The Economic and Trade Effects of Anti-Competitive Practices in a Globalising World Economy. [mimeo], Paris, OECD, 26 April 1995. 105 See P. Lloyd and G. Sampson, “Competition and Trade Policy: Identifying the Issues after the Uruguay Round,” The World Economy, vol. 18, no. 5, September 1995, p. 686. 106 See L.G. Soete, Technology Policy and the International Trading System: Where Do We Stand? Paper presented at the conference “Towards a New Global Framework for High- Technology Competition,” Kiel, Germany, 30–31 August 1995, p.15. 107 For instance, Microsoft, the leading computer-software firm, has been accused of using profits from the operating-system market to subsidize applications programs. See “Thoroughly Modern Monopoly,” The Economist, 8 July 1995, p. 92. 108 However, empirical evidence on competitive restraints involved in, and flowing from, technological cooperation is scarce. A possible loss of variety seems to be more than compen- sated for by substantial economies of scale and gains in efficiency. For an overview of trends in (national and international) technological partnering see J. Hagedoorn, The Economics of Cooperation Among High-Tech Firms—Trends and Patterns in Strategic Technology Partnering since the Early Seventies, in: G. Koopmann and H.E. Scharrer (eds.), The Economics of High- Technology Competition and Cooperation in Global Markets, Baden-Baden, 1996, p. 173ff . 

236 CONFLICT AND COOPERATION areas that either were not addressed by the GATT or did not violate a GATT obligation. An example could be an exemption by the competent antitrust authorities granted to private enterprises, that effectively reduces market access opportunities for products of third countries by establishing difficult-to-penetrate distribution channels.109 In conjunction with a signifi- cantly strengthened dispute settlement procedure, nonviolation complaints could therefore lead to the opening of foreign markets closed by private restrictive practices. However, the wording of the clause is too general, and its reach too limited, for it to serve as a solid base for an effective interna- tional competition policy. Its “structural” weakness is to be, by definition, not based on explicit provisions concerning anticompetitive business con- duct in international trade (with the exception of dumping). This is different with the Draft International Antitrust Code (DIAC) pro- posed by an international group of legal experts.110 The Draft Code pro- vides for minimum standards to be observed by the antitrust authorities in the participating countries (which need not include all WTO members, as the DIAC is designed as a plurilateral GATT-WTO agreement) in dealing with private restraints of competition that affect international trade flows. The practices covered range from the formation of international cartels to the abuse of dominant market positions and from horizontal to vertical restraint. An international competition agency, equipped with the right of an International Procedural Initiative, would be entrusted to safeguard the application of the national law in cases where an inactive member state would not take its own initiative; if necessary, it could sue in the national courts to ensure application. Conflicts arising from impacts in one country of competitive restraints originating in another would be addressed by tak- ing advantage of the new WTO dispute settlement procedure. Even though this proposal as a whole is fraught with a number of difficulties regarding its practical implementation and enforcement, as well as inevitable losses of national sovereignty,111 it seems in principle to be well suited to meet the challenges noted above, including the specific antitrust issues related to high technology. It could effectively flank the international trading order with an international competition regime. 109 See B.M. Hoekman, and P.C. Mavroidis, Antitrust-Based Remedies and Dumping in International Trade, Policy Research Working Paper 1347, The World Bank, Washington, D.C., August 1994, p. 20. A comprehensive discussion of non-violation in the competition context is given in B.M. Hoekman, and P.C. Mavroidis, “Competition, Competition Policy and the GATT,” The World Economy, vol. 17, 1994, pp. 121–150. 110 For the text and a detailed explication see W. Fikentscher, “Competition Rules for Private Agents in the GATT/WTO System,” Außenwirtschaft, vol. 49, no. 2/3 1994, pp. 281– 325. 111 Klodt, regarding the proposed minimum standards rather than maximum standards, points to the particular problems of establishing an international merger control, as well as of rules for the control of an abuse of market power. See H. Klodt, “Internationale Regeln für den Wettbewerb,” Wirtschaftsdienst, October 1995, p. 560.