PAGE 1
Competing Economies: America, Europe, and the Pacific Rim October 1991 OTA-ITE-498 NTIS order #PB92-115757
PAGE 2
Recommended Citation: U.S. Congress, Office of Technology Assessment, Competing Economies: America, Europe, and the Pacific Rim, OTA-ITE-498 (Washington, DC: U.S. Government Printing Office, October 1991). For sale by the U.S. Government Printing office Superintcndent of Documents, Mail Stop: SSOP, Washington, DC 20402-9328 ISBN 0-16 -0359333 ii
PAGE 3
In Appreciation H. J OHN H EINZ III 1938 to 1991 The staff of the assessment on Technology, Innovation, and U.S. Trade dedicates this report to Senator John Heinz. Committed to solving the problems of American manufacturing and workers, Senator Heinz contributed to the legislative process an enthusiasm for new ideas and optimism that solutions could be found. He was a requestor of this study and one of the first in Congress to grapple with competitiveness issues. His sense of humor delighted those privileged to work with him, his sense of seriousness made him a strong champion of American manufacturing. Ill
PAGE 4
Foreword Despite macroeconomic adjustments and a lot of discussion, U.S. manufacturing is still in trouble. A test of national competitiveness is whether standards of living improve while world market share holds steady or increases. The United States is failing on both counts. Over the past two decades, the U.S. share of world imports has increased, its share of world exports has decreased, and imports have captured an increasing share of the U.S. market. At the same time, real incomes of the large majority of all Americans have declined. Studies of specific industries also support the claim that American manufacturing is not keeping up with the competition, the Japanese in particular. While U.S. auto manufacturers have improved in productivity and quality, the Japanese still make more reliable vehicles at lower costs. Japanese computer manufacturers have made rapid progress: even in supercomputers, one of the most advanced sectors, three Japanese firms are now threatening to overtake the U.S. leaders. Like Japan, Korea and Taiwan have mostly succeeded in promoting development of their economies in general and several advanced industries in particular. We can learn several lessons from these examples. First, the successful governments are partners with the private sector in developing particularly risky technologies and industries. Second, they combine many different policies including, when necessary, limitations on foreign companies competition in the domestic market. Third, they patiently persevere with development strategies, often for decades. Finally, they preserve important market disciplines, making sure that domestic companies competed with their most proficient rivals. This report ex amines how the economic environment of the United States can be made more conducive to improving manufacturing performance. It considers how Federal institutions, in cooperation with industry, can develop competitiveness strategies for high-tech, fast growing industries; and how trade, financial, and technology policies could be combined into a strategic competitiveness policy. This is the final report in a series of three in OTAs assessment of Technology, Innovation, and U.S. Trade. The first two reports, Paying the Bill: Manufacturing and Americas Trade Deficit and Making Things Better: Competing in Manufacturing, concluded that there are many signs of weakness in American manufacturing technology and identified policy options to improve the ability of American manufacturers to develop and diffuse technology. u JOHN H. GIBBONS Director
PAGE 5
Advisory Panel Technology, Innovation and U.S. Trade Peter J. Sharfman, Chairman Director of Policy Analysis The MITRE Corp. Alice Amsden New School for Social Research Economics Department Marietta L. Baba Professor of Anthropology and Acting Chair Wayne State University Wolfgang Demisch Managing Director, Director of Research UBS Securities Jean-Jacques Duby Group Director, Science and Technology IBM Europe Paris LaDefense, France Mark Eaton Director, International and Associated Programs Microelectronics & Computer Technology Corp. John T. Eby Executive Director, Corporate Strategy Ford Motor Co. Larry French Consultant Glen S. Fukushima Director, Public Policy and Business Development AT&T Japan Ltd. Tokyo, Japan Edward Graham Research Fellow Institute for International Economics Joseph A. Greenwald Attorney/Consultant Howard Greis President Kinefac Corp. Peter Hall Professor of Government Harvard University F. William Hawley Director, Governmental Relations for International Affairs Citicorp Sanford Kane President and CEO PCO, Inc. Robert Kuttner Economics Correspondent The New Republic Alek A. Mikolajczak Vice President, Technology and New Business Rohr Industries Susan Pharr Professor of Government and Director, U.S.-Japan Relations Program Harvard University Margaret Sharp Senior Research Fellow Science Policy Research Unit University of Sussex East Sussex, England Laura Tyson Professor of Economics University of California, Berkeley John D. Warner Vice President, Engineering Boeing Commercial Airplane Group Stephen Woolcock. Fellow Royal Institute for International Affairs London, England NOTE: vi OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members. The panel does not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for the report and the accuracy of its contents.
PAGE 6
OTA Project StaffCompeting Economies: America, Europe, and the Pacific Rim Lionel S. Johns, Assistant Director, OTA Energy, Materials, and International Security Division Audrey B. Buyrn, Program Manager Industry, Technology, and Employment Program Julie F OX Gorte, Project Director Katherine Gillman, Deputy Project Director Robert Weissler, Senior Analyst John Alic George Eberstadt Robin Gaster Jeffrey B. Lewis Sebastian Remoy Robert Wade Elizabeth Sheley, Editor Administrative Staff Carol A. Guntow, Office Administrator Christopher N. Chuy, Office Administrator Diane D. White, Administrative Secretary Publishing Staff Katie Boss, Publishing Officer Cheryl Davis Dorinda Edmondson Denise Felix Steve Kettler 2 Chip Moore Christine Onrubia Susan Zimmerman Contractors Marie Anchordoguy BSI Consulting, Inc. Stephan Haggard Keith Hayward Hugo Letiche Takashi Mashiko Todd Watkins James Womack and Daniel Jones IUntil June 1990. Wntil September 1991. vii
PAGE 7
Page Chapter 1. S ummary and Findings . . . . . . . . . . . . . Chapter 2. Policy Issues and Options . . . . . . . . . . . . Chapter 3. American Firms, Foreign Firms: Contributions to the Nation . . . . Chapter 4. U.S. Trade Policy . . . . . . . . . . . . . . Chapter 5. EC-92: Trade and Industry Policy . . . . . . . . . . . Chapter 6. Japanese Industrial Policy: The Postwar Record and the Case of Supercomputers Chapter 7. The New Competitors: Industrial Strategies of Korea and Taiwan . . . Chapter 8. Government Support of the Large Commercial Aircraft Industries of Japan, Europe, and the United States . . . . . . . . . . . Index . . . . . . . . . . . . . . . . . . . . 3 41 85 117 191 239 295 341 365 WI
PAGE 8
Chapter 1 Summary and Findings
PAGE 9
. CONTENTS THE COMPETITIVENESS PROBLEM . . . . . . . . . The Role of Competition: Are Free Markets Always Best? . Industrial Targeting . . . . . . . . . . . . . . . . . . Import Dependence . . . . . . . . . . . . . . SUMMARY OF POLICY ISSUES AND OPTIONS . . . . . . . Institutions for a Strategic Competitiveness Policy . . . . . . . . . A Technology-Friendly Environment . . . . . . Trade Policy Options . . . . . . . . . . . . . . . . . . . GOVERNME NT INDUSTRY PARTNERSHIPS . . . . . . . Risk Sharing in Technology Development . . Financial Risk-Sharin g . . . . . . . . . . . . . Government Purchases . . . . . . . . . . . . . . . . . . . . AMERICAN FIRMS, FOREIGN FIRMS, AND AMERICAN INTERESTS . . . . U.S. TRADE POLICY ...,.,.....,,.,.. EUROPE AND THE SINGLE MARKET . . . . . . . . . . . . . . . . JAPAN . . . . . . . . . . . . . INDUSTRIAL POLICIES IN TAIWAN AND KOREA l . . . . . . . . . . . GOVERNMENT SUPPORT OF LARGE COMMERCIAL AIRCRAFT . . . Figures Figure 1-1. U.S. Imports as a Percent of GNP . . . . . . . . 1-2. Merchandise Trade Balance, Percent of GNP l . . . . . . 1-3. Hourly Wages, Manufacturing : 1982-84 dollars 1-4. Weekly Wages, Manufacturing : 1982-84 dollars . . . . . . . . . . . 1-5. Weekly Wages and Salaries of Full-Time Employees, 1969-90, 1982-84 1-6. Japanese Share of World Trade . . . . . . . .,.,*,.*. 1-7. Imports as a Percent of GNP, Japan . . . . . . . . 1-8. Trade Balances as a Percent of GNP, Japan ... ... ... .*. +., ,**. ..+ O**** . . . . . . . dollars . . . . . .+ . ., . .* Page 3 7 9 11 13 15 16 20 23 23 ., 25 25 26 28 30 31 l 33 34 Page 4 4 4 5 5 6 6 6 Table Table Page 1-1. U.S. Share of World Imports and Exports . . . . . . . . . 3
PAGE 10
Chapter 1 Summary and Findings THE COMPETITIVENESS PROBLEM U.S. manufacturing is falling behind the competition. The standard of comparison is Japan; if Japanese companies were not such successful competitors in many important industries, Americas competitiveness problem would be far smaller. Eventually, other East Asian nations like Korea and Taiwan probably will rival Japan as top international competitors, and the European Community is trying to improve its own competitiveness. But for now, it is Japanese industry that poses the most formidable challenge. What is the evidence that American manufacturing has lost competitiveness? There is no widely accepted single measure of competitiveness. Analysts use many proxies: international trade balances, comparative international figures on productivity or standards of living, manufacturings share of gross national product (GNP), and comparative studies of the performance of individual industries are common ones. However, examin ing U.S. performance piecemeal in a few of these areas does not give a consistent picture. So, despite growing acceptance of the notion that U.S. manufacturing is in competitive trouble, the debate persists. A logical way to sort out the usefulness of these measures is to begin with a definition of competitiveness. One that stands out as being most useful is defined in the 1985 report of the Presidents Commission on Industrial Competitiveness: Competitiveness is the degree to which a nation can, under free and fair market conditions, produce goods and services that meet the test of international markets while simultaneously maintaining or expanding the real incomes of its citizens. There are two pertinent criteria here: meeting the test of international markets, and maintaining or expanding real incomes. Free and fair market conditions is a qualifier. We shall consider these items one by one. The most obvious way to interpret meeting the test of international markets is in terms of world market share. There is no direct, single measure of U.S. manufactured goods as a share of global manufacturing output, but if we combine data on U.S. shares of world imports and exports with figures on the proportion of U.S.-made goods in domestic consumption, the picture that emerges is unequivocal: the United States has lost world market share in merchandise. l The U.S. share of world merchandise imports has climbed from less than 15 percent in 1968 to over 17 percent in 1988; its share of world merchandise exports has fallen from nearly 16 percent to less than 13 percent over the same period (table 1-1). 2 At the same time, imports captured an increasing share of the U.S. domestic market, going from about 3 percent of GNP in 1960 to over 9 percent in 1989 (figures 1-1 and 1-2). To be sure, it was natural and expected for the United States to lose market share as the worlds developing and war-tom economies improved their performance. Sinking market share alone is not proof of failing competitiveness. Had Americans become better off in the process of producing a smaller share of world output and domestic consumption, competitiveness would not be the issue it is now. If American manufacturing had stayed robustly competitive, we would expect to see the living standards of manufacturing workers increase. Instead, they have pl ummeted. Real hourly wages of manufacturing production workers peaked in 1978 at almost $9.50 per hour; 3 by 1990, they had sunk to almost $8.00, below the 1964 level (figure 1-3). Real weekly wages of manufacturing workers also Table 1-1U.S. Share of World Imports and Exports Percent Percent Year of imports of exports 1970 . . . . . 12.9 13.8 1973 . . . . . 12.4 12.4 1975 . .. .0...... . 11.7 12.7 1977 . . . . . 13.6 10.8 1978 . . . . . 13.8 11.1 1979 . . . . . 13.1 11.1 1980 . . . . . 12.5 11.1 1981 . . . . . 13.4 11.9 1982 . . . . . 13.4 11.6 1983 . . . . . 14.4 11.1 1984 . . . . . 17.2 11.5 1985 . . . . . 17.9 11.1 1986 . . . . . 17.5 10.3 SOURCE: United Nations, Department of International and Social Affairs, 1985/86 Statistical Yearbook, 35th Issue (New York, NY: United Nations, 1988). -3
PAGE 11
4 l Competing Economies: America, Europe, and the Pacific Rim Figure I-l U.S. Imports as a Percent of GNP Percent 0 ~ 2 4 b 6 8 I 1 1 I I 1 1 I 1 1 I 1 1 1 1 1 1 -1 0 I I 1 1 1 1 1 1 1 1 1 1 19601963196619691972 19751978198119841987 1990 SOURCE: Economic Reporf of the President, (Washington, DC: U.S. Government Printing Office, February 1990), table C-1; and U.S. Department of Commerce, Bureau of Economic Analysis, international Transactions, Survey of Current Business, June 1991. Figure 1-2Merchandise Trade Balance, Percent of GNP Percent 9. I -1 1 I 3 w 1 1 1 I 1 1 1 c 4 [ 1 1 1 1 1 I I 1 1 I \ 1960 1964 1968 1972 1976 1980 1984 1988 SOURCE: Economic Report of the President (Washington, DC: U.S. Government Printing Office, February 1990), table C-1; and U.S. Department of Commerce, Bureau of Economic Analysis, International Transactions, Survey of Current Business, June 1991. peaked in the late 1970s, but have since fallen to levels of the late sixties (figure 1-4). 4 Looking more broadly at the workforce does not improve the picture. Real hourly and real weekly wages of all production and nonsupervisory workers--over 74 million people by the end of 1989, or 63 percent of the employed civilian workforce-have been sinking too. After peaking in 1972, real hourly wages dropped back to where they were in the mid-sixties, and real weekly wages declined much further. A still broader measure of living standards is wages and Figure 1-3-Hourly Wages, Manufacturing: 1982-84 dollars Hourly wage $1 O $9. 5 $ 9 $8. 5 I $ 8 I I 1 1 1 1 I 1 I 1 1 I I I 1 1 1 1 1 I I I 1 I 1 1 1 1 1 66 68 70 72 74 76 78 80 82 84 86 88 90 SOURCE: U.S. Department of Labor Statistics, Employment and Earnings, various issues. salaries of all full-time workers (73 percent of the workforce). Here, too, are losses: real weekly wages were about $330 in 1969, and below $320 in 1990, and have been falling since 1987 (figure 1-5). In sum, the living standards of the large majority of all Americans have dropped. So U.S. manufacturing fails the test of improving competitiveness on two counts: decisively on meeting the test of international markets, and substantially on increasing standards of living. Have these happened under conditions of free and fair markets? Here, the evidence is not conclusive. While markets have grown more free over the postwar period in terms of the general level of tariffs and quotas, many analysts would argue that nontariff barriers have proliferated. Whether the overall effect is one of increasing market openness, at least for a majority of our largest trading partners, is unproven. There is another way of looking at this issue. The only explanation of U.S. losses of world market share in merchandise and losses of real income for the majority of Americans other than a drop in competitiveness would be a substantial, progressive closure of international markets. Furthermore, we would also have to make a case that abroad array of American industries needed access to other big markets (Canada, Japan, and Europe) in order to maintain competitiveness. While there are some products for which increasing returns to scale make efficient-sized enterprises too large for all but global markets, for example, large commercial aircraft and supercomputers, it is hard to argue that the U.S.
PAGE 12
Chapter 1--Summary and Findings l 5 Figure 1-4Weekly Wages, Manufacturing: 1982-84 dollars $380 $360 1 $340 [ 1 1 1 1 1 ( 1 1 I 1 1 1 1 I I I I I 1 1 1 $320 I 1 I 196466 68 70 72 74 76 78 80 82 84 86 88 90 SOURCE: U.S. Department of Labor Statistics, Employment and Earnings, various issues. Figure 1-5-Weekly Wages and Salaries of Full-Time Employees, 1969-90, 1982-84 dollars $380 $300 1969 1972 1975 1978 1981 1984 1987 1990 SOURCE: U.S. Department of Labor Statistics, Employment and Earnings, various issues. market is too small to support competitive industries in most of manufacturing. In view of what is known about decreasing formal trade barriers, the burden of proof is on anyone who would argue that market closure accounts for the drops in American world market share and standards of living since the 1970s. This does not mean that every American industry is uncompetitive or is growing less competitive. In fact, competitiveness is best understood at the level of industries and even companies. What these figures tell us is that, at least in the most important sectors, U.S. companies are not holding their own against foreign competition. In particular, American industries are beleaguered by Japanese competitors. Japans record over the postwar period is in many ways a mirror image of Americas. Japans share of world exports increased 3 percent per year between 1968 and 1988, while its share of world imports increased 0.8 percent annually (figure 1-6). Japanese companies also held their own in their home market (figures 1-7 and 1-8). At the same time, Japan has been able to sustain brisk growth in living standards compared with the rest of the developed world. Real gross domestic product (GDP) per capita rose an average of 6 percent per year between 1950 and 1989, faster than in any other developed nation (the United States averaged only 1.9 percent per year, and Western European countries between 2 and 3.5 percent). To be more accurate, Japan in the 1950s may have fit more into the category of developing than developed nation, and therefore had more potential for very rapid growth. But even after this development period, Japans growth in real GDP per capita was higher than that of any other developed nation between 1979 and 1989 as well, averaging 3.5 percent per year, about double the rate of most European countries and the United States. s Real earnings per employee in manufacturing increased 3.6 percent per year, on average, between 1968 and 1985, and gross national income per capita went from 55 percent of U.S. levels in 1968 to 88 percent two decades later. 6 While citizens in the United States still earn more and live better than citizens of Japan, most Americans are not becoming better off, and most Japanese are. The complication in comparing the records of the two countries comes in the caveat free and fair market conditions. Japans market was anything but open in the early postwar decades. Due to a combination of business practices and government policies, it is still one of the worlds most difficult markets to penetrate. Under these conditions, it is not surprising that Japans companies did well in their domestic market, or conversely, that manufactured imports did so poorly. But while this lack of permeability kept Japans people from consuming as much or living as well as they would have with the same income in the United States, it is also true that Japanese companies in many industries did meet the test of international competition in the more open
PAGE 13
6 l Competing Economies: America, Europe, and the Pacific Rim Figure 1-8-Japanese Share of World Trade 12% 1 o% 8% 6%? 4% 2% o% I I I I I 1 I I I I I I I I I I I I I 1968 1972 1976 1980 1984 1988 + % of world exports +% of world imports SOURCE: Economic Report of the President, (Washington, DC: U.S. Government Printing Office, February 1990), table C-1; and U.S. Department of Commerce, Bureau of E&momic Analysis, International Transactions, Survey of Current Business p June 1991. Figure l-7imports as a Percent of GNP, Japan 1.2% 1.0% [ 0.8% 0.6% 1 0.4% o% 19601963196619691972 19751978 1981 19841987 SOURCE: Organization for Economic Cooperation and Development, Department of Economics and Statistics, National Accounts, various issues. markets of the United States, Canada, and Western Europe. In fact, there are other conditions that affect competitive performance. Currency value is an important one. In the early 1980s, the high value of the dollar was widely held to be primarily responsible for the nations plunge into deep trade deficits, Figure 1-8Trade Balance as a Percent of GNP, Japan 0.4% 0.3% A 0.2% I 0.1% o% -O.l% [ -0.2% [ 1 i [ 1 t I 1 1 1 1 1 1 1 I I 1 1 1 I 1 1 I 1 -0.3% I t 1 1 1 1 I 1960 1965 1970 1975 1980 1985 SOURCE: Organization for Economic Cooperation and Development, Department of Economics and Statistics, NationalAccounts, various issues. and conversely, the falling dollar in the late 1980s for the improvement in the trade accounts. But over the long run, adding currency value into the picture simply accentuates the difference in competitive performance of American and Japanese manufacturers; American manufactured goods have lost world market share in spite of the fact that the dollar has been on a long-term decline against a trade-weighted
PAGE 14
Chapter 1--Summary and Findings l 7 average of foreign currencies. Japans manufacturers have increased their market shares even with a rising yen. American manufacturers are aware of their competitiveness problems; many have made commendable efforts to improve their performance. But particularly in sectors that contribute heavily to employment, trade, knowledge, and income, there is still a gulf between Japanese and American company performance. In many cases, the gap is widening, driven by the fact that Japanese companies, flush with the profits of their market success, are investing more heavily in technological improvement and global expansion. If there are no major changes in government policies of developed nations, we expect U.S. manufacturing competitiveness to continue to sink, compared with Japan. There will be more emerging technologies in which the dominant power is Japan, not the United States, and established industries will remain behind the Japanese world leaders. This situation also faces producers in Western Europe, who are likewise behind the Japanese (and in some cases American) manufacturers in a variety of important sectors. The difference between the United States and Europe, at this point, is that European governments are taking an active role in trying to bolster their competitiveness, while the U.S. Government takes the position that the best aid to competitiveness is a free market. The Role of Competition: Are Free Markets Always Best? Market freedom is a relative thing. By the standards of economics, there is no free market in the world today, and there never was. There are, on the other hand, quite a number of economies whose markets are thought of as free because large segments of the economy are shaped at least as much by market forces as by policy and regulation. The United States likes to think of itself as one of the most free markets, and it probably is. That, plus the fact that the United States has been the dominant economic power of the world throughout the postwar period, is often taken as proof that the market is superior to government planning as the way to economic prosperity. The recent events in Eastern Europe, with the demonstrated failure of heavy state pl anning, are regarded as additional conflation. In fact, both government and the market are, in the words of one eminent economist, instruments of social policy. 7 They are different ways of arranging the activities of production and distribution in the economy; neither is clearly superior to the other. We do not really know what the economy would look like if markets were not limited and constrained in countless ways by government regulation, but one hypothesis is this: The only industry that knows no bounds or rules is the illicit drug trade, where the market process resembles what Thomas Hobbes called the Warre of Each Against. He described life under those conditions as Nasty, Brutish and Short. 8 The miserable record of the command and control economies of Eastern Europe, compared with the record of the U.S. and Western European economies, does not imply wholesale superiority of market forces under all conditions. Most of the restraints the U.S. and other governments impose on markets exist because the market serves some interests (e.g., long-term values of society or provision of social goods like scientific knowledge, clean air, and safety) very poorly. If we examine the difference between the performance of Japan and the United States, it is appropriate to suspect that more competition is not always better. The government of Japan has never been as sold on the tenets of neoclassical economics as the government of the United States. Japan and the fast developing East Asian economies of South Korea and Taiwan have restrained and shaped competition at various points in modern history and have benefited as a result. 9 One of the most visible forms of restraint on free markets is the market protection that all three countries have used extensively to nurture infant industries and, in a few cases, to permit orderly rationalization of mature ones. 10 Japan employs government intervention to harness, channel, or restrain the forces of competition. During the 1960s, the government coped with what it called excessive competition by organizing antirecession cartels and vigorous export campaigns, erecting barriers to foreign direct investment, and structuring the financial system to favor industrial investments (especially in targeted industries) over consumption. 11 These measures were aimed at both restraining competition from more advanced foreign firms and restricting the cutthroat investment and price competition among Japanese firms. 12 Starting in the 1980s, the government still reined in competitive forces, but with different
PAGE 15
8 l Competing Economies : America, Europe, and the Pacific Rim measures (e.g., voluntary restraints on exports and bargains to increase foreign fins access to the Japanese market) and for a different purpose: to soften foreign, often U. S., complaints of unfair competition. Although the forms and aims are new, the effect is still an interference with the workings of the market. The Japanese Government has never trusted the market to achieve, by itself, large-scale investments for basic research in high technology, or to overcome the disadvantages of being behind. 13 In this, Japan is hardly alone. Few developing nations trust market signals alone to generate the investments and provide the resources necessary to improve industrial development and living standards. Korea and Taiwan, too, have altered market signals and outcomes significantly. Both share a commitment to long-term planning, industrial targeting, or strategic visionsforcing their firms to compete on world markets and n urturing them at home. Taiwan, while more open to foreign trade than Korea, and with less interventionist industrial policies, has relied more on public enterprises. Korea has been more protected from foreign competition, with the protection tied to export expansion. Both countries depend on government policies to promote an indigenous technology base. (See ch. 7 for discussion of the industrial policies of Korea and Taiwan.) Japan, Korea, and Taiwan all developed exceptionally fast, with industrial policies that signifi cantly altered, but did not destroy, market signals. Protection of the domestic market and direct funding of R&D were forms of intervention, as were policies to steer low-cost capital, preferential access to foreign exchange, assistance in negotiations with foreign companies for access to technologies, and support of domestic technology development and implementation through a variety of fiscal incentives. At the same time, these governments were careful to maintain incentives that forced domestic firms to compete with the dominant foreigners, often in third markets or in the home markets of the foreign competitors. Market competition, in short, was viewed as having a proper place; it was superior to government plannin g in providing incentives to improve productivity and quality and reduce costs. But unbridled competition was not, and still is not, regarded as always yielding the best possible outcome for the nation. The prevailing ideology in the United States is very different. We have accepted that market competition will not secure the outcome we want in some areas. For example, market incentives do not provide sufficient incentives to invest in as much R&D as would be optimal for the nation, nor do they provide incentives to preserve such public goods as clean air and water. But the United States views anything less than free trade as dangerous interference. This is consistent with the ideology of the great European powers when they were the highest value, lowest cost, most advanced producers of their day; it is nothing new for the prevailing economic view to coincide with commercial interests. 14 But the fit between the two, never perfect, is growing more uncomfortable as U.S. industries watch foreign competitors, usually Japanese, attain technological dominance in many important industries. More U.S. industries are finding themselves in the position of having to catch up, while learning that intelligent responses to market signals are insufficient to close the gap between them and the market leaders. A case in point is capital cost. Throughout most of the postwar period, as a result of government control of financial markets, Japanese and German companies have enjoyed much lower capital costs for plant and equipment acquisition and R&D than American fins. As a result, they have invested more in R&D, plants, and equipment, and been more patient in recouping those investments. The responses have been especially strong in industries that demand high investments in R&D and capital equipment. In Japan, special tax measures, such as accelerated depreciation, have sweetened the investment incentives in particular industries chosen by the government as strategic. As a result of these and other measures, Japanese firms invested more in technology development and adoption than American firms and have advanced faster and gained greater market shares. 15 What the Japanese Government, and more recently the Korean and Taiwanese Governments, has done is to use a combination of market signals and government planning to speed economic development and growth. At times, the interventions have backfired, but overall, the policies of the Japanese, Korean, and Taiwanese Governments have been essential to fast development. They are not the only contributors, as shown by the failure of similar government policies to lift dozens of other less
PAGE 16
Chapter 1--Summary and Findings l 9 developed countries out of poverty. What sets these East Asian nations apart? Some of the answer, of course, is that they have relied heavily on market forces to shape the behavior and strategy of businesses. Many developing and some developed nations have erred in the direction of overprotection, and protected sectors have failed to become competitive with world leaders. The successes of the East Asian nations lie in their combination of import substitution and market protection with export promotion, which allowed domestic industries some potential for output growth and access to needed equipment and components while forcing them to compete with the best performers in the world. l6 The governments were able to force or influence firms to conform with public policies through a variety of disciplinary measures. Firms that failed to improve export performance, for example, might have trouble getting the necessary permission and foreign exchange to import needed equipment. Another part of the explanation is that the companies crafted intelligent, patient strategies for developing, producing, and marketing their products. Some of this strategic genius is attributable to government policies and some to good business strategies, but another part is sheer diligence; East Asian companies are famous for scouting and adopting the best of the strategies of other successful companies and countries. Policies and cultures that emphasized very high product quality and universal education also helped, as did ethics that valued hard work. Industrial policies were not the whole answer, but the path of development would have been slower and rockier without them. Industrial Targeting To Americans, industrial targeting is one of the most controversial aspects of industrial policy. The idea that some industries contribute more to national well-being and knowledge than others is not particularly contentious. It is other things in targeting that we fear. One is that government will not be rational in its choice of industries to support and that the process will end up being hijacked by special interests. Another objection is that the market is better suited than government to choose the industries that make disproportionately large contributions to national welfare. Most developed nations and many developing ones do not share that faith in the market. The Japanese Government considers certain industries crucial to its economic health. Immediately after the war, policymakers felt that Japan should be strong in manufacturing iron and steel, ships, machinery, heavy electrical equipment, and chemicals. Later, the automobile, petrochemical, nuclear power, computer and semiconductor, and aircraft industries were added to the list. 17 Though less so than in the 1950s and 1960s, Japan continues to provide particular benefits to targeted industries and the users of their outputs. 18 Korea and Taiwan, too, selected industries for special support. In the 1950s, Korean industrial policy focused on import substitution in light manufacturing. In the late 1960s, emphasis shifted to steel and nonferrous metals, chemicals and petrochemicals, machinery, automobiles, and shipbuilding. Still later came emphasis on semiconductors, computers, and consumer electronics. The pattern in Taiwan was similar, emphasizing light manufacturing and import substitution in the first two postwar decades, shifting to heavy industries in the 1960s and later into more technology-intensive sectors. The tools and methods were different, but the selection of industries was similar. Industrial targeting is not limited to developing countries. Most of the nations of the European Community (EC) have long had policies of supporting European producers of motor vehicles, telecommunications equipment, semiconductors, consumer electronics, and aircraft. While now discouraging support for national champions by individual countries, the ECs plans for the single market in 1992 and the Framework Program and EUREKA 19 are aimed at developing technological and productive prowess in many of the same sectors. Electronics and telecommunications sectors receive greatest emphasis and heaviest support in European R&D programs, while EC trade policies are being structured to cushion the European automakers encounters with Japanese producers. Even in the United States, which has mostly rejected the use of policies designed specifically to improve civilian industrial competitiveness, 20 there are a few examples of industrial support for nonmilitary purposes. One of the most prominent is the National Aeronautics and Space Administration (NASA), which has, as part of its mission, the responsibility of improving aircraft technology .21 NASAs Aeronautics Program budget in 1991 was
PAGE 17
10 l Competing Economies: America, Europe, and the Pacific Rim almost $920 million; 22 in 1992, the budget is expected to exceed $1 billion. 23 Though this is partly intended to support military technologies, NASA officials state that 90 percent of the technology developed is common to both military and civilian needs. That industrial targeting exists is not a question. But can the U.S. Government select the right industries for support? To detractors, the idea of picking winners is dangerous because it opens the possibility that any industry could be selected if it had a powerful enough lobby. Another potential drawback is the argument raised by many economists, that industrial supports could skew investment incentives and create excess capacity, and thus increase vulnerability to business cycles in targeted industries. This, in turn, could make additional protective measures necessary-beginning a cycle of ever-widening support, at increasing cost to taxpayers and diminishing benefit to consumers. These are real problems and should be taken seriously, but they are not the inevitable outcome of any exercise in supporting the competitiveness of critical industries. Japan, Korea, and Taiwan have had some problems resulting from their support of targeted industries, but their overaIl economic performance-and, with few exceptions, the performance of the targeted sectorshas been better than that of the United States. Some argue that the success of these economies, and of targeted industries, is more coincidental with industrial policy than causally related. But after weighing the evidence, OTA concludes that the industrial policies of these East Asian countries is crucial in their economic performance. Japanese industrial policy aimed at changing Japans industrial structure from one characterized by labor-intensive industries immediately after the war to one dominated by capital-intensive industries in the 1950s and 1960s, and from that to one led by knowledge-intensive industries from the 1970s to the present. At each stage of policy formulation, industries selected for special support were those that made disproportionately large contributions to national well-being. 24 They had high growth prospects, higher wages, and the possibility of higher profits, with more positive spillovers to other sectors in terms of contributions to technology and knowledge. In some cases, the spillovers were downstream: microelectronics and computers add to the technology intensity and productivity of industries that employ them. In the case of automobiles, primarily a consumer product, the contributions to other industries came primarily upstream, in the machinery industry. The idea that certain sectors can be identified as incubators of larger economic change is gaining currency in the United States as well. An economy with strength in these industries will have higher wages, faster growth, and better developed processes of generating, diffusing, and using new technology than economies that do not. It is no coincidence that many of these industries-e. g., semiconductors, computers, telecommunications equipment, aircraft and spacecraft, and advanced materialsalso present formidable barriers to entry. Capital requirements for efficient production are often quite high, and the requisite knowledge of science, technology, and production is even more forbidding. If government can help potential entrants overcome entry barriers, the whole economy can be put on a path of faster growth and higher incomes. With most of the governments of large developed economies providing some kind of support for critical industries, the market signals American firms get for these same industries often point in the opposite direction--down. In many critical, hightechnology sectors, American firms are facing competitors whose business risks are shared by their governments (and thereby, the taxpayers of their home nations). Faced with the necessity of assuming most of the risks and costs of entering or even maint aining operations under such conditions, some eventually abandon operations, as have many American manufacturers of memory chips. Can the U.S. Government afford to be indifferent to the mix of industrial goods produced here? It would be dangerous, and in many ways unprecedented, to adopt this course. While we have let a few industries, such as consumer electronics, wither in the face of superior foreign competition, the government has stepped in many times to support industries that served various economic needs, including employment (the Chrysler bailout), technology development (aircraft in the 1920s and 1930s), and national/economic security (semiconductors). This is not meant to imply that every effort to develop or support an industry has been successful; every country that has tried industrial support has made some errors. In many developing nations, the whole enterprise of supporting industrial development has
PAGE 18
Chapter l-Summary and Findings l 11 been disastrous; government planning cannot replace the market and the forces of competition. But the signals the market sends to American firms in the majority of high-technology, economically critical industries are not promising if we want those industries to thrive financially and technologically. It is equally dangerous to trust ad hoc, often crisis-motivated, efforts to intervene in industrial structure and performance and expect the results to make consistently positive contributions to economic well-being. Import Dependence One of the biggest problems in sorting out how to treat critical industries in public policy is choosing among them. All nations do not have to, and indeed, cannot, be competitive in all high-technology industries or critical sectors, if only because some, such as aircraft, will not support more than a few competitors without massive government assistance. Moreover, there are many sectors regarded as critical, and self-sufficiency in all of them may be beyond the means of any single nation, or it might spread available resources for support too thinly to have an impact. Finally, and most importantly, selfsufficiency by-passes the benefits of specialization and trade. It should do a nation no harm to import certain critical products and export others. This is how Germany manages, and the Germans have the highest standards of living and most competitive manufacturing in Europe. Manufactured imports accounted for 14.4 percent of German GDP in 1987, and for nearly 45 percent of German manufacturing GDP. In the United States, manufactured imports were 7.3 percent of GDP and 37.8 percent of manufacturing GDP; corresponding figures for Japan (which is an outlier among both industrialized and industrializing nations) are 2.4 and 8.3 percent. 25 This kind of evidence often leads some to question why the United States should care about depending on foreign manufacturers, even for key inputs. In fact, competitiveness might be improved as a result; computer manufacturers could be more competitive if they have access to low-cost foreign semiconductors than if they have to pay higher prices for domestically made ones. Certainly, after the 1986 Semiconductor Trade Agreement (STA) with Japan resulted in (though did not necessarily cause) high prices for 256K DRAMs, American computer makers suffered declines in profitability, and Japanese computer makers improved market positions and profits. 26 In fact, the fallout of the STA is often used to support the argument that trade protection hurts more than it helps. Whether that is true, even in the one case of the STA, is disputed. The STA was signed after several years of mounting disputes with Japan over the fairness of their trade in semiconductors. 27 American producers alleged that Japanese semiconductors were dumped, both in the United States and in third country markets, and that American semiconductor chips were unfairly excluded from Japanese markets. After formal investigation, the International Trade Administration found that dumping charges were substantiated. The STA stipulated that both practices (dumping and exclusion) should stop. 28 American officials apparently hoped for substantial cutbacks in Japanese production as a means to raise prices of Japanese semiconductors, thus ending the dumping. Shortly after the STA was signed, MITI (the Japanese Ministry of International Trade and Industry) took to issuing forecasts of chip production that were widely interpreted, here and in Japan, as administrative guidance to cut production. The companies, however, were initially unresponsive, and the U.S. Government announced sanctions in early 1987. Others dispute that the STA caused Japanese producers to raise prices. 29 According to the counter argument, joint action by Japanese producers was underway for at least a year before the STA was concluded, and had begun before the STA was even a topic of discussion. The collapse of DRAM prices in 1984-85 that precipitated the withdrawal of many non-Japanese producers from the market also cost Japanese producers billions of dollars. By late 1985, the market was dominated by Japanese producers. Newspaper accounts about Japanese companies taking joint action to raise DRAM prices began to appear in 1985, as did rumors of meetings of Japanese producers aimed at addressing the price disaster. These stories and rumors are made more believable by the many instances of coordinated Japanese industry/government management of prices and production that occurred in other industries (e.g., iron and steel) in the past. Both sides of the story agree on one point: that MITIs forecasts and guideposts are used as targets for production and/or prices. The difference arises
PAGE 19
12 l Competing Economies: Arnerica, Europe, and the Pacific Rim over whether the production cutbacks and price increases, which both sides agree were deliberate, were a result of the STA. Without a formal investigation, the dispute will likely remain unresolved. Whether or not U.S. sanctions were an important cause, the fact remains that by early 1988 prices of Japanese DRAM chips rose, and production fell, a sharp break with past price behavior. While MITI denies any official guidance, its continued quarterly forecasts and their pinpoint accuracy strongly suggest to some analysts that MITI was controlling production and, possibly more significantly, investment in new facilities. Ordinarily, sustained high profits brought on by elevated prices would be expected to generate a wave of investment in new capacity, which, as of mid-1989, had not materialized, in contrast to the past and contrary to what analysts expected. Some of the restraint in investment in new capacity was attributed to quiet guidance by MITI, and some to a new spirit of cooperation among the Japanese DRAM manufacturers. Both can be attributed in part to fears of reigniting trade disputes with the United States, but there was something else at work, too. The Japanese producers, probably with MITIs encouragement, had begun to act like a cartel, controlling output and prices and reaping higher profits as a result. While that was a predictable outcome of the U.S.-initiated STA and accompanying sanctions, it is one of the more ominous developments on the trade scene, and it symbolizes one of the things most feared about dependence on foreign suppliers. When suppliers act collusively to manipulate production and prices, everyone else is stuck paying higher prices. If the products of the cartels are inputs to key industries, whole economies can suffer. No nation is more aware of this than Japan, which suffered a severe economic downturn after the first oil shock in 1974. But how much real danger is there? Cartels that can effectively manipulate supplies and prices globally are not very common. It is tempting to think that if we dont meddle with free trade we wont have a problem. But this is probably wishful thinking. Japan maintains several legal cartels, and while the number is diminishing the practice is familiar. Moreover, the historical links between the semiconductor companies and the Japanese Government mean that under ordinary circumstances they share a degree of knowledge of each others plans and behavior that would be considered extraordinary, probably collusive, in the United States. Even without the prodding of the STA, it is conceivable that Japanese companies might have begun to exercise restraint over production and prices. Two pieces of evidence support the notion that these producers were amenable to such restraints. One is the fact that similar price rises did not occur in EPROMs (erasable progr ammable read-only memories), another semiconductor product covered in the STA. Unlike the situation in DRAMs, production of EPROMs was not dominated by a few Japanese companies. Another is the difference in the degree of compliance with STA in two areas: dumping and foreign companies share of the Japanese market. Cutbacks in production resulted in price increases that eliminated dumping in fairly short order, while the share of foreign semiconductors in Japans market has increased much more slowly than originally called for. 30 It took much less time and effort to arrange production cutbacks, which resulted in higher profits for Japanese companies, than increased market share for foreign companies, which gave Japanese companies no particular benefits. Another problem created by dependence on one or a few suppliers for critical components is access to the latest technologies. Again, under the conditions of competition envisioned in economics, a supplier of critical components would be foolish to deny the most advanced products to any customer, since a competitor is always ready to do it. For example, the worlds industrial producers of textiles depend on machinery from a handful of suppliers in a few countries-Germany, Switzerland, Japan, Italy, and Sweden account for most of it. Yet textile manufacturers are routinely able to get the most advanced equipment from any supplier; textile makers in the countries that manufacture textile production equipment have no particular advantage. The story is different in electronics. American and European systems makers, or policymakers speaking on their behalf, are concerned that it may be difficult to get the most advanced production equipment or chips from Japanese vendors. In most cases, those Japanese vendors are also systems makers, or else they have close ties through the keiretsu 31 system with a Japanese systems maker. It is a logical, and probably not uncommon, business practice to reserve access to the latest technologies to users within the develop-
PAGE 20
Chapter I-Summary and Findings l 13 ing firm or to special customers. According to anecdotes, when IBM and Perkin-Elmer had a special relationship, IBM got first access to new machinery developed by Perkin-Elmer. While no blame attaches to Japanese companies that give themselves or their closest customers first crack at new generations of technology, the competitive advantage it confers on Japanese firms can be significant in the fast-moving electronics industry. Dependence on imports, particularly when those imports come from only a few suppliers, and particularly when those suppliers are also competitors, can create vulnerabilities that nations sometimes choose to avoid. Four European countries, for example, continue to provide financial support to Airbus in part because they do not wish to depend on two American companies, Boeing and McDonnell Douglas, for all large commercial jet transports. While Airbus is a financial drain on its government sponsors, it has forced the American producers to compete on price to a greater extent. Probably the foremost example of unwillingness to rely on foreign suppliers is Japan, which has a long tradition of limiting dependence on imports to a practical minimum. 32 While Japan is criticized for its peculiarly strong aversion to imports, and is slowly changing, many of Japans premier industries probably would not have developed, or would have developed far more slowly, had it permitted much greater imports and investment when Japanese industries were catching up. Now, with an increasing number of American industries in the position of latecomer, the vulnerabilities created by import dependence have assumed more importance. SUMMARY OF POLICY ISSUES AND OPTIONS The idea that troubled American manufacturing industries could use help from their government is gaining acceptance. Although the form such aid should take is not a settled question, the areas of agreement are widening. Agreement embraces more than the traditional areas for government action: macroeconomic policies that create stability and lower the Federal deficit, and human resource policies that produce the well-educated and welltrained workforce that American industry needs. Beyond this, a consensus is forming for more focused government policies to help industry develop and adopt technologies that can boost the competitiveness of U.S. manufacturing. Among these are policies to diffuse technologies throughout manufacturing, as with, for example, governmentfunded technology extension services. The area of agreement is expanding to include R&D partnerships between government and industry to develop high-risk technologies of generic commercial interest. Congress created a small program of this sort in 1988, the Advanced Technology Program in the Department of Commerce, and first funded it in 1990. The Program has now gained the backing of the Bush administration. More generally, the Presidents 1992 Budget endorsed government support for generic or enabling technologies at the pre-competitive stage of R&D. It said: The Administration believes that appropriate Federal investments in applied civilian R&D can result in high payoff to the economy. 33 Helpful as all these policies can be in restoring U.S. competitiveness, in some critical cases they are not enough. Certain industries characterized by expanding markets, good jobs, increasing returns to scale, and technological spillovers to other industries are so essential to the Nations economic growth that standards of living will suffer without them. Computers and electronic components are the clearest example. Important parts of these U.S. industries are in trouble, and it is doubtful that they can regain their competitive edge without much more substantial technology assistance than that provided by Sematech. Japanese firms are so large, so adept, and so dominant that, without trade technology, manpower, and other policy help from the U.S. Government, the U.S. electronics industry probably will continue to decline in comparison. The help from the STA was on an ad hoc basis that did not signal to the industry any sustained government commitment to the industry, and therefore it did little to encourage investment in technological improvement. While there may be times when trade policy is a necessary complement to other policies to improve competitiveness, it is emphatically not a stand-alone fix. By itself, trade policy will do little to restore competitiveness and can have deleterious effects on downstream producers and consumers. It is much more likely that policies to encourage technology development and diffusion will be helpful, even without relief from foreign competition, than that trade policy alone will improve competitiveness.
PAGE 21
14 l Competing Economies: America, Europe, and the Pacific Rim An approach that combines trade policy with technology assistance, and sometimes financial assistance as well, in support of critical commercial industries could be proactive and planned to avoid pitfalls. Any trade protection involved could be framed in a frank and self-respecting way as guarding important American national interests, rather than based on findings of unfair trade by others (as is now the case with most U.S. trade protection). This would skirt three problems that bedevil current U.S. trade policy. First, it is difficult and time-consuming to make charges of unfair trade stick. Second, unfair trade is often only a minor reason for a U.S. industrys decline, secondary to the ability of foreign firms (often aided by industrial policies) to lower production costs, acquire new technologies, and make genuinely superior products at a good price. Finally, whatever the mix of causes, the moralizing tone in our present trade policy is an irritant to our trading partners, and sometimes makes it harder to find reasonable solutions to trade disputes. The difficulties of creating a coherent government strategy for supporting critical commercial industries should not be underrated. There is merit in the argument that our form of government is open to capture by special interests, so that the chances of getting rational, disinterested government decisions on industry support are slim. Experience suggests, however, that a contrary argument also has merit. The United States does employ some trade protection and does sometimes subsidize certain economic sectors (e.g., farm price supports). These departures from our free trade and free market philosophy are often politically motivated. It is possible that a coherent strategy to select a few industries for their contribution to the national good and support them with a tailored mix of technology, financial, and trade assistance would result in more rational exceptions to a general rule of free trade. A related problem is that the U.S. Government does not have the experience or institutional capacity to operate a strategic industry and trade policy. Ideally, we would need an institution capable of identifying critical industries, analyzing their needs, and planning measures to fit the needs while taking care to keep the government support modest, make industry a full partner, and foster competition among firms within the critical industries. A tall order, and one that could only be filled over time. The rule would be to start small, gradually building expertise and a spirit of mutual trust with industry. Another condition for a strategy in support of critical commercial industries is commitment from both Congress and the Administration, backed by wholehearted support from the American people. Congress has taken the lead in recent years toward giving some support to critical industries (e.g., in creating and funding Sematech and the Advanced Technology Program). The Administration also now supports a government role in developing generic technologies of commercial interest. It is hard to imagine, however, that Congress could unilaterally create broader industry and trade policies in support of critical industries and the institutions to carry them out. For the policies to work, both branches of government must be committed to them, and that commitment must rest on the understanding and support of the American people. There would have to be widespread comprehension that U.S. economic security is at risk, agreement that governmentindustry partnerships for improved competitiveness can pay off, and acceptance that short-term costs are worth paying in return for long-term gains in restoring excellence to U.S. manufacturing. While the obstacles to crafting an effective strategy to support competitiveness are great, so are the payoffs. Improved competitiveness can come only from improvements in productivity and technology; these, in turn, can support higher standards of living for most Americans. This not only gives individuals more choices and comfort in their own lives, it also increases U.S. resources to do things that only rich nations are equipped to do: fight poverty and illness here and abroad, protect environmental amenities, expand the frontiers of science. The policy issues and options discussed below include: building new institutions to plan and implement a government strategy in support of critical commercial industries, fostering a supportive environment for technology development and adoption, altering trade policies so they are more effective and more attuned to competitiveness needs, and formin g government-industry partnerships for technology development and low-cost production.
PAGE 22
Chapter 1--Summary and Findings l 15 Institutions for a Strategic Competitiveness Policy Actions and Policy Tools A necessary first task for a government agency or commission given responsibility for industry and trade strategies is to identify the critical commercial industries that will be the focus of policy attention. This selection does not mean that other industry will be ignored; an economic and policy environment that supports industrial advance across the board is also necessary. But the concept of a critical industries policy is that some industries contribute more to continued technological advances and rising prosperity than others. Most advanced countries do lend policy support to certain industries, and the criteria for selection are generally the same; they favor industries that are knowledge intensive (with a high proportion of technical workers and high R&D), that have good prospects for growing markets, and that are built on versatile core technologies with spillovers to other industries. Another principle arguing for government involvement is high barriers to entry arising from exceptional capital costs, large economies of scale, and a steep learning curve, so that late entrants are at a great disadvantage. These principles usually lead to selection of much the same industries, including electronic components, computers and software, communication equipment, precision machining equipment, advanced materials, robotics, biotechnology, and aerospace. Governmental and private bodies in several countries have made lists of emerging technologies, around which new critical industries (or advanced versions of existing industries) might coalesce. These lists too have many similarities-including the assessment that the United States will lag behind its major competitors, especially Japan, if current trends continue. 34 The specific policy tools available to the agency cover trade, technology, and financial areas, and are discussed below. Its overall responsibility would be coherence and coordination. It could be given the duty not only to develop critical industry strategies but also to champion them throughout government, reminding and urging other agencies to give critical industries priority in their decisionmaking. Congress might wish to emphasize this function by requiring Federal agencies to prepare a competitiveness impact evaluation before taking major actions affecting critical industries. The extra paperwork involved could be well worth its cost if it served as a constant reminder of the potential impact of government policies on industrial competitiveness. Finally, interaction with industry is a must. A government agency with strategic trade and industry responsibilities should have some independent knowledge and goals, but it must also work in alliance with industry-which would be expected not only to help shape the strategic plans, but also to put up at least half the funds in any venture where the gov ernment provides financial backing. Alternative Institutions Institutions that might develop and implement government policies to support critical commercial industries could take several forms. A certain amount of restructuring and reorganizing of current Federal functions would help. But it takes a lead agency to plan and carry out a coherent strategy in which high reward industries are selected for particular attention and in which elements of technology, trade, and financial policies are combined as needed. Industry-specific advisory committees, established by Congress, might be one way to begin. There is a precedent for this. In the 1988 trade act, Congress created the National Advisory Committee on Semiconductors (NACS), made up of industry and Federal Government leaders, to devise a strategy for strengthening the U.S. semiconductor industry. The Committee has issued two interim reports and will publish a final one by the end of 1991, laying out a comprehensive strategy. This is just a frost step, however. No agency has responsibility for coordinating and carrying out the NACS-recommended strategy, and there is certainly no guarantee that it will do more than gather dust. An existing agency with the potential for combining strategy development with action is the Competitiveness Policy Council, created by Congress in 1988 and launched in 1991. 35 With members appointed both by the President and by leaders from both parties and both Houses of Congress, the Council is structured to take a bipartisan approach that could be effective. However, it would need broader powers and a longer life than it has now. Its present duties are only to develop recommendations for greater competitiveness, and unless continued by
PAGE 23
16 l Competing Economies: America, Europe, and the Pacific Rim Congress, it will go out of existence in 1992. It might be turned into a commission, given at least a 5-year life, and directed explicitly to identify critical commercial industries and strategies to support them. It would also have to be given some power to implement the strategies if it is to have a real impact. Another possibility is to lodge the responsibility for industrial strategies and trade policy in a small office in the executive branch, possibly in the Executive Office of the President. That location would be a good position from which to remind and encourage other agencies to consider effects on competitiveness in all their major decisions. This could only work, however, if the office is seen as truly competent and well-informed, with close interaction with industry and a staff of exceptional people. And it could have little effect on other government agencies unless it had strong Presidential backing. Major government reorganization is not required, but it could help. If Congress wishes to adopt a more proactive approach to trade and competitiveness issues, a reorganization could produce more focused policymaking and more direct lines of authority. In trade matters, for example, the U.S. Trade Representative (USTR) is charged with coordinating all relevant government agencies to formulate trade policy, and then with negotiating from that position. However, the USTRs own staff is thin and lacks continuity; it can do little more than concentrate on current issues. Yet responsibility for a durable, strategic approach to trade policies that guard basic American interests is lodged there. Much the same is true of policies that influence industrial competitiveness. The absence of a strong voice in government for international competitiveness just about guarantees that other objectives (e.g., foreign policy, national security) will win in a dispute. Several bills in Congress have proposed a reorganization to focus Federal trade and competitiveness policy functions. Some would establish a Department of International Trade and Industry, assuming many of the functions of the Department of Commerce and the USTR. 36 Some concentrate on trade; they would set up a department that consolidates USTR and the trade policy units from several departments and would establish a Cabinet committee to coordinate international economic policy. Others focus on technology and industry. There have been several bills to create a Department of Industry and Technology, expanding the Commerce Departments export promotion authority and creating a Civilian Technology Agency (CTA); 37 some of these bills would also create an independent U.S. Trade Administration, consolidating the USTR and the Commerce Departments trade agencies. Whatever bureaucratic arrangement is chosen matters less than the substance of the strategic policies and the commitment of both the Admin stration and Congress. No arrangement will solve all coordination problems; there are always competing government objectives related to trade and industrial competitiveness. And no arrangement will create a U.S. equivalent of Japans powerhouse Ministry of International Trade and Industry. What it could do is make possible a modest start in pulling together policy strands that would promote critical industries and our national economic welfare. A Technology-Friendly Environment Some of the most important options to help critical commercial industries perform better could also improve the competitiveness of all American industry, across the board. These are options to create a hospitable environment for the development and adoption of new technologies generally, throughout manufacturing, and they merit close consideration whether or not more targeted efforts are undertaken to nurture particular industries. OTAs earlier report, Making Things Better: Competing in Manufacturing, considered in detail options to help manufacturers improve their performance through better use of technology. Defined broadly, technology includes not only new products and advanced production machinery, but also efficient organization of work and effective use of people. Industry and government both have parts to play in building a better technology base for U.S. manufacturing. The report defined four areas in which government could usefully contribute: l l l l Improving the financial environment for U.S. fins, which means taking action to reduce capital costs and relieve other pressures to show high profits every quarter. Upgrading the education and training of the managers, engineers, technicians, and workers needed in manufacturing. Diffusing technologies throughout the manufacturing sector. Forming a strategic technology policy to promote the development of new technologies
PAGE 24
Chapter I-Summary and Findings l 17 with commercial promise through collaborative ventures with industry. Options in the first three categories outlined above would benefit all U.S. manufacturing with no distinctions among industries. The fourth is for a more targeted effort, and is discussed in a later section. Most of the options s ummarized in this section are analyzed in greater detail in two recent OTA reports, Making Things Better and Worker Training, as well as in chapter 2 of this report. The Financial Environment The U.S. financial environment is not hospitable to long-term investment in new technologies and production equipment. High capital costs favor taking short-term profits rather than investing for the longer run, as do pressures from the stock market. Capital costs are affected by several factors, including interest rates, the economic depreciation of investment and its tax treatment, and other fiscal incentives for investment. Recent studies that take all these factors into account provide solid evidence that U.S. capital costs have been substantially higher than those of Japan and Germany for more than a decade, through 1988. 38 Moreover, the terms on which capital is made available are more favorable to long-term investment in both Japan and Germany. An example is the stable shareholding system in Japan, in which the majority of shares in large corporations are held by either companies in the same group or by stable shareholders, and these companies do not trade their shares. In the United States, government policy has contributed damagingly to high capital costs. The main culprit at present is the huge, accumulating Federal budget deficit, which puts upward pressure on interest rates. Also, the U.S. tax system has many fewer incentives for productivity-enhancing investment in manufacturing than those of our competitors, especially Japan. The dilemma is that some specific fiscal measures that might help firms modernize and invest in new technologies would also tend to worsen the budget deficit, because they would lower revenues, at least in the short run. The budget agreement of 1990 forbids this unless there is a compensating rise in tax revenues or decline in spending elsewhere in the same segment of the budget (nondefense domestic programs). If Congress wishes to lower capital costs through tax breaks, it will need to find something else to cut, or get agreement to raise taxes in compensation. Ultimately, economic growth based on better competitive performance could ease budget problems, but in the short run there will be a price to pay. Another choice is to increase savings, and thus ease pressure on interest rates. Options include a national savings campaign, with appeals to patriotism plus attractive interest rates for regular payroll savers, or perhaps a consumption tax designed to escape the severe regressive effects of a flat tax. Another option is further restrictions on deductions for home mortgages. Tax breaks to industry, such as accelerated depreciation for investment in new equipment or a tax credit for R&D, have shown positive effects in the past, though the exact size is debatable. Also, they are expensive--especially accelerated depreciation, which can cost the U.S. Treasury tens of billions per year; whatever they cost would have to be made up elsewhere. Incentives to hold investments longer might relieve some of the pressure to focus on short-term profits. One option would be a capital gains tax that favors long-term gains and penalizes short-term turnover. It would be most effective if applied to pensions and other funds that are now tax-free, since these funds hold one-third of all stocks and probably account for half of transactions on the stock market. Human Resources Success in manufacturing depends on having well-trained people, comfortable with the demands of advanced technology, at every level from the managers office to the shop floor. The failures of our public schools in turning out well-educated young people with good work habits is well-known. Unfortunately, training of adult workers in the United States is deficient compared with that in several other countries, in particular Japan and Germany. The quickest payoff may be in improved training of the active workforce, since improvements in schooling take many years to show up on the job. An aggressive, far-reaching option, which guarantees more training without any direct cost to the government, is a payroll-based training levy. Employers would have a choice of spending a certain amount on training their workers or paying the same amount into a national training fund. Several foreign countries, including France, Germany, Ireland, and the Republic of Korea, use the system. Government
PAGE 25
18 l Competing Economies: America, Europe, and the Pacific Rim might also offer technical assistance on training needs and best practice training methods to trade associations, labor-management groups, or industrial consortia. Restoration and improvement of formal apprenticeships is another option for government-industry partnership. And the Federal vocational education program has many useful features, including school-to-work programs, but they lack adequate funding. Training linked with technology assistance is effective, and a few States provide it. However, technology extension services are scarce and spotty in the United States, and the link with training is scarcer still. The National Institute of Standards and Technology (NIST) is in charge of the Federal Governments modest technology extension effort; it could add a training component. The Federal Government could also take a more active hand in testing and evaluating computer-aided training technologies, including adapting and transferring instructional technologies developed for the armed forces. Because education and trainin g of engineers is a central competitiveness issue, the Federal Government could also be more active in this regard. While the supply of engineers seems adequate now, it could dry up in future years as the proportion of white males in the work force declines; white males predominate in engineering. This is essentially a problem to be solved in public school education; attitudes to math and science are formed early. Meanwhile Federal grants to women and minorities to encourage science and engineering careers seem to get results, and deserve support. Retraining of midcareer engineers is another way to enlarge the supply over the next few years, especially at a time when engineers are losing jobs in defense industries. Technology Diffusion U.S. institutions for diffusing new technologies throughout manufacturing are thin. Even large firms with the resources to develop or acquire the technologies they need often neglect to take what they could from outside the firm. Many of our 350,000 small and medium-sized manufacturing firms are worse off, with only scant exposure to new technologies. Technology extension services funded by government could improve the manufacturing of small and medium-sized firms, but so far it is more potential than actuality. Defining industrial extension as one-on-one technical advice given to individual fins, 16 States had such programs in early 1991, and another 7 had technology demonstration or assistance centers. Spending by the 23 States for 27 centers amounted to about $50 million. A small Federal program was established under NIST administration in 1988; it now includes five centers, with one more planned, each with Federal funding of about $1.5 million per year and an equal amount from State, local, and private sources. For perspective, compare these scattered programs with technology extension in Japan. Besides the nationwide system of 185 technology extension centers, funded at about $500 million, half from the national government and half from the prefectures, many Japanese cities, wards, and other localities support industrial halls that offer similar services. These include regular workshops on common manufacturing problems, use of specialized equipment at low fees, demonstrations of new technologies, plant visits by field agents, and referral to expert consultants for advice on special problems. If Congress wished to support a wider network of technology extension centers, it might set a minimum goal of 120 centers, serving about 24,000 small and medium-sized firms per year and costing about $120 to $480 million a year, depending on the level and quality of service. Some of the funds could come from State or private sources, though it may be unrealistic to demand that these sources take overall the funding within a few years (as the law provides in the case of the NIST centers). A program of this size might soon prove insufficient. It would serve about the same percentage of small and mediumsized manufacturers (7 percent) as is served by Georgia Techs well-regarded industrial extension service for firms in the State of Georgia. That service does not advertise for fear of being swamped with requests. Another promising option with at least two major advantages is a system that would allow manufacturers to lease modern production equipment, or buy it on the installment plan, at subsidized rates. The system would encourage firms to use up-to-date equipment, such as computer numerically controlled (CNC) machine tools. If the system bought U.S.made equipment, it would also benefit U.S. builders of the machinery by offering a stable assured market for part of their output. An equipment leasing system for CNC machine tools, for example, could start with modest government finding-probably about $3
PAGE 26
Chapter I-Summary and Findings l 19 million per year. 39 It could be open only to small firms, or open to all with lower rates for small firms. An option with somewhat similar effects is special tax incentives for investments in advanced manufacturing. This is one of the many inducements to modernize that the Japanese Government offers to businesses, especially small ones. Commercialization of technology from Federal laboratories is a goal that Congress has actively pursued for more than a decade through laws and oversight. There has been progress, but the goal is far from fully realized. On the industry side, many firms fail to pursue energetically what they could get from the Federal labs. On the government side, the two main obstacles are too little money and too much red tape. With other missions taking priority, lab funding of technology transfer has been scanty. Bureaucratic hoops in the parent agency, especially delays for legal review, have many times stalled technology licensing and the conclusion of cooperative agreements between the labs and private industry. Since Congress passed its latest (1989) law promoting technology transfer, approval of industrygovernment cost-shared cooperative R&D agreements has speeded up (it now often takes less than the 90 days allowed under the new law), and such projects are becoming more accepted. The Department of Energy (DOE) has moved to grant some of its big multiprogram National Laboratories the freedom to conclude most such agreements with industry with only a limited wait for agency review. A high level DOE group, the Technology Transfer Project, is working on easing licensing as well as cooperative R&D agreements. Problems remain, however. An umbrella agreement for an Advanced Manufacturing Initiative, negotiated between the National Center for Manufacturing Sciences (representing over 100 companies) and DOE (covering several of the national labs) was not formally launched until July 1991, nearly a year after its announcement. Moreover, funding to promote commercialization is still limited. Congress might wish to earmark some of the labs R&D appropriation for commercialization efforts, perhaps mandating that a few percent of the budget be set aside for the purpose. Also, continued congressional oversight seems to be necessary for getting over bureaucratic roadblocks. Improved protection of intellectual property and modest changes in antitrust law might bolster the competitive position of some U.S. manufacturing industries. Better patent protection could start at home with speedier enforcement-patent cases that go to trial usually take 2 years. Congress might consider designating special patent judges with the technical knowledge to move cases through faster (similar in principle to the tax courts). In foreign markets, the Japanese system has been a special problem. It is slower than the U.S. system in issuing and enforcing patents, and it strongly favors licensing of patents, which U.S. companies often do not wish to do. While the U.S. Government negotiates with Japan on these problems, Congress might establish a program in the Patent Office to provide U.S. firms with information on the Japanese system. Antitrust law and enforcement have been greatly relaxed in the past decade. Changes have been made to allow cooperative endeavors that could improve U.S. competitiveness; perhaps some further moves in that direction could be considered. For example, Congress might allow the Justice Department to certify in advance that joint projects do not violate the law, or to establish safe harbor market shares, so that shares below a certain percentage would not be in violation. Probably most important, Congress could recognize that joint ventures or mergers between U.S. firms are sometimes necessary to fend off foreign competition, and could instruct the courts to listen seriously to such arguments. Congress might also instruct the Justice Department and the courts to weigh carefully the long-term competitive effects of a foreign fros taking over a U.S. firm. Finally, information and exhortation to American manufacturers on how to make things better, given under U.S. Government auspices, have proven surprisingly effective. In 3 years, starting in 1988, the Commerce Department has given the Malcolm Baldrige National Quality Award to nine companies meeting the awards high standards. In that time, 203 companies applied for a possible 18 awards (2 each in 3 categories for each year). The award costs the taxpayers next to nothing. But it has become an excellent means of technology diffusion in several ways. Just filling out the application is instructive. Then, all applicants, win or lose, receive reports from examiners outlining their strong and weak points. Finally, the winners are obliged to share the details of what they did to win. Company representatives give hundreds of speeches a year and hold briefing sessions for executives of other companies, including their competitors. One manager who
PAGE 27
20 Competing Economies: America, Europe, and the Pacific Rim attended came away amazed at the level of detail everything but the financial data. Trade Policy Options Promotion of liberal trade (often termed free trade) has been the policy of the United States since World War II. For two or three decades, this policy served national interests well enough. In the postwar world, the United States was far enough ahead of most other countries that America prospered even when trade was more liberal for imports into the United States than for U.S. exports to some other countries. Today, with several key U.S. industries fallen to second rank, free trade is not necessarily to the countrys advantage, certainly not one-way free trade. Many U.S. industries are struggling to meet foreign competitors equipped with plentiful supplies of patient capital and cutting-edge technology. The outlook is particularly bleak for small or startup firms trying to break into markets dominated by powerful multinationals. Today, most of the highreward industries-characterized by high knowledge intensity, particularly large economies of scale, positive spillovers to other industries, and well-paid jobsare tough to survive in without government help. Indeed, except for the United States, most developed nations have some kind of government program to promote the competitiveness of hightechnology industries. Why has the United States been so vulnerable to these foreign programs? Industry is partly to blame. Many U.S. managers have stuck far too long with outmoded technologies and management styles. But government is also to blame. Unlike its ablest competitors, the U.S. Government has not pursued domestic programs to develop its own important industries. The only way that the U.S. Government has responded to tough foreign competition is as a trade issue: it has attempted to open foreign markets, and it has at times levied extra import duties intended to compensate precisely for foreign subsidies and dumping. However, important foreign market barriers have often taken many years to remove. Similarly, U.S. law regarding subsidies and dumping has done at best a slow, incomplete job of compensating for advantages that foreign industrial policies confer, and the extra duties levied at the border are inherently inadequate to compensate for another countrys domestic policies. Rather than reacting to foreign governments initiatives, always at least one step behind, it would be easier and more effective to improve U.S. competitiveness using domestic programs. Measures might include R&D incentives, tax breaks to encourage R&D and capital investment, increased commitment to technology diffusion, and support for education and training. Trade measures-trying to open foreign markets, and protecting the U.S. market--could be used when necessary, but in a subordinate role. Moreover, these trade measures could be used more strategically. Efforts to open markets could focus on areas of the greatest strategic importance; and protection could be based not on legalistic criteria, but on the industrys need and place in the economy. Some say that government cannot and should not pick industries to promote, and that attempted government assistance to particular industries is likely to make them less rather than more competitive. However, there is general agreement on many of the technologies (e.g., electronic components and information technologies) that are key drivers of industrial performance, and it is not hard to identify industries that use those technologies. And while government intervention has sometimes been counterproductive, the experience of several governments provides guideposts for what approaches work best. For example, industry should take the lead in proposing joint government-industry R&D programs and should shoulder much of the cost; policies should conform with market forces as much as possible; U.S. industry must compete with the best in the world; and industry must work to improve its competitiveness and outgrow the need for assistance. Responses to Foreign Market Barriers Foreign market barriers often hurt U.S. industries. While GATT has reduced quotas (quantitative restrictions on imports) and tariffs (taxes on imports, also called duties), there are many other barriers to imports. These include burdensome customs procedures; discrimination in standards, regulations, and government procurement; and private agreements, tolerated or encouraged by the government, that tend to exclude foreign products. Normally, the U.S. response, if any, is to negotiate to eliminate barriers, with negotiations led by the Office of the United States Trade Representative (USTR). The United States is hindered by lean
PAGE 28
Chapter I-Summary and Findings l 21 USTR staffing, frequent turnover of senior government officials, and the representation of foreign interests by former key government officials. To address these problems, Congress might wish to increase USTR staffing, create more high-level career positions at the USTR and other government agencies, and prohibit senior trade officials from representing foreign interests for several years after they leave government service. Other reasons why foreign markets can take many years to open are harder to address. Barriers take time to identify; other countries can stall negotiations; countries hedge on promises to remove barriers; and when one barrier is removed, another can take its place. And some barriers are ingrained in a countrys business practices and domestic policy, making their removal difficult even if the foreign government is willing. The United States has only limited leverage to induce foreign countries to remove barriers. GATT dispute resolution procedures, while recently improved, are still slow and uncertain. Continuing U.S. efforts in the Uruguay Round to improve these procedures might be productive. Congress might also wish to consider the ambitious task of creating a new multilateral trading system, with a much stronger commitment to and enforcement of free trade. Such a system would require a great deal of pl anning and commitment, and probably would have to be limited to a very small group of like-minded nations in the beginning. The United States can also threaten to retaliate with barriers of its own to imports from the country in question. Under Section 301 and related sections of the Trade Act of 1974, as amended, 40 the USTR can investigate foreign trade barriers, negotiate for their removal, and if necessary retaliate. However, even the investigation phase angers trading partners; and retaliation would most often violate GATT, could provoke counter-retaliation or GATT challenges, would not solve the problems of the U.S. industry facing the market barriers, and could cause problems for downstream U.S. industries. Therefore, Section 301 is not a very serviceable tool. Limited U.S. leverage makes negotiations often slow and ineffective. Barriers that cause particular damage-e. g., Japanese barriers to the sale of semiconductors and supercomputers-often persist for years. While market opening is a worthwhile long-term goal, attempts to remove specific barriers often provide little or no relief in the interim, during which time the affected U.S. industry can suffer serious and irreversible damage. U.S. policy toward market barriers could be reoriented to emphasize domestic measures to maintain competitiveness rather than negotiations to remove barriers. Normally domestic measures, such as R&D support and tax breaks, could keep an industry competitive even in the face of trade barriers. While opening foreign markets could still be pursued as a long term goal, there would be no urgency requiring measures that would anger trading partners. It would also make sense to allocate the governments limited negotiating resources according to an industrys strategic importance. In exceptional cases, domestic measures might not be enough. This might be the case with semiconductors; limited access to Japans market, the largest and most disc riminating in the world, can be an important handicap to U.S. fins. In the rare case of an important U.S. industry facing substantial harm from foreign market barriers that domestic measures cannot alleviate, the national interest might be served by pulling out all the stops to remove the barrier: quick, aggressive negotiations led by high-ranking officials, perhaps with cabinetlevel or even Presidential involvement, followed if necessary by the threat of substantial retaliation, carried out if necessary. Use of Protection U.S. industries normally receive protection against imports only when they are dumped or subsidized. In this case, an extra duty can be assessed in an amount that in theory precisely counteracts the trade-distorting subsidy or dumping. However, U.S. law and practice regarding subsidies and dumping by and large fails to compensate for the advantages foreign governments create for their fins. The reasons include delay, difficulty in proving subsidies or dumping, the laws ignoring or devaluing certain subsidies, difficulty in proving the required injury, and the high expense of legal proceedings. A further problem is that the effects of government assistance can increase over time, rather than dissipating as the law assumes. To some extent, the laws limited effectiveness stems from adherence to GATT requirements. A more effective approach would be to assist industries beleaguered by imports primarily by
PAGE 29
22 Competing Economies: America, Europe, and the Pacific Rim domestic measures to promote competitiveness through cost control, productivity enhancement, and quality improvement. Where domestic measures alone might not suffice, protection could also be used, lasting only as long as strictly necessary. Criteria for awarding protection would include the industrys need, its merit (including whether the industry was making reasonable efforts on its own and showed promise of effectively competing on its own), and the importance of the industry in the U.S. economy. Protection roughly along these lines already exists under Section 201 and the following sections of the Trade Act of 1974, as amended, which is patterned after GATTs so-called escape clause (GATT Article XIX). Section 201 permits the President to grant import relief for up to 8 years when the International Trade Commission (ITC) finds that increased imports cause or threaten serious injury to an existing U.S. industry .41 However, Section 201 has rarely been used in recent years, and as currently written and interpreted is not very serviceable. The injury requirement usually will not be met before serious damage is done, and meeting it is especially hard for high technology industries with rapidly growing markets. While Section 201 could be amended to cover these situations, there is some question as to whether that would be consistent with GATT. Congress could empower the President to grant protection apart from Section 201. Under GATT Article XXVIII, the United States could negotiate with other countries to accept higher U.S. tariffs on certain goods in exchange for reduced U.S. tariffs on other goods. While reduced tariffs on the other goods could adversely affect other U.S. industries, the government could possibly mitigate such effects with tax breaks or other programs. Also, continued protection ideally would depend on sufficient effort from industry to improve its competitiveness. If other countries would not agree on compensation, in rare cases as a last resort the United States might impose protection anyway and risk a GATT challenge. To some other nations, that might be interpreted as a signal that the United States was abandoning its commitment to GATT and free trade. While this is a risk, it might be preferable to the alternative of losing critical industries altogether. Promoting Exports Nearly all industrialized nations promote exports. Exporting is difficult, requiring firms to overcome differences of language, geography, and custom. Governments help firms to learn about markets, to identify potential customers and distributors, and to comply with administrative requirements. Such assistance seems particularly needed in the United States, whose firms must export more manufactured goods than before. However, the United States spends far less promoting manufactured exports than many of its important trading partners. In the late 1980s, low funding even led to situations such as commercial officers in U.S. embassies not having funds to return phone calls from U.S. firms. While budgets have improved somewhat in the last few years, Congress might wish to consider funding more on a par with that of other countries. Congress might wish to make a policy statement that export promotion should be a priority not only for commercial officers abroad but for the whole diplomatic staff. Cabinet-level involvement in promotion activities, such as Commerce Secretary Mossbachers presence in Tokyo in April 1991 to kickoff the Japan Corporate Program, could also be encouraged. Nearly all industrialized nations also assist firms with export financing. In the United States, applications for financing assistance for manufactured exports must be justified on a case-by-case basis; the need for justification increases delay and the burden on the exporter. Congress might consider adopting the approach used by Japan and many European countries, which determine in broad policy terms what exports to assist and then assist all creditworthy exports within the guidelines. Many other countries condition some of their foreign aid on purchases of capital goods, construction services, and the like from the donor country. While the United States also ties some aid to purchases, U.S. nondefense aid focuses on agriculture, nutrition, health, and education, rather than large capital projects. U.S. effort to improve international agreements limiting the use of tied aid is worth continuing. However, it is uncertain how successful that effort will be. Congress could expand the so-called War Chest for matching foreign tied aid offers to more effectively discourage foreign tied aid. The War Chest was funded at $150 million in grants for fiscal year 1991, though as of July 1991 only $58 million had been used. Some other countries spend many hundred million dollars annually.
PAGE 30
Chapter I-Summary and Findings l 23 Congress could expand the Trade and Development Program (TDP), funded at $35 million for fiscal year 1991. TDP helps to pay for feasibility studies or other plannining assistance performed by U.S. firms for capital projects in developing and middle-income countries. U.S. participation in the pl anning phase has often helped U.S. firms win contracts for the actual project. So far, $161 million in program funds have led to documented U.S. sales totaling $3.2 billion, with an estimated $18 billion more sales expected as projects mature. 42 In expanding TDP or otherwise increasing the emphasis on capital projects, care should be taken to avoid adverse environmental and social effects, which in the 1970s turned the United States away from such projects. As well as promoting exports, the U.S. Government at times impedes export of high-technology goods and data by its system of export controls. Export of dual-use itemsthose with both military and civilian use-is regulated by the Export Administration Act of 1979, as amended; 43 this requires firms to obtain a license to export certain items for certain destinations. In 1990, perhaps $90 billion worth of U.S. exports of manufactured goods required a license. There is a genuine need for some control over exports to guard advanced technologies and products that could be used in weapons against the United States. Yet export controls have also proved an unnecessary hindrance to some manufactured exports, at times merely shifting the business from U.S. to foreign firms. For example, until mid-1990 U.S. export controls limited exports of personal computers based on Intels 80386 processor, although they were easily available from foreign sources. While U.S. export controls are being reformed, the process is incomplete. Reform could be facilitated, as could ongoing administration of export controls, by a competitiveness policy agency. Such an agency could identify exports with strategic economic importance and help to expedite their approval when possible. While an Administration sympathetic to competitiveness concerns is essential for fully effective reform, Congress could take measures on its own to further decrease unnecessary burden on commercial competitiveness. These measures include better funding and staffing of the office that determines when foreign availability makes U.S. controls ineffective, encouraging more political appointees with technical backgrounds, encouraging more use of automatic indexing to track technological change (subject to yearly review), and court enforcement of congressionally mandated deadlines. Congress could take measures to stop abuse of the State Departments Munitions List, which is supposed to contain only items with purely military use but has been used to control some dual-use items. An emerging area of concern is the use of so-called foreign policy controls, which are largely untouched by the recent reforms. Congress might wish to extend the recent reforms to foreign policy controls where feasible. For example, if the purpose of a foreign policy control is to guard technology for making chemical weapons, Congress could require that items not be controlled without multilateral agreement to control them, and that controls be removed if they are ineffective because of foreign availability. GOVERNMENT-INDUSTRY PARTNERSHIPS In addition to trade policy, a comprehensive strategy for greater competitiveness includes promotion of technology development and diffusion, risk-sharing between government and industry, and a generally supportive environment for adoption of new technologies. Risk Sharing in Technology Development The least intrusive and least expensive of several risk-sharing options is an R&D partnership for developing new technologies of commercial interest. 44 The main reason for government to support such ventures is that potential benefits to society are great but the likely payoff to individual firms is too small or the possibility of failure too great to make it worth their taking the risk. In the U.S. financial environment, with its high costs and emphasis on short-term profit, government contributions to risky ventures are especially significant. The Federal Government has in the past given some technology assistance for commercial purposes, but not on a coherent, strategic basis. By far the largest existing Federal program for precommercial technology development is NASAs aeronautical R&D program (amounting to over $900 million per year, including expenses for wages and salaries, R&D, and facility construction), which supports military and commercial technologies. The 5-year
PAGE 31
24 l Competing Economies: America, Europe, and the Pacific Rim Sematech program, to help the semiconductor industry develop a manufacturing process technologies for memory chips, is large ($ 100 million per year for 5 years in government funds and a matching sum from industry) and was created ad hoc in response to strong industry pressure and the argument that a competitive U.S. semiconductor industry is essential to national defense. A small beg inning for a more general R&D partnership is the Advanced Technology Program (ATP), operated by NIST in the Department of Commerce. ATP was established in the 1988 trade act, which authorized the Program to assist businesses in doing research in precompetitive, generic technologies. In fiscal year 1990, Congress provided ATPs first tiding, $10 million, and raised funding to $36 million in fiscal year 1991. ATP might in time become a full-fledged civilian technology agency (CTA), although it was not created with that specific mission. Bills to establish more formally an Advanced CTA in a new Department of Industry and Technology, which would replace the Department of Commerce, were introduced in the I00th and 101st Congresses. 45 These bills defined the agencys mission as contributing to U.S. competitiveness by supporting long-term, highrisk projects that are likely to yield important benefits to the Nation but lack adequate private support. A bill that passed the Senate in 1989 would have given the ATP a similar mission, and authorized funding up to $100 million per year. 46 Any CTA would have to start small, as ATP has, and grow only with experience. A mature agency might have a research budget of $300 to $600 million per year. This very rough estimate is based on the list of about 100 technologies developed by the private-sector Council on Competitiveness, 47 each of which might merit government participatory grants of about $1 to $2 million per year, with enough redundancy that two or three grants might be made in each field. Collaboration with industry in choosing technologies for support would be essential. If private companies are not interested enough in the technology to put up at least half the money and do much of the work, then the chances for commercial success are probably remote. Joint finding helps the government resist special interests or political pressure in choosing technologies for support. At the same time, a CTA would need a set of guiding principles-e. g., it should look for technologies that are knowledgeintensive and have wide applications in many products and industries. NASAs experience underscores the importance of collaboration with industry, if the goal is competitive success. Most of NASAs R&D is for military as well as commercial applications, and much is basic research that is quite freely available to the world and has little or no near term application to commercial production. NASAs greatest contributions to competitiveness are in two areas: its facilities, such as wind tunnels and its Numerical Aerodynamic Simulator, which are solely or preferentially available to U.S. companies; and R&D projects in which U.S. companies are close collaborators. So far, the U.S. Governments offers of collaborative R&D projects have been snapped up by private companies. The Advanced Technology Projects frost batch of l-year cooperative grants, amounting to $9 million, attracted 249 applicants requesting a total of $150 million. Eleven projects, half of them joint ventures or consortia, were chosen; they initiated R&D projects that are expected to cost $100 million (including private funds) over 5 years. The total dollar amounts in the governments few cooperative R&D programs are tiny compared to the $100 billion per year that U.S. companies spend for R&D. It is remarkable that such small programs have drawn so large a response, including proposals from such industrial giants as Du Pent, AT&T Bell Laboratories, and IBM. If Congress wishes to continue the expansion of cooperative R&D programs, with the ultimate goal of having an agency of similar size and importance as the Defense Advanced Research Projects Agency (DARPA), it may opt for another modest rise in the ATP budget for fiscal year 1992. It might also consider putting into law a more formal statement of goals for the Program. Congress has already responded in an innovative way to the question of how to handle participation by foreign-owned firms in cooperative R&D projects. In appropriating funds for the ATP for fiscal year 1991, it set standards that apply to U.S.-owned as well as foreign fins, thus bypassing ownership as the central criterion for participation. The Secretary of Commerce is authorized to decide on a fros eligibility based on its contributions to high-value-
PAGE 32
Chapter I-Summary and Findings l 25 added manufacturing production and manufacturing employment within the United States. Further conditions apply to foreign firms, based on equal treatment for U.S. firms in the foreign fins home country. These provisions provide guidance but give the Secretary of Commerce great latitude. Congress may wish to exercise substantial oversight for a time on how these provisions are carried out, and to extend these eligibility criteria for at least a few years to permit evaluation of their merit. Financial Risk-Sharing Strategic technology policy goes only so far. It is up to industry to make the much larger investments in product design, manufacturing equipment and tooling, worker training, and acquisition of knowhow by managers and production engineers needed for commercial production. It is normal and expected for private industry to make these investments and take the risks. Sometimes, however, in some critical sectors, private investment is inadequate from the standpoint of social benefits. U.S. investment in production equipment is meager compared to Japan s. Japanese investment in machinery and equipment, as a share of GNP, has been twice the U.S. rate since the mid-1970s, and the discrepancy has recently widened. The discrepancy is especially damaging in critical commercial industries, such as semiconductors, that require continuing large investments for new generations of products every 3 years or so. Government policies to share financial risks with industry can take the explicit form of loans or loan guarantees on advantageous terms, or they can take the form of tax breaks, which are implicit expenditures. Both put burdens on the Federal budget, and Congress cannot opt for either without compensation tax rises or spending cuts. Macroeconomic policies that help to lower capital costs and provide stability are probably the most important help government can provide to encourage greater investment in technology development and deployment by U.S. manufacturing companies. Specific options for financial risk-sharing are worth considering, however. Of the two forms of risk-sharing, tax breaks are more within U.S. tradition and experience. In the past, U.S. companies have received accelerated depreciation and tax credits for capital investments. Although certain activities (e.g., real estate) have been singled out for special treatment, Congress has not in the past designed tax incentives to improve the competitiveness of particular industries. Across-theboard tax breaks for capital investment would cost the U.S. Treasury tens of billions per year, while loosening the rules for selected industries would cost less. The National Advisory Committee on Semiconductors, for example, estimated that allowing 3-year rather than 5-year depreciation for new investments in semiconductor manufacturing equipment would cost the U.S. Treasury $180 million per year in lost revenues, and would prompt $450 million in added capital investments by the industry. 48 If Congress wishes to target tax breaks to critical commercial industries, the best way is to make these tax measures part of a comprehensive strategy that includes such things as technology support and trade policy. This presupposes the existence of an institution able to form such a policy. Direct financial aid to commercial industries is mostly foreign to U.S. experience; the one previous effort, with the Reconstruction Finance Corporation after World War II, was not targeted to critical industries, and is generally considered an expensive failure. If Congress wishes to consider direct aid in exceptional cases, it may wish to start at a very modest scale. Government Purchases Government purchases have been an important factor in the birth and growth of several industries, including aircraft engines, semiconductors, and computers; the big buyer in each case was the Department of Defense (DoD). Today, defense purchases are not a very promising source of financial support for critical commercial industries. Technological spillovers from military to commercial products are probably declining (though the evidence on this is mixed). More important, laws and regulations governing DoD purchases are so restrictive and cost-inflating that it is increasingly difficult to combine military and civilian production and to take advantage of whatever spillovers may exist. Post-Cold War declines in defense spending have removed some of the potential support. Federal, State, and local government spending for nondefense goods amounted to about $97 billion, $10 billion of it Federal, in 1990. Although the total is modest as a share of the GNP for goods (about 5 292-889 0 91 2 QL:3
PAGE 33
26 l competing Economies: America, Europe, and the Pacific Rim percent), certain kinds of government purchases can be significant. For example, Federal purchases greatly aided the supercomputer industrys development. This rolebeing an assured customer for a startup productis especially significant. Suppose the Postal Service (a quasi-public agency) decided that electric vehicles could help reduce pollution, and that their limited range and need for frequent recharging would not handicap mail runs. A substantial order might give a real boost to a U.S. producer of electric vehicles, providing it with a head start over foreign competitors. Preference for U.S. firms would be consistent with the GATT Procurement Code, which does not yet cover the Postal Service. Internationally, procurement tells a familiar story: the U.S. market is more open than those of many major trading partners; many major trading partners use procurement as a strategic tool to develop important industries, while the Federal Government generally does not do so intentionally. Congress might wish to reorient U.S. policy to be more proactive: as well as negotiating to make procurement markets more open, the United States could take stock of opportunities to use its own procurement strategically, such as the Postal Service example. There are likely many such opportunities that are not yet prohibited by international agreements. Using procurement strategically would help U.S. competitiveness pending further market-opening agreements and would give the United States more leverage in negotiations. Awareness of strategic opportunities would make the United States a more informed negotiator. For example, if Postal Service procurement of electric cars were deemed a strategic opportunity, the United States might not agree to subject the Postal Service to the GATT Procurement Code unless other countries give up equivalent opportunities. AMERICAN FIRMS, FOREIGN FIRMS, AND AMERICAN INTERESTS Government involvement in efforts to boost civilian industrial competitiveness invariably raises the issue of who is eligible to participate. Three decades ago, there were only a handful of true multinationals. Today there are hundreds of firms that have a substantial interest and presence in more than one nation. The stock of direct investment abroad, on the part of firms from all nations, has increased over 10 percent per year since 1960, and by 1988, stood at over $1.1 trillion. 49 Direct investment abroad is increasing much faster than world merchandise trade or world economic output. All developed nations must make decisions about how to treat foreign firms, and increasingly, those decisions center on foreign firms eligibility to participate in government-sponsored projects to improve industrial competitiveness. Foreign direct investment in the United States (FDIUS) is also on the increase-in fact, FDIUS is increasing faster than the world average. Between 1967 and 1988, the stock of FDIUS increased from $9.9 to $328.9 billion, or over 18 percent per annum. The world stock of direct investment abroad increased from $105.5 billion to $1.2 trillion over the same period, or 12.4 percent annually. The gross product of foreign affiliates in the United States accounted for 3.4 percent of GNP in 1987, nearly double their 1.8 percent in 1977. The presence of foreign direct investors is especially prominent in manufacturing. In 1987, U.S. manufacturing affiliates of foreign parents accounted for 12.2 percent of the assets of all American manufacturing, compared with 8.9 percent of the total net worth of all nonfinancial corporations. So while foreign companies are relatively small players in the U.S. economy, they control or influence a significant chunk of manufacturing. The U.S. Government funds relatively fewer industrial-competitiveness programs that could admit foreign firms than governments in other developed nations, especially in Europe and Japan. Nevertheless, we are in the thick of the debate; foreign companies have asked for access to Sematech and the Advanced Technology Program of the Department of Commerce. How we end up dealing with foreign firms or their American affiliates depends greatly on how they behave. American firms too are held up to scrutiny in terms of their contributions to national well-being in the debate, often in caricature; some believe that the typical American firm does little manufacturing or R&D in the United States. When we ex amine the contributions American firms make to the U.S. economy and living standards and compare it with foreign fins, the caricatures fall apart. The U.S.-based parent companies of American multinationals act, on the whole, like American companies: their U.S. operations accounted for 78
PAGE 34
Chapter 1--Summary and Findings l 27 percent of the companies total assets, 70 percent of their sales, and 74 percent of their employment in 1988, slightly more on all three counts than in 1977. 50 Contrary to the popular stereotype, there is a slight tendency for U.S. multinationals to do more of their business in the United States than formerly. American multinationals also keep good jobs and develop technology at home: the vast majority of their R&D is done here, 51 compensation per U.S. employee is 39 percent higher than compensation per employee in offshore affiliates, and assets per U.S. employee are 30 percent higher than assets per employee in offshore affiliates. U.S. affiliates of foreign multinationals behave much like American firms in America, although there are some key differences. Affiliates do not differ significantly from American companies in terms of compensation per employee and investment in plant and equipment as a percent of sales. 52 They treat their employees similarly during economic downturns, being about as likely to lay off workers as American companies. U.S. affiliates of foreign multinationals are, on the other hand, less likely to do R&D in the United States than American companies; the R&D intensity (spending on R&D as a percent of sales) of U.S. manufacturing affiliates was about half that of American manufacturers. This is consistent with common behavior by multinationals (including American multinationals), most of whom do the bulk of their R&D at home. Aggregate data on the hours and quality of training given in U.S. affiliate companies are not available. We do know that workers and managers in Japanese motor vehicle transplants receive substantially more training than workers and managers in American companies auto plants. overall, however, we would expect to find little overall difference in the amount or quality of tr aining given to workers in U.S. affiliate companies, compared with workers in American companies. There are two sources of real difference between the behavior of affiliates and the behavior of American fins. Foreign affiliates are far more likely to import than American-owned companies, and affiliates of Japanese parents more so than affiliates of European or Canadian parents. The merchandise trade deficit associated with affiliates of foreign parents is substantial; in 1988, it was $90 billion, three-fourths as large as the entire U.S. merchandise trade deficit. 53 It is a mistake, however, to hold foreign companies primarily responsible for Americas poor trade performance. In large part, we are responsible for that: our anemic savings rate does not generate enough capital to cover our investments; our appetite for consumption is greater than our production; foreign goods are often better and/or cheaper than domestically made ones. The fast increases in foreign direct investment, and the associated trade deficit, are mainly results of these underlying weaknesses, not the causes. In other words, if we somehow prohibited American affiliates from importing, or forced them to export more (like some developing countries do), and changed nothing else, we would expect the dollar to fall to compensate. Japanese direct investors also behave very differently in many respects than other foreign investors or American fins. Their propensity to import is higher; in 1988, affiliates of Japanese companies imports totaled $75.9 billion, 51 percent of total imports of affiliates. Japanese affiliates share of U.S. sales was only 26 percent of the sales of all affiliates. Moreover, Japanese affiliates imported almost exclusively from Japan; 93 percent of their imports came from Japan. Other foreign affiliates import from home, but to a much smaller extent; European affiliates got 70 percent of their imports from Europe (although the attachment of affiliates of any nation to imports from that particular nation is less strong) and 73 percent of Canadian affiliates imports were from Canada. Japanese direct investment appears to be far more oriented to selling Japanese products abroad than the investment of European nations or Canada. This shows up not just in the propensity of Japanese affiliate companies to import from Japan, but in the profile of investment; Japanese firms are more likely than European firms to invest in wholesaling affiliates, particularly in motor vehicles, and their primary activity is to sell Japanese products. Possibly as a result, Japanese parents are more likely to maintain tight control over their affiliates than are European parents. American affiliates of Japanese electronics companies mostly employ Japanese managers, while American electronics affiliates in Japan mostly employ Japanese managers. This pattern of affiliates of Japanese companies having Japanese CEOs holds true in other countries as well as in the United States. Is there a reason to be especially concerned about--or even wary ofJapanese investment?
PAGE 35
28 l Competing Economies: America, Europe, and the Pacific Rim There are concerns that Japanese investment in high-tech firms may simply be a vehicle for fast transfer of technology to Japan, and possibly the eventual benefits (manufacturing, employment, value added) that flow from innovation. But Japanese direct investment has had many positive effects as well. Japanese direct investment in automobile manufacture has raised the standards of quality among American auto parts manufacturers that have sold to Japanese transplants. The transplants have also been largely responsible for demonstrating how efficient, high quality Japanese auto assembly works and, to some extent, transferring knowledge of how to manage such enterprises. Japanese transplants in autos have admirable records of training employees and managers. Given our inclination to consume and unwillingness to save, increased foreign investment was inevitable. It is little wonder that it should come from Japan, whose companies are richer and more competitive than any others. As to differences in Japanese affiliates behavior and investment patterns, it is possible (though by no means universally accepted) that these stem more from the inexperience of Japanese firms in investing abroad and managing affiliates than from peculiarities of Japanese business practice, and that in time their behavior will come to resemble that of other multinationals. On the other hand, it is also possible that there are certain kinds of Japanese investment or business practice that will not contribute to American well-being. Careful monitoring over the next few years will be needed in order to determine how to treat Japanese affiliates. In the meantime, we are still faced with the choice of including or excluding U.S. affiliates from government programs like the Advanced Technology Program. Private research consortia also receive inquiries from foreign firms and their U.S. affiliates. While such private organizations often restrict membership based on a fros nationality, government programs (except those whose mission is national defense) often cannot. Nominally, all nations that belong to the Organization for Economic Cooperation and Development (OECD) subscribe to the principle of national treatment, which stipulates that foreign firms should be treated exactly the same as domestic firms. 54 In practice, most nations of the OECD base decisions on participation in governmentsponsored competitiveness programs on other criteria. One is mirror reciprocity, meaning that affiliates of foreign firms are given the same treatment in the host country that the host countrys firms are given in the foreign affiliates countries. Another is performance standards. In Europe and the United States, affiliates or subsidiaries of foreign parents must meet certain standards, for example, establishing manufacturing and R&D facilities in the country, or adding a certain percentage of value domestically, to be considered for participation. As of mid-1991, both reciprocity and performance standards govern participation in the Department of Commerces Advanced Technology Program. The EC uses similar criteria, with respect to American firms at least, to govern participation in the Framework Program. (See ch. 5 for further discussion of the ECs Framework Program.) The issue of how affiliates of foreign firms are treated in the United States is assumin g greater importance as a competitiveness question, but at the moment it is of less concern than what we do to improve the competitiveness of American fins. There is increasing agreement (far from unanimous) among analysts that there are many things the government could do to assist U.S. firms to become more competitive, including many of the options noted above, in the summ ary of policy options. U.S. TRADE POLICY Since World War II, the United States overriding objective in trade policy has been to promote free trade throughout the world, using the GATT system and, to a lesser extent, bilateral negotiations. The GATT system has reduced quantitative barriers to trade (quotas and tariffs), and as a result is often given credit for the increase in world trade. For most of the postwar period, U.S. firms prospered under this regime. To be sure, some industries had problems, even in the 1960s when most U.S. industry was at the technological forefront of global competition. The textile and apparel industries, for instance, relied heavily on unskilled and semiskilled labor, and as a result faced competitive pressure from low-wage countries quite early; treaties limiting textile imports were signed in the 1950s. Television manufacturers came under pressure from imports in the 1960s, as a result of both high production costs and, toward the end of the decade, superior technologies (solid state circuitry) in Japanese products. Until the early 1980s, the industries that had competitive trouble were regarded as outliers, which the United States could
PAGE 36
Chapter l-Summary and Findings l 29 probably afford to lose as it shifted into hightechnology sectors. But in the 1980s these trade troubles spread. The indisputable fact emerged that American technology development and diffusion was deficient in even the most high-technology industries. Now, it is difficult to find an American industry that is in no competitive trouble at all, and there are a few where only fast and drastic action can preserve domestic manufacturing. Moreover, American firms are significantly behind in an increasing number of emerging technologies and industries. Trade increasingly exposes U.S. companies to competition from foreigners with superior technologies, deeper pockets, better trained workers, and governments determin ed to provide their indigenous firms with advantages. Some of these advantages are nationwide-e. g., frost-rate education, encouragement of household savings, and tax breaks for R&D and capital investment. Some governments, notably in Japan, Korea, and Taiwan, have also targeted for support specific industries, such as semiconductors and computers, that seem to contribute disproportionately to a nations wealth and economic development. Developing such industries is often a race in which the firms or nations that get ahead will likely stay ahead for some time. A company with technical advantages or greater market share can reap economies of scale or learning which will let it capture additional market share or finance more R&D than its competitors, enabling it to pull still further ahead. Governments have targeted critical industries with both domestic policies and home market protection. Domestic policies include R&D support, special tax breaks, preferential financing, and tolerance or encouragement of cartel pricing in specific industries. R&D programs can give firms a technical advantage over competitors abroad or at home. Special tax breaks or other financial support can help domestic companies pay for their investments or charge lower prices. Trade protection has rarely if ever been successful when used alone, but in combination with domestic policies it can be a powerful tool. A protected home market can enable domestic firms to catch up with more advanced foreign companies without having to compete with them for domestic customers. Profits in a protected home market can bankroll forays into export markets at low prices, R&D, and investment in worker training and equipment. In the short term, foreign producers could probably meet these low prices; but in the long term, foreign firms not similarly supported can lose market share and the revenues to fund new investments. Of course, protection can easily go astray, leading to an industry ill-suited to international competition, but when managed properly it can aid a nations economic development. Other countries domestic programs and market protection have often delivered a one-two punch to U.S. industries. For the most part, the U.S. Government does not have comparable proactive programs to promote its own industries. U.S. trade policy plays out by noticing some of the advantages foreign firms enjoy, and then trying after-the-fact to eliminate or offset them, usually after substantial delay and often incompletely. Important foreign market barriers often persist for years, despite U.S. attempts to eliminate them. While some advantages enjoyed by foreign firms are recognized by U.S. dumping and subsidies law, various problems prevent or limit redress even in deserving cases. These problems include the expense required to prepare a petition and fight a legal case, the time it takes to conduct investigations, ways by which foreign firms circumvent duty orders, the interpretation of the injury requirement so as to inhibit timely relief, and the laws failure to recognize the impact of many subsidies. U.S. policy thus puts important industries at risk. No matter how hard U.S. firms work, under current conditions they might not be able to compete with foreign industries backed by their governments. Other aspects of U.S. policy are also ineffective in promoting the competitiveness of U.S. industry. While many foreign governments procurement policies are attuned to fostering national industries, U.S. procurement policy is not. The Commerce Departments export promotion programs, while useful, are small and ineffective compared with programs in other countries. Export financing by the Export-Import Bank of the United States is sometimes less attractive than that offered by other countries export financing agencies. Finally, U.S. national security export controls unduly hinder high-technology exports; while many controls truly are necessary for national security, some are not.
PAGE 37
.30 l Competing Economies: America, Europe, and the Pacific Rim EUROPE AND THE SINGLE MARKET The United States is not alone in facing questions of what to do about lagging industries and technologies. The nations of the European Community, individually and together, have a long record of attempts to use industrial policy, and with few real successes in past attempts, are launching a new initiative. Known as the Single Market, or, after the proposed date of its inception, Europe 1992, the initiative is really a wide variety of new policies and agreements broadly aimed at increasing European unity, improving technology, and increasing competitiveness. Unity has been an elusive goal for Europe. The first step was taken with the Treaty of Rome in 1957, but progress toward true harmony was slow and painful. In 1985, the Commission of the European Communities (CEC) proposed a set of some 300 specific policy actions that would be needed to eliminate barriers to movement of goods, services, people, and capital throughout the 12 nations of the EC. 55 For several reasons, one of the most important being determination to escape the economic stagnation that had bedeviled Europe for more than a decade, the CEC adopted the 1985 White Paper, and progress toward implementing the 300 specific resolutions began. Even though progress was immediate and rapid, the history of past disappointments led many American analysts to discount it for several years. Now, however, it is obvious that the EC will have some kind of single market in place at least by the end of 1992, although there will likely be some unfinished items still on the agenda. Significant changes in European economic activity are very possible. How significant? And what does it mean for the United States? Nearly everyone expects that removing sources of commercial friction among the 12 EC nations impediments to movement of goods, people, services, and capital-will mean faster growth in the GNP of the European Community. The range of estimates of the increase in growth is wide. The closest thing to an official estimate of the EC is a report done in 1988 (known as the Cecchini report), which estimated gains at 4.3 to 6.4 percent of GDP accruing over a 6-year period, or up to 1 percent additional growth in GDP each year. Another 2.5 percent (over the 6 years) is possible if appropriate accompanying macroeconomic policies are added, according to this estimate. The Cecchini report has been hailed as an impressive technical work, but its growth estimates are also regarded as optimistic. 56 In contrast, the gains in GDP from the elimination of tariffs on industrial products among Common Market countries in 1968 were on the order of 1 percent, total. 57 What this means for the United States, in the short or long run, is murky. Additional growth, even if it were substantially below the levels estimated by the Cecchini report, would ordinarily mean increased opportunities for U.S. firms to sell goods to and produce goods in Europe. The former (increased exports) would further the national interests of the United States directly; the latter only indirectly, to the extent of contributing a bit to the prosperity of firms headquartered here. But the Cecchini report also makes it clear that some of the added growth in Europe is expected to come at the expense of imports from outside the EC; the Cecchini growth forecasts assume a reduction of imports from outside the EC by 7.9 to 10.2 percent. 58 Whether there will be growth in Europe due to factors not anticipated by the Cecchini team, and whether these increase the possibilities for U.S. exports, is simply not clear. Most of the fears that EC 1992 would be a Fortress Europe have been put to rest, but there are a few signs of increasing protectionism in Europe. Two areas where Europe has taken specific steps that limit imports are in automobile trade and semiconductors. The former is not expected to have any significant negative impact on American exports; most U.S. autos sold in Europe are manufactured there already. The story is different for Japanese producers. The other area, semiconductors, could be more problematic for the United States. A change in the rule of origin for semiconductors 59 and a stiff existing tariff on imported semiconductors favor producing in Europe to exporting. This may not have a detrimental effect on companies that already own wafer fabrication plants there, or on the large, rich Japanese producers, but smaller American producers without existing plants in Europe are faced with the painful choice of losing European markets or making expensive investments there. Certainly, the change in the rule of origin makes it more difficult for the United States to export semiconductors to Europe.
PAGE 38
Chapter l-Summary and Findings l 31 There are also developments, such as the liberalization of European government procurement, that could open new markets to non-EC firms, although it is not likely that this will mean much in the way of increased exports. Most of the explicit liberalization of trade is intra-EC. In the short run, then, we would expect few major changes in U.S .-EC trade as a result of the single market. But what about the long run? One of the aims of EC 1992 is to make European firms more competitive with American and Japanese fins. The single market may contribute somewhat to that by enabling European firms to achieve new economies of scale in a market that will have about the same GNP as the United States. Another contribution could come from the Framework Program and EUREKA. These two programs, the first an EC program and the second a program with 19 members (all the EC countries are members, as is the EC itself), fund R&D intended to improve civilian industrial competitiveness on an impressive scale, by U.S. standards. The third Framework Program (1990-94) is funded at ECU 5.7 billion, and EUREKA projects announced between 1985 and 1990 came to ECU 7.4 billion. At a rough estimate, public funding of European cooperative research in both the Framework and EUREKA programs comes to about $2 billion per year. 60 The largest parts of the Framework Program, ESPRIT and RACE, are aimed at microelectronics, computer, and telecommunications technologies (including services and software as well as hardware). BRITE is a large program that funds R&D in products and processes to improve basic manufacturing. Other programs address technology development in many areas: medicine and health, energy, advanced materials, biotechnology, agriculture, and road transport efficiency and safety. Both the Framework Program and EUREKA projects encourage cross-border collaboration to promote unity and exchange of scientific and technological information in Europe, and the resulting enthusiasm for cross-border collaboration is due, in no small part, to this encouragement. Whether this increased international exchange and collaboration, or the money spent on developing new technologies, will contribute significantly to European competitiveness is questionable. It is probably too soon to judge most of the programs, but those that have been going for awhile, and several past efforts at crossborder collaboration in technology, have produced few unambiguously successful results. JAPAN Japan is the economic phoenix of the postwar period. Throughout the nearly five decades following the war, its growth of GNP and productivity have consistently been higher than in the rest of the developed world. That it should be so was by no means obvious in the frost decade after the end of the war. Japan was desperately poor, short of most raw materials, and faced labor strife. Now, one of the biggest problems Japanese bureaucrats face is how to contain the robust productive power of its premier corporations enough to avoid exacerbating trade disputes. The Japanese Government has long used industrial policy to push its economy toward more high-value-added, knowledge-intensive industries that use more highly skilled labor and fewer natural resources. The primary tools are financial aid, government sponsorship of price, investment, and R&D cartels, and protection of the domestic market. These policies were instrumental in improving competitiveness in industries like steel, motor vehicles, semiconductors, and computers. A few caveats are in order. The impression is often given that Japanese policies alone are responsible for Japans economic success, and that the record of success is unblemished. In fact, Japans policies were creative and innovative but they would have been much less effective in a society with less well educated people that placed lower value on hard work and ceaseless pursuit of improvement. Japans culture, with its emphasis on achieving consensus, and on the performance and interests of groups rather than individuals, played a role, although the prominence given to cultural explanations of Japans success in the popular literature is often overdone. There are also several examples of failure in Japanese industrial policies. For instance, the longterm goal of promoting an indigenous large civilian air transport industry has remained elusive, and MITIs expectations have been scaled back considerably. There is widespread disagreement, at least among American analysts, about the overall effect of Japanese industrial policy on Japans national income and standards of living. Japanese consumers have long been able to live less well than American
PAGE 39
32 Competing Economies: America, Europe, and the Pacific Rim consumers on an equivalent amount of income, in part because of policies that sheltered many industries from foreign competition. Some of those policies, in turn, were made to foster industrial development; the inference that Japanese consumers pay for Japans industrial policies is quite correct. But the tradeoff is not just between Japanese industry and consumers; it is also a sacrifice of short-run gratification in favor of enhanced prospects for long-run growth. Even as Japanese standards of living and wages approach those of the richest nations, there are few signs of impending stagnation, and it is likely that faster growth of Japanese living standards will continue, surpassing ours. That does not mean that Japanese policy remains the same as always. The hand of the government in directing industrial development is considerably less heavy than it was during the high-growth period (which ended in 1974, with the first oil shock). Japans government has liberalized financial markets and consumer credit, reduced formal, quantitative import barriers, liberalized foreign investment, and reduced the number of cartels. Some have interpreted this as proof that Japans economy is a modem, capitalist, free-market one along the lines of America, Canada, Germany, or Great Britain. Yet Japans trade patterns remain peculiar by the standards of other developed countries; manufactured imports are quite low, and a strong preference remains for adding as much value as possible in Japan. Japanese direct investment abroad is more oriented to exports than the direct investment of other developed nations, and it is an outlier among developed nations in that foreign direct investment plays a much smaller role in its own economy. Many in America and Japan argue that all this is simply because foreign exporters or investors are not diligent enough; their products are inferior or their knowledge of Japanese business practice is weak. Some of that is true, but it is not the whole story. Japans Government is still actively involved in creating an advantageous environment for Japanese business. The computer industry was targeted for development nearly three decades ago, and within the past 5 years has come of age; many Japanese computers from mainframes to laptops are now as good as or better than American models. That payoff is the result of three decades of company diligence and experimentation, combined with tax incentives (general and specific), R&D funding in strategic areas, subsidized leasing, and market protection. 61 Policies changed over time, in response to different industry needs, and even now, with competitive and technological advantage increasingly weighted on the side of the Japanese computer makers, policies to support specific segments continue. One such area is supercomputers, where Japanese Government support has continued through the 1980s and into the 1990s. Some support comes in the form of funding for research consortia. From 1981 to 1989, the Japanese Government spent 18.2 billion yen (about $121 million) on the High Speed Computing System for Scientific and Technological Uses Project, aimed at producing a machine with a speed of 10 Gigaflops. NTT, the Japanese telephone company, also supports supercomputer technology development in its own supercomputer project, and several public and private projects are exploring parallel and massively parallel processing. Another important element of the strategy is procurement. Until very recently, American supercomputers were superior to Japanese supercomputers, yet while U.S. machines only were bought and installed in America and Europe, they were a small share of Japanese purchases. In 1987, for example, Cray and Control Data, American supercomputer makers, accounted for 73 percent of installed supercomputers in the world; Japanese companies for 27 percent, which consisted entirely of sales within Japan. Moreover, the Japanese companies Fujitsu, NEC, and Hitachi accounted for 87 percent of all Japanese installations, and Cray for only 13 percent. In part, that could be attributed to the Japanese preference for buying goods from and doing business with other members of keiretsu, but American supercomputers had a far more difficult time in the Japanese public sector than in the private sector. A few Japanese private companies bought Cray machines because they were better and faster, and buying an inferior Japanese machine would have been a real handicap; in the public sector, however, procurement was almost exclusively of Japanese machines. The Japanese Government apparently was determin ed to provide Japanese companies with a secure market while they worked hard to catchup to or surpass Crays technology. There are many who regard such practices as unfair or underhanded. In fact, they are logical, reasonable things for governments to do; Japan is hardly alone among industrialized countries in using the power of public procurement to foster domestic
PAGE 40
Chapter 1--Summary and Findings l 33 business and competitiveness. The story is not told for the purpose of castigating Japanese policy, but to illustrate that policies designed to create competitive advantages for Japanese firms (and compensate for the advantages of foreign fins) are not relegated to Japanese history. MITI, and other Japanese Government agencies that are genkyoku (sections of the bureaucracy with primary responsibility for developing and supervising policies for an industry), may have less ability to manipulate industries and the economy than they once had, but they still wield considerable power. Is Japan at a crossroads? Legions of writers have said so; one of the most popular themes of current writing on competitiveness is how much and how fast Japan is changing. In a sense, Japan has never stopped changing; policies that supported a particular industry or activity were shifting in the 1950s and 1960s as well as the 1980s. But the implicit corollary to the Japan is changinggenre is also that it is becoming more like us in ways that will make its industrial performance more like ours. At best, this is unproven; more likely it is a delusion. Japans government and private sector are still working, independently and together, to improve the competitive performance and market share of Japanese companies in a wide range of industries. They will probably succeed. INDUSTRIAL POLICIES IN TAIWAN AND KOREA Like Japan, Korea and Taiwan have used industrial policies to encourage the development of high-technology, high-wage industries. They, too, have been successful. Their successes indicate that industrial policies can contribute to industrial competitiveness under differing circumstances--in other words, that Japans industrial policies were not mere adjuncts to a culture that provides hothouse conditions for business. The Republic of Korea and the Republic of China (referred to throughout this chapter as Korea and Taiwan) have advanced remarkably fast, in comparison to developed and developing nations, and especially compared to what most observers expected. In 1965, Taiwan was the worlds 28th largest exporter of manufactured goods; in 1986, it was 10th. Korea moved up from 33rd to 13th. In 1989, Taiwan and Korea were, respectively, the fourth and fifth largest suppliers of manufactured goods to the U.S. market. Both nations are still poor compared with the developed worldtheir combined GDP is 1.5 percent of the free worlds GDPbut they have done remarkably well nonetheless at moving from manufacture of light industrial products requiring large contributions from low-wage labor to compete, to high-technology industries like computers and semiconductors. Taiwan is the worlds 10th largest producer of machine tools, with particular strength in low-end numerically controlled machines; Korea is the first developing country since Japan to make a strong debut in world automobile manufacture. Korean and Taiwanese industrial policies share many similarities, but there are important differences as well. They are similar in that they rely on long-term planning--overall visions of the directions of economic growth and developmentand use industrial targeting in addition to broader measures to encourage industrial activity generally. They educate their people superbly and share a cultural commitment to hard work. Finally, they both forced their companies to compete with the most proficient of world competitors, using competition abroad to provide the impetus for cost reduction and productivity improvement, while shielding them from competition at home. In Korea in particular, the protected home market was also used to make firms compete more effectively; the ability of firms to import needed inputs and machinery depended on their export performance. The differences are also interesting. The Taiwanese market has long been more open than Koreas and the industrial structure much less concentrated. Taiwanese firms have performed well across a broader range of industries than Korean fins, reflecting the choice of market niches that rely on standardized technologies that can be purchased and used effectively by small fins. Korea has organized production into large, conglomerate firms that have very few competitors at home and have performed well in many sectors where the economies of scale that large firms can gain are advantages, such as motor vehicles, consumer electronics, semiconductors. Both countries have had setbacks. Some attempts to develop industries or rationalize production failed, as was the case in Japan. But Japans success has also made the world a more difficult place for Taiwan and Korea; developed countries, afraid of what could happen to their own industries if
PAGE 41
34 l Competing Economies: America, Europe, and the Pacific Rim another Japan appeared, have been much less tolerant of Korean and Taiwanese policies like controlling currency values, protecting their own markets, and loose protection of intellectual property than was true for Japan. Both countries have, in response to increasing pressure from the United States and other trading partners, liberalized controls over their markets and currencies, and permitted more imports. Their own success has made it more difficult for them to pursue the policies responsible for success. Whether they can continue to develop, and raise their living standards above levels that are still only at the high end of poverty by the standards of developed nations, will probably depend as much on the performance of American and European economies as on their own. If America and the EC are successful in getting their own manufacturing back on track, by whatever standards they adopt, the world will be a more amenable place for developing countries, including Korea and Taiwan. If, on the other hand, American and European manufacturing continue to lose competitiveness, and only Japan gains, things could be different. Japans role in promoting world development is now larger than Americas, but would Japan be able to compensate for the retaliatory and self-protective policies likely to grow if American and European industries continue to lose competitiveness? It is possible, but perhaps unlikely; Japan, too, is concerned about the economic success of her neighbors, and continues to pursue industrial policies of her own in response to the challenge of these newly industrializing countries. GOVERNMENT SUPPORT OF LARGE COMMERCIAL AIRCRAFT More than most other civilian industries, the large commercial aircraft industry owes its existence to government policies in whatever countries it exists. The U.S. aircraft industry 62 is the largest in the world, and is the largest single trade-surplus category in Americas international accounts; in 1990, the United States ran a $13 billion surplus in commercial transport aircraft and parts. 63 This was not always so. For a few years in the early part of this century, European producers were more advanced technologically; without substantial support from the government, the U.S. aircraft industry might not have gotten off the ground. Today, government support is second to corporate strategy as a determinant of the strength of the industry in the United States, and it is an extremely important contributor to the European aircraft consortium, Airbus Industrie. Although support for aircraft is regarded by some as one of the more prominent failures of Japanese industrial policy, the Japanese are hardly out of the race yet, and government supports are an important reason. Government is a major player, but its role varies widely in America, Europe, and Japan. In the United States, most of the government support for the aircraft industry has been indirect, a byproduct of defense programs. Though their contributions have decreased considerably, military R&D and contracts for production of military aircraft are the largest government contributions to aircraft industry competitiveness. To be sure, civilian industrial breakthroughs have also given DoD programs many boosts as well; the contributions of military and civilian developments to modern jet engines, for example, are tough to disentangle. Also, military business does not always follow the business cycle, giving aircraft and engine companies some ability to maintain expensive R&D programs and staffs through economic downturns to which this industry like so many other durable goods industries, is particularly vulnerable in recessions. On the other hand, on the occasions when government spending on defense aircraft and NASA programs declines during an economic recession, the damage to the industry is substantial. The U.S. Government also has contributed to the civilian aircraft industry through nonmilitary programs. NASA is spending $800 million in fiscal year 1991 on R&D that can contribute to the civilian aircraft industry. Still, there are problems; NASAs programs are not designed to contribute most effectively to competitiveness, but instead serve a range of other purposes understanding of basic scientific principles of aerodynamics and materials, and the like. As a result, the programs make a smaller contribution to competitiveness than they could, dollar for dollar, if they were designed to improve competitiveness. Finally, military programs have become much more burdensome, and less of a contribution to civilian industrial needs, in the last decade. Intricate and intrusive DoD procurement rules, and changes in the funding of R&D, have shifted more of the burden of developing technologies for civilian aircraft to the companies own coffers, and diminish the usefulness of military contracts. 64 At the same time, military purchases are declining sharply.
PAGE 42
Chapter 1--Summary and Findings l 35 Whether government policies designed to serve other purposes could have similar beneficial effects on other industries is quite doubtful, and reliance on serendipitous externalities is not likely to be the salvation of American manufacturing. The aircraft industry is a special case in several ways. Its formidable development and R&D costs, capital intensity, strong scale economies, and heavy reliance on accumulated experience of integrating production of millions of components make it different from most other industries. Electronics industries share some of these attributes (high development and R&D costs, capital intensity, and in some sectors, scale economies that also make it hard for more than a very few producers to make a profit), but differs in other respects (the half-life of commercial technology is generally far shorter in electronics than in aircraft, for example). Several decades ago, the government employed policies similar to ones that it uses now in the aircraft industry to jump-start a U.S. semiconductor industry, and with similar results for U.S. competitiveness. The governments of some European countries and Japan are less willing than the U.S. Government to leave competitive outcomes to chance in the aircraft industry. France, Germany, Great Britain, and Spain have contributed billions of dollars over more than two decades to Airbus Industrie, a consortium of four companies (Aerospatiale, Deutsche Airbus, British Aerospace, and CASA), to enable them to produce large commercial air transports that would compete directly with the products of Boeing and McDonnell Douglas. Support for Airbus is hard to pin down, since the company itself keeps no records on the amounts contributed by the governments, but the low-cost loans given to the consortium members over the years have enabled the companies to overcome the substantial barriers posed by high development and capital equipment costs and long, uncertain payback periods. While the investments made in Airbus have, according to the best estimates, not been repaid and may never be fully recompensed, there are benefits: increased competition, which forced Boeing and McDonnell Douglas to lower prices; the direct benefits of high-wage jobs and increased exports; and, indirect benefits consisting primarily of spinoffs of technology. Considering these benefits, so far the four European governments seem to believe the investment in Airbus is worth what it costs. Whether such a belief would stand up to the rigors of a standard economic benefit-cost analysis is much less certain. Japans Government has supported its aircraft industry for more than three decades. The industry was targeted for special support and development in MITIs planning documents since the 1970s, yet Japan has never achieved MITIs original goal for the industry: to become an independent producer of large commercial jet transports. For this reason, the aircraft industry is often used as an example of a failure of Japanese industrial policy, and occasionally as an example of the eventual fate of all attempts at industrial policy. Closer examination reveals that neither the policies nor the industry are such failures as some of the more deprecatory analyses indicate. In 1989, Japanese companies produced commercial aircraft-related products worth $1.2 billion, 65 and exported half of it. Japanese companies have become major partners in new jet engine production, and continue to progress in other segments of the business. Their accomplishments owe much to Japanese industrial policies; the fact that the aircraft industry has not moved ahead as much or as fast as many other targeted industries is not an indictment of those policies. Instead, the aircraft industry illustrates the limitations of industrial policy in the face of formidable obstacles, including Japans unique constitutional limitations over much of the postwar period on the production or development of military aircraft. 1 We cannot be more precise than merchandise trade; figures for U.S. share of world manufactured exports and imports are not available. However, the majority of merchandise trade is in manufactured products. 2 The World Bank World Tables: 1989-90 Edition (Baltimore, MD: The Johns Hopkins University Press, 1990). 3 1982-84 dollars. 4 Real weekly wages have not fallen as much as real hourly wages, because hours worked per week have increased. 5 U. S. Department of Labor, Bureau of Labor Statistics, Comparative Real Gross Domestic Product, Real GDP Per Capita, and Real GDP Per Employed Person, Fourteen Countries, 1950-1989, unpublished data, p. 8. 6 The World Bank, World Tables, op. cit., pp. 8-9 and pp. 328-329. Figures are in 1980 U.S. dollar terms. 7 Robert Heilbroner, A Tune-Up For the Market, The New York Times Magazine, Sept. 24, 1989, p. 61. 8 Ibid., p. 75. 9 This is OTAs conclusion based on studies of Korean, Taiwanese, and Japanese industrial policies of the postwar period (see chs. 6 and 7). Many analysts disagree that these nations industrial policies were either effective or significantly responsible for their economic success. On the anti-industrial policy side, see, e.g., Gary Saxonhouse, What Is All This About Industrial Targeting in Japan? The World Economy, vol. 6, No. 3, 1987, pp. 253-273; Philip Tresize, Industrial
PAGE 43
36 l Competing Economies: America, Europe, and the Pacific Rim Policy Is Not the Major Reason for Japans Success, The Brookings Review 2, spring 1983, pp. 13-18; Richard Samuels, The Business of the Japanese State: Energy Markets in Comparative and Historical Perspective (Ithaca, NY: Cornell University Press, 1987). Among those who think that Japans industrial policy had an important effect on its growth see Ira Magaziner and Thomas Hout Japanese Industrial Policy (London: Policy Studies Institute, 1980); Chalmers Johnson MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925-1975 (Stanford, CA: Stanford University Press, 1982); John Zysman Governments, Markets, and Growth: Financial Systems and the Politics of Industrial Change (Ithaca, NY: Cornell University Press, 1983); M. Aoki, Information, Incentives, and Bargaining in the Japanese Economy (Cambridge: Cambridge University Press 1988); R. Lawless and T. Shaheen Airplanes and Airports: The Subtle Skill of Japanese Protectionism SAIS Review, fall 1987, pp. 101-120; K. Yamamura, Caveat Emptor: The Industrial Policy of Japan in P. Krugman Strategic Trade Policy and the New International Economics (Cambridge, MA: The MIT Press, 1986). 10 For instance, Japan recently signed a voluntary restraint agreement (VRA) with Korea, under which Korean exporters of apparel agreed to restrain their exports to Japan. See "Japan South Korea Work Out Knotty Knitwear Dispute, Japan Economic Journal, Feb. 11, 1988. 11 Daniel I. Okimoto, Between MITI and the Market: Japanese Industrial Policy for High Technology (Stanford, CA: Stanford University Press, 1989), p. 47. 12 Ibid., pp. 38-48. 13 Ibid., p. 49. 14 Guy Routh, The Origin of Economic Ideas, 2nd ed. (Dobbs Ferry, NY: Sheridan House, 1989), passim. 15 For a more complete discussion of the role of capital costs on industrial performance, see ch. 2 of this report and U.S. Office of Technology Assessment, Making Things Better: Competing in Manufacturing, OTA-ITE-443 (Washington, DC: U.S. Government Printing Office, February 1990), ch. 4 passim. 16 An eloquent explanation of how these strategies worked is in Robert Wade, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization (Princeton NJ: Princeton University Press, 1990). 17 Ryutaro Komiya, Masahiro Okuno, and Kotaro Suzumura Industrial Policy of Japan (Tokyo: Academic Press, 1988), p. 7. 18 See ch. 6 for more discussion of Japanese industrial policies. 19 The Framework Program is a set of EC-funded R&D programs. EUREKA is a publicly coordinated and supported R&D effort of 19 European countries, including the 12 members of the EC. See ch. 5 for discussion of the single market the Framework Program, and EUREKA. 20 Except where noted, this is the definition of industrial policy used in this study. 21 See ch. 8 and vol. II for additional discussion of the aircraft manufacturing industry in the United States, Europe, and Japan. 22 Including salaries of employees, R&D, and construction of facilities. 23 Unpublished budget figures, NASA. 24 Marie Anchordoguy, The Nature and Effectiveness of Japans postwar Industrial Policy, contractor report for the Office of Technology Assessment February 1990. 25 The World Bank World Tables (Washington DC: The World Bank 1989). 26 Kenneth Flamm, Policy and Politics in the Intcrnational Semiconductor Industry, Paper Presented to the SEMI ISS Seminar, Newport Beach, CA, Jan. 16, 1989. 27 Material about the STA in this paragraph is drawn from Flamm, ibid. 28 This refers to both the formal STA and the side letter that specified that foreign shares of Japans semiconductor market should grow. 29 The following account is drawn from Thomas Howell, Brent Bartlett, and Warren Davis, Semiconductors and Government Industrial Policy in the 1990s (Washington, DC: Dewey Ballantine, draft in progress), pp. 126-139. 30 There is some conflict on this point. MITI maintains that the foreign share is over 17 percent, while the U.S. Government and industry estimate the foreign share to be between 13 and 14 percent. The difference is that MITI counts (while U.S. Governm ent and industry do not): 1) captive sales-i.e., in-house sales of a chip of a type that is not offered for sale to outsiders and 2) private label sales of a chip made by a Japanese firm with a U.S. firms label. 31 Keiretsu are groups of Japanese companies that do much of their business within the group and hold significant amounts of each others stock, See ch. 6 for additional discussion. 32 Edward J. Lincoln, Japans Unequal Trade (Washington, DC: The Brookings Institution, 1990), pp. 62-67. 33 Budget of the United States: Fiscal Year1992, Part Two, p. 47. 34 U.S. Department of Commerce, Technology Administration Emerging Technologies: A Survey of Technical and Economic Opportunities (Washington DC: 1990); and Council on Competitiveness, Gaining New Ground (Washington, DC: 1991) 35 Not to be confused with two other groups with similar names: the non-statutory Council on Competitiveness in the office of Vice President Quayle, which is concerned with the effects of Federal regulation and a private Council on Competitiveness, composed of chief executive offices from industry, academia, and labor. 36 These bills include S. 121, as reported by the Senate Committee on Governmental Affairs in the 98th Congress; S. 1365 in the 99th Congress; H.R. 1338 and H.R. 2135 in the 100th Congress; and H.R. 1274 in the 101st Congress. 37 S. 1978 in the 101st Congress; S. 1233, as reported by the Senate Committee on Governmental Affairs in the 1OOth Congress. 38 Robert N. McCauley and Steven A. Zimmer, Explaining International Differences in Capital Costs, Federal Reserve Bank of New York Quarterly Review, summer 1989, pp. 7-28. See also OTA, Making Things Better, op. cit., ch. 3. 39 OTA, Making Things Better, op. cit., p. 58. 40 19 U.S.C. 2411-2420. 41 19 U.S.C. 2251-2254. 42 U.S. Trade and Development Program, 1990 Annual Report, pp. 27,29,55-61. 43 50 U.S.C. App. 2401-2420. This Acts authorization lapsed as of Oct. 1, 1990, but President Bush continued its provisions in force by invoking the International Emergency Economic Powers Act 50 U.S.C. 1702. See Executive Order 12730-Continuation of Export Control Regulations, Sept. 30, 1990. 44 This option is termed strategic technology policy and described in detail in OTA Making Things Better, op. cit., pp. 71-88. 45 S. 1233 in the 100th Congress; H.R. 3838 and S. 1978 in the 10lst Congress. 46 S. 1191 in the 10lst Congress. 47 Council on Competitiveness, Gaining New Ground, op. cit. 48 National Advisory Committee on Semiconductors, Toward a National Semiconductor Strategy, Second Annual Report (Arlington, VA: 1991), p. 15. 49 John W. Rutter, Direct Investment Update: Trends in International Direct Investment U.S. Department of Commerce, International Trade A dministration September 1989. The stock of investment given here is the position of foreign direct investors, or the value of the foreign
PAGE 44
Chapter 1Summary and Findings l 37 investors equity in and loans to offshore affiliates. For the United States, a foreign direct investor is one that owns or controls at least 10 percent of a companys voting stock (or an equivalent amount in an unincorporated enterprise). 50 1988 is the latest year for which data are available. 51 The latest figures are, unfortunately, for 1982; that year, foreign affiliates of U.S. parent companies accounted for less than 9 percent of the companies total R&D. 52 In the late 1970s, manufacturing affiliates of foreign parents had much higher investment in plant and equipment as a percent of sales, than U.S. manufacturers; by the early 1980s, most of the difference had disappeared. U.S. affiliates of foreign multinationals also pay slightly higher compensation than American companies on average, but that is due to relatively heavy concentration of investment in high-wage industries, like banking, and not to higher wages in comparable establishments. 53 In the past, the trade surplus generated by U.S. direct investment abroad was also substantial, but it has dwindled. In 1988, U.S. parent companies were associated with a merchandise trade surplus of only about $8 billion. 54 It is important to remember what makes a firm foreign. Obviously, any wholly owned subsidiary of a foreign firm qualifies. In the United States, a firm is counted as an affiliate of a foreign parent if one foreign parent company controls 10 percent or more of its voting stock, or an equivalent stake in an unincorporated enterprise. 55 The 12 members are Belgium, Denmark France, Germany, Greece, Ireland, Italy, Luxembourg, The Netherlands, Portugal, Spain, and the United Kingdom. 56 See, for example, Merton J. Peck, Industrial Organization and the Gains from Europe 1992, Vittorio Grilli, Financial Markets and 1992, Richard N. Cooper, Europe Without Borders, and Rudiger Dombusch, Europe 1992: Macroeconomic Implications, all in William C. Brainard and George L. Perry (eds.), Brookings Papers on Economic Activity 2 (Washington, DC: The Brookings Institution, 1989), pp. 277-381 passim. Peck for example, says, It is important to emphasize this overoptimism, given the significance of the report both as a work of economic advocacy and as an impressive scientific study. Peck p. 278. 57 Peck, ibid., p. 277. 58 Dombusch, op. cit., p. 353. 59 While it is not a statutory requirement, most nations distinguish foreign from domestic goods by a rule of origin that states that the good comes from the nation where the last substantial transformation is made. The EC used this rule for semiconductors for many years, but recently changed it to state that semiconductors originated where the most substantial transformation which the EC defines as diffusion, is made. 60 The ECU 5.7 billion for the Framework Program is all public money; it is generally matched by the projects participants, except for certain nonprofit organizations such as universities, which can receive up to 100 percent of project costs from the European Community. EUREKA is primarily funded by the private sector; less than 10 percent of its funding is public money according to the U.S. International Trade Commission. A rough calculation assuming that 10 percent of EUREKA costs are public and that both programs spend about the same amount of money each year, yields annual public funding of cooperative research amounting to ECU 1.6 billion, or about $1.9 billion. Source of funding numbers is U.S. International Trade Commission, 1992: The Effects of Greater Economic Integration Within the European Community on the United States: Second Followup Report, USITC Pub. No 332-267 (Washington, DC: September 1990), pp. 16-6 and 16-10. 61 See, for example, Marie Anchordoguy, Japanese Policies for the Supercomputer Industry, contract report for the Office of Technology Assessment, February 1991; Marie Anchordoguy, Computers, inc.: Japans Challenge to IBM) (Cambridge, MA: Council on East Asian Studies, Harvard University, 1989); and Kenneth Flamm, Targeting the Computer: Government Support and International Competition (Washington DC: The Brookings Institution, 1987). 62 Throughout this assessment, the large commercial aircraft industry will be referred to as the aircraft industry. Only when specified does the term refer to all segments, including military aircraft. 63 This is unusually high; in 1990, a strike at British Aerospace delayed deliveries of some Airbus aircraft purchased by U.S. airlines. Aerospace Industries Association, "1990 Year-End Review and Forecast An Analysis, addendum to Aerospace Facts and Figures 90-91 (Washington, DC: 1990). 64 This should not be read as an unqualifled indictment of military procurement and R&D contracts. The changes over the past decade were made for understandable reasons--i.e., avoiding fraud and reducing costs-and this study did not look into how the changes have performed in doing what they were designed to do. For further information on that topic, see U.S. Congress, Office of Technology Assessment, Holding The Edge: Maintaining the Defense Technology Base, OTA-ISC-420 (Washington DC: U.S. Government Printing Office, April 1989). 65 The conversion rate is Y130 to $1.
PAGE 45
Chapter 2 Policy Issues and Options
PAGE 46
CONTENTS Page A TECHNOLOGY-FRIENDLY ENVIRONMENT . . . . . . . . 44 The Financial Environment . . . . . . . . . . . . . 44 Human Resources . . . . . . . . . . . . . . . 45 Technology Diffusion . . . . . . . . . . . . . . 47 TRADE POLICY TOOLS . . . . . . . . . . . . . . 52 Responses to Foreign Market Barriers . . . . . . . . . . 53 Use of Protection . . . . . . . . . . . . . . . . 55 Domestic Content Requirements . . . . . . . ,. . . . . 57 Export Promotion . . . . . . . . . . . ....*.,? . . 58 Export Controls . . . . . . . . . . . . . . . . 59 GOVERNMENT-INDUSTRY PARTNERSHIPS . . . . . . . . . 62 Strategic Technology Policy . . . . . . . . . . . . . 62 Financial Risk-Sharing . . . . . . . . . . .*....., ....*, 65 Government Purchases . . . . . . . . . . . . . l ....., 69 INSTITUTIONS FOR A STRATEGIC COMPETITIVENESS POLICY . . . 71 Selecting Industries . . . . . . . . . . . . . .,*..... 73 What Kind of Support? . . . . . . . . . . . . . . 74 Institutional Alternatives . . . . . . . . . . . . . . 76 Boxes Box 2-A. 2-B. 2-C. Page The Baldrige National Quality Award . . . . . . . . . . 51 The Advanced Technology Programs F irst Round . . . . . . . 66 National Advisory Committee on Semiconductors . . . . . . . 73 Table Table Page 2-l. Comparison of National Critical Technologies with Department of Cornmerce Emerging Technologies and Department of Defense Critical Technologies . . 75
PAGE 47
Chapter 2 Policy Issues and Options The U.S. approach to promoting particular industries has been mostly not to do it. Our underlying belief is that the national economic interest is best served by free and fair competition in the marketplace, at home and abroad. Whatever industry and trade profile results from this competition is in the national interest, and government interference is justified only when the competition is unfair or when national security is at stake. In practice, the government has intervened from time to time to support or protect certain industries. But this has been mostly ad hoca response to political pressures, not part of a strategy to build up competitive industries in areas of special importance. Does this approach still make sense in a world where governments in most advanced industrial nations, including those of our most able competitors, are cooperating with private business to promote critically important industries? This study has concluded that, on the whole, free trade and vigorous competition are worthy, indeed essential goals, that must be steadily pursued. It has also found evidence from Japan, Korea, and Taiwan that selective, flexible government support of particular industries can pay off in rapid advances in technological achievement, export success, and national income. The European Community, preparing for the Single Market in 1992, is developing its own versions of support for critical industries within the Community. The core issue in this chapter is whether the United States, with its particular traditions and form of government, can combine selective promotion of strategic industries with a firm overall commitment to competition and free trade, and if so, what institutions and policy tools would be needed. It can be argued that the U.S. form of government is open at so many places to capture by special interests that the chances of rational gov ernment decisions on aid to selected industries are slim. The good that government-industry cooperation might do in strategic fields (e.g., electronics) where we have lost competitive advantage could dissipate in expensive, useless handouts to industries whose main claim to special treatment is political clout. This argument, however, can be stood on its head. Despite our free trade philosophy and comm itment to open markets, the United States does make exceptions; it does grant some trade protection, and does sometimes subsidize favored economic sectors (e.g., price supports for farm products). Nearly always, these exceptions are motivated by politics. It is possible that a coherent strategy that first selects industries for support based on their contribution to the national good and then applies a judicious mix of supportive measures, including trade protection if needed, would result in more rational exceptions to a general regime of free trade. Trade protection is only one part of the mix in a policy to promote strategic industries and is not by any means the dominant one. Trade protection on its own is unlikely to improve competitiveness, and in fact can have the opposite effect. If the only government action taken is to shield companies from competition, they often do not (except where there is still vigorous domestic competition) have the proper incentives to invest in technology development and diffusion, worker training, plant modernization, and other things that improve competitiveness. Another essential part of the mix is government partnership with industry in technology development, especially of high-risk technologies where the potential benefits to society are great but the prospect of returns to individual firms is too small or remote to justify a big investment. Still another part encompasses measures that spread the financial risks of adopting advanced product and process technologies to government as well as industry; such measures include indirect means, such as tax breaks that favor technology adoption. Beyond these industry-specific strategies we broader government policies that shape the overall environment for all U.S. industries and affect their competitiveness. Taxes, spending, and the Federal budget deficit directly influence the cost of capital to firms. They wield indirect but powerful effects on our ability to compete through their influence on the growth and stability of the domestic economy and the international value of the dollar. Equally critical to industrys performance is the education of the Nations children (a government responsibility) and the reeducation and trainin g of adult workers (a responsibility shared by government and industry). Important as these policies are, this report cannot do -41
PAGE 48
42 l Competing Economies: America, Europe, and the Pacific Rim full justice to them all. Options to improve the financial environment for industry; to upgrade the training and education of managers, engineers, and workers; and to diffuse advanced and best practice technologies throughout industry are s ummarized here but have been covered in greater depth in earlier OTA reports. l This chapter gives more detailed consideration to strategies that select and combine trade, technology, and fiscal or financial policies to improve the competitiveness of particular U.S. manufacturing industries. The focus is on manufacturing, not only because it dominates international trade, but also because it pays for most privately funded R&D and provides well-paid jobs, not only in manufacturing but also in related service sectors. The first question to ask is whether there is any need for exceptions to the Nations free trade and free market policies. The answer starts with the strong evidence, in this report, other recent OTA reports, and elsewhere, that U.S. manufacturing is in trouble. 2 Moreover, certain industries that most people regard as essential to the further technological advance of the American economy and to rising standards of living are in trouble. This is true, for example, in parts of the semiconductor and computer industries. On the other hand, the United States is not in the position Japan was 40 years ago, when that nation was behind in nearly all advanced technologies and manufacturing industries and was prepared to make sacrifices for many years to catch up. Our troubles are not those of a poor or war-ravaged country but of a rich country that has lost its edge. All this suggests that it may make sense to offer government help to particular industries. The U.S. Government has made several starts in that direction through technology policy, most notably in the Sematech project, where the government has gone halves with industry to develop better manufacturing processes for dynamic random access memory (DRAM) semiconductor chips, at a cost to the government of $500 million over 5 years. Yet in some critical cases, and major parts of the microelectronics complex may be among them, technology assistance is not enough. U.S. electronics companies lost their lead to Japanese competitors partly because of the Japanese fins manufacturing excellence, but also partly because of support from the Japanese Governments trade, technology, and financial policies. Government support helped to create the large, rich, integrated Japanese electronics firms of the 1990s, whose deep pockets and command of certain critical technologies keep them a pace ahead of much of the U.S. industry. There is good reason to believe that the U.S. semiconductor industry will continue to lose ground without some trade policy help from the U.S. Government. And indeed, it has already received some trade relief as part of a broader arrangement with Japan, but while U.S. firms have gained a few percentage points market share in Japan, mostly the results have been disappointing. 3 Nevertheless, the trade agreement and Sematech were unusual, perhaps unique, attempts by the U.S. Government to strengthen a domestic industry of strategic importance. An advantage to the strategic approach to trade policy is that it could be proactive and planned to avoid pitfalls. And it could be based on a frank appraisal of the needs and interests of the United States without casting blame on other nations. In most cases, U.S. trade policy actions are based on findings that foreign firms are competing in ways labeled unfair: that they are dumping in the United States (selling below cost or below the prices they charge in protected markets at home), or that they are taking subsidies from their governments and thus undersell U.S. fins. There often is truth in these charges. But often the whole truth is more complex, including both genuine superiority in quality and price of the foreign goods and industrial policies of foreign governments that help their firms lower production costs and acquire new technology. Dumping or subsidies may be only a part, sometimes a minor part, of the problem. To require all U.S. trade policy actions to be based on findings of unfair trade is an irritant to amity among otherwise friendly nations and is particularly galling to our trading partners when it brands their governments industry policies as unfair. Imposing limited, conditional, and temporary trade restrictions as part of an overall U.S. strategy to strengthen vital domestic industries is less abrasive and, under certain limited conditions, is legal under current international agreements and U.S. trade law. If the U.S. Government should opt to develop a strategic competitiveness policy, two essential conditions would have to be met. The first is an institutional capacity to plan and carry out the policy. At present the United States lacks this capacity. Responsibility is diffuse in the executive branch even for carrying out our present policy of urging free trade and threatening sanctions for unfair
PAGE 49
Chapter 2-Policy Issues and Options l 43 practices. Judging by experience in the Pacific Rim and Europe, it is more effective to place the principal control over selective support for strategic industries in the hands of one or two strong agencies than to scatter it among many. Given enough interest in Congress and backing by the public, it is possible to create anew Federal agency with powers that did not exist before; witness the creation of the Environmental Protection Agency in 1970. Although an agency with power and prestige equal to that of Japans MITI is highly improbable-and perhaps undesirable--in the American setting, it might nevertheless be possible to create an agency with the lead in competitiveness policy, a seat at the table in Cabinet meetings, and the ear of the President. An alternative to creating a new lead agency is to use existing institutions, but assign them a much clearer mission of promoting the competitiveness of American industry. U.S. trade law already allows some leeway in pursuing this aim. What would be needed among the Federal agencies involved in trade, technology, and financial policy is a stronger understanding of which industries are critical to national economic security, better analysis of their competitive situation, closer relations with industry, and coordination of policy to support them. This brings us to the second condition. Any strategic competitiveness policy, whether directed by a lead agency or coordinated among several lesser ones, needs wholehearted support from both Congress and the Administration. Congress can move independently in the direction of support for strategic industries. It has done so repeatedly in the past few years; for example, by creating the Sematech project via appropriations for the Defense Advanced Research Projects Agency (DARPA); by awarding DARPA funds to advance other technologies of commercial as well as military interest (e.g., flat panel displays for computers); and by setting up a purely commercial Advanced Technology program in the National Institute of Standards and Technology (NIST), under which government can take part with industry in R&D consortia. Recently, the Bush administration has moved toward supporting a government role in developing generic technologies that could have commercial application. If Congress were so inclined, it might by resolution declare its intent to adopt policies in support of selected industries. It is harder to conceive that Congress could unilaterally create both the capacity and the will within the government to carry out competitiveness policies. Backing for such policies by both branches of government must rest on the understanding and support of the American people. There would have to be widespread comprehension that our national economic security is at some risk, agreement that government support for critical industries could be fruitful, and acceptance that government collaboration with industry to regain excellent performance in manufacturing is not cost-free. Trade protection always costs consumers. Programs for partnership between government and industry to develop new technologies or share the financial risks of adopting them cost the taxpayer something. More basically, massive Federal budget deficits, combined with our low personal savings rate, exert a steady upward pressure on interest rates that hikes the cost of capital for all U.S. industries. And the budget deficit cannot be greatly reduced without the pain of higher taxes, cuts in favorite government programs, or both. One explanation for the success of the Japanese government-industry partnership is the existence of a high-growth consensus among consumers, workers, and taxpayers as well as government and business leaders, and a willingness to make sacrifices for that purpose. 4 Sacrifices of current income allowed long-term investments in technology, capital equipment, and human resources, which in turn helped to produce the large steady rises in income that Japanese citizens have enjoyed for more than 40 years. The same kind of consensus in the United States is a condition for the adoption of new government policies to restore our national competitive performance. And the same kind of future rewards-faster economic growth, increasing incomes for most citizens, healthy growth of well-paid jobs--could be expected for American citizens. The following sections consider, first, options to foster a supportive environment for technology development and adoption. Next are options to carry out a comprehensive competitiveness policy in two areas: trade policy attuned to competitiveness needs and government-industry partnerships. Many of these options could be considered on their own merits, whether or not a more comprehensive policy is adopted. Last, there are options for new or altered institutions that would be needed to plan and implement a strategic competitiveness policy.
PAGE 50
44 l Competing Economies: America, Europe, and the Pacific Rim A TECHNOLOGY-FRIENDLY ENVIRONMENT OTAs earlier report, Making Things Better: Competing in Manufacturing, found that U.S. manufacturers have lost out in one industry after another to competitors who are able to make things better to produce quality products at lower cost. And the key to this better performance is technology. Defined broadly, technology includes not only new products and advanced production machinery but also efficient organization of work and effective use of people. Industry and government both have roles in building a better technology base for U.S. manufacturing. The report delineated four areas in which government could contribute: Improving the financial environment for U.S. firms, by taking action to reduce capital costs and relieve other pressures in the financial markets to show high profits every quarter. Focus on short-term profits at the expense of longer term investments in advanced equipment and new technologies has hobbled U.S. competitiveness. Upgrading education and training of the managers, engineers, technicians and workers needed in manufacturing. U.S. manufacturing suffers not only from the well-known defects in American public education but also from failures by managers to train workers and organize work to make best use of workers abilities. Beyond improving public education, government can also help in the training of workers and managers and the education of manufacturing engineers. Diffusing technologies throughout the manufacturing sector. Many companies, especially small ones, are slow to take advantage of modern production equipment and manufacturing processes. Government technology extension services in several States and foreign countries have shown they help manufacturers select and learn to use up-to-date technologies. Other devices for technology diffusion might include a government-subsidized equipment leasing system, or easier access to technologies developed in government labs. Forming a strategic technology policy to promote the development of new technologies with commercial promise through collaborative ventures with industry. Several of the options outlined above would benefit all American manufacturing, with no distinctions in kind. A hospitable environment for the generation and adoption of new technologies throughout manufacturing is needed regardless of more targeted efforts to nurture particular technologies or industries. The option for a strategic technology policy is one of those more targeted efforts, and is discussed in a later section. The discussion immediately below covers the more general options to lift the performance of U.S. manufacturing as a wholes The Financial Environment Americas financial climate is not hospitable to long-term investments in new technology and production equipment. High U.S. capital costs favor short-term profits over long-term investments, as do pressures from the stock market. Recent studies provide solid evidence that U.S. capital costs have been substantially higher than those of Japan and Germany for more than a decade, through 1988. 6 Capital costs are influenced by interest rates, the economic depreciation of investment and its tax treatment, and other fiscal incentives for investment. In the United States, government policy has contributed to high interest rates, particularly in recent years. The combination of high Federal budget deficits and low personal and business savings rates has kept a relentless upward pressure on interest rates. Congress and the Administration agreed on some genuine budget discipline in late 1990, but the 1991 deficit was still projected to climb to an all-time high (declines over the next 4 years were also projected). The dilemma is that some specific fiscal measures that might help firms to modernize and invest in new technologies would also tend to worsen the budget deficit, because they involve raising tax expenditures or lowering revenues. The budget agreement of 1990 forbids this, unless there is a compensating rise in tax revenues or decline in spending elsewhere in the same part of the budget (domestic non-defense programs). 7 The serious efforts in the budget agreement to curb the deficit make sense from the standpoint of improving competitiveness since, otherwise, additional measures to reduce capital costs would probably have no more than a marginal effect. If Congress wishes to lower capital costs
PAGE 51
Chapter 2-Policy Issues and Options l 45 through tax breaks, the difficulty will be to find something else to cut in the nondefense budget, or get agreement to raise other taxes in compensation. Eventually, economic growth based on stronger competitive performance could help to ease the budget problems, but in the short run there would be a price to pay. Increased savings could help to ease the upward pressure on interest rates and thus lower capital costs. A combination of carrots, sticks, and appeals to patriotism might induce greater savings. Congress could consider inaugurating a national savings campaign that rewards regular savings. For carrots, one option would be to offer government bonds at an attractive, guaranteed interest rate to people who sign up for a regular program, such as a payroll savings plan. Another option might be a tax reduction on the interest income from payroll savings. The sticks would be policies to discourage consumption. One option is a consumption tax, crafted to escape the severe regressive effects of a flat tax either by taxing necessities lightly (or not at all) and luxury items heavily, or by granting substantial exemptions. A consumption tax might serve the dual purpose of encouraging savings while raising tax revenues. Another option would be to limit tax deductions for interest on home mortgages more strictly than the law already does. Although home equity is a form of savings for householders, these savings are tied up and unavailable for capital investments in manufacturing industries. Tax breaks could help to lower the cost of capital to industry even if interest rates remain high. Congress could consider several options for tax inducements for technology development and capital investment. The United States has a long, but inconsistent, history of offering investment tax credits and accelerated depreciation to promote capital investment. There is evidence that these measures do have positive effects, though the magnitude is a matter of debate. They would undoubtedly be expensive; they cost tens of billions of dollars a year in lost tax revenues when they were in force. 8 The tax credit for R&D is far less expensive (amounting to hundreds of millions in forgone revenue per year) but it also has less effect on competitiveness; while it rewards technology development, it has little effect on the actual adoption of new technologies in manufacturing practice. As with the investment tax credit, it is hard to pin down how much extra R&D is due to the tax credit. Yet despite the uncertain payoffs, the arguments in favor of both kinds of programs are substantial enough that they deserve a careful hearing. It is worth repeating that under the current budget agreement (lasting through fiscal year 1994) such measures cannot be adopted unless the revenue losses they entail can be made up in some other way. Incentives to hold investments longer might take some of the pressure off managers to focus on short-term profits. One option would be to create a capital gains tax that favors long-term gains and penalizes short-term turnover of holdings. A variable rate scheme might be adopted, with a high tax rate (say, 50 percent) for gains on assets held less than 1 year, and the rate declining through several steps to a low level (perhaps 10 or 15 percent) after 5 or 6 years. The measure would be most effective if its application were extended to pension and other funds that are now tax-free, since these funds hold one-third of the stock and probably account more than half of the transactions on the capital markets. Stability in the economic and political environment is a great asset to business, and has been well provided in Japan and Germany, much more so than in the United States. Germany has successfully concentrated on keeping prices and exchange rates stable. In Japan, business has benefited not only from long-sustained economic growth and low inflation but also from policymakers sensitivity to the effects on business of macroeconomic changes. For example, in the mid-1980s when the yen suddenly rose greatly against other currencies, the Japanese Government made low-interest loans easily available to firms (especially small ones) so they could ride out the period of adjustment. American manufacturers penalized by the very high value of the dollar in the early 1980s got no such help. Human Resources Success in manufacturing depends on having well-trained people, comfortable with the demands of advanced technology, at every level from the managers office to the shop floor. In frost class competitive industries, production workers must sharpen reading and math skills, take more responsibility, cooperate more closely with others, and understand their own roles in the entire production system. In other words, more is being demanded of workers. At the same time, the typical American
PAGE 52
46 l Competing Economies: America, Europe, and the Pacific Rim education leaves many young people ill-prepared for work, and training of the people already in the work force is equally deficient. Our major trade competitors (i.e., Japan and Germany) offer more and better training, both to young people preparing for work life and to active workers. The situation in the United States is likely to get worse before it gets better. About half the new entrants to the U.S. work force up to the year 2000 will be members of minority groups, and two out of five minority children grow up in poverty. Poor children drop out of school disproportionately and many never learn the skills they need for productive work. A great deal of public policy attention is being given to the critical need for improved public schooling. But even if help arrived tomorrow, it would be many years before the results showed up in the work place. Meanwhile, another approach is government-industry programs to offer better training to those already in the work force. This would include not only financial commitments from both government and industry but also a management style that gives workers a real stake in the enterprise and real responsibilities for quality and efficiency, and organizes work to take advantage of workers training and abilities. Some of the options for training active members of the work force are as follows: 9 Government encouragement to industry to train workers encompasses several possibilities. The most aggressive and far-reaching of these options, which guarantees more training without any direct cost to the government, is a payroll-based training levy. Employers would have a choice between spending a certain amount on training their workers (say 1 percent of their payroll) or paying the same amount into a national training fund. Several foreign countries (including France, Germany, Ireland, and South Korea) use the system, and four States in the United States (California, Delaware, Rhode Island, and Alaska) raise training funds through a small payroll-based levy. A less pervasive option is government technical assistance to help trade associations or joint labor-management groups identify industry-wide training needs and acquaint their members with training materials and best practice approaches. U.S. trade associations are far less active in training than their European counterparts, yet it is a low-cost way of wholesaling training information to individual firms. Cost-sharing by governments could get the activity started. A similar option is for the government to offer small grants to help firms set up training consortia. Small companies could share the costs of instruction and facilities, and large ones could use consortia to help their supplier networks develop common training appreaches. 10 Training delivered to individual workers includes several existing government-funded programs that need either revitalizing or more financial support to fulfill their promise. Apprenticeship can also bean excellent way for workers to get real skills training with recognized credentials. The American apprenticeship system once served industry and a small segment of the work force well, but fell into decline in the 1980s with cuts in funding and staff for the Labor Departments Bureau of Apprenticeship and Training, and shrunken support from industry. The Federal vocational education program has many excellent features including school-to-work programs, cooperative business-labor-education training programs for active employees, and some support for apprenticeships. The problem is finding enough money to support these programs in a meaningful way. For example, a demonstration workplace literacy program has generated keen interest with hundreds of application, far more than the program could accommodate. Training linked with technology assistance is an effective combination. Many U.S. manufacturing fins, especially small ones, have trouble adopting new technologies. One source of the trouble is inadequate training. The United States is far behind many other countries in helping individual firms learn about and use new technologies, but some States and a small Federal program are making efforts in this direction (as discussed more fully below). The best of these programs integrate training into their industrial extension efforts. The National Institute of Standards and Technology (NIST) in the Commerce Department is the agency mainly responsible for the Federal Governments modest efforts in technology diffusion. It could link training with technology in its own programs and those of other Federal agencies, and could also serve as an information clearinghouse for State agencies trying to do the same thing. Improving the quality of worker training is, in part, a Federal responsibility. The military is the largest training institution in the United States. Besides using conventional classroom and on-the-
PAGE 53
Chapter 2-Policy Issues and Options l 47 job trainin g, the armed forces have developed instructional technologies that might be adapted and transferred to civilian use. Although the law calls on the Department of Education to take on this task, no funds have been provided for it. The Department and other Federal agencies might also take a more active hand in testing and evaluating workplace training, with particular attention to computer-aided training technologies. Education and training of engineers is a central issue for competitiveness. Although there seem to be plenty of engineers now (the United States has about as many engineers per thousand workers as Germany and Japan), the supply could dwindle a few years hence because of the demographic facts. The group most inclined to enter engineering, white males, is shrinking gas a proportion of children in the educational pipeline. In the long run, an adequate supply of engineers depends on our success in giving all our children a good education, including better math and science education starting in the early grades. Meanwhile, Federal scholarships and grants that encourage minorities and women to take up science and engineering careers seem to be getting results and deserve support. Retraining of midcareer engineers is another way to enlarge the supply over the next few years, especially as many engineers are losing jobs in defense industries, with the cutback in military spending. With government support, retraining courses might be targeted to fit the needs of commercial manufacturing. Problems of manufacturing are generally neglected in university engineering departments. The elitism of U.S. engineers and their remoteness from the shop floor are weaknesses of American manufacturing. While this is primarily an issue for managers, government might support education and research in manufacturing engineering. One option might be to create a Manufacturing Sciences Directorate in the National Science Foundation. Technology Diffusion U.S. institutions for diffusing new technologies throughout manufacturing are thin. Even large firms with sufficient resources often neglect to take what they could from outside the firm. Many of our 355,000 small and medium-sized manufacturing firms are worse off, with only scant and spotty exposure to new technologies. While some small fins, such as Silicon Valley startups, are at the technological forefront, a great many others find the effort to keep informed beyond their means. The part played in U.S. competitiveness by these bread-andbutter small manufacturing firms can be critical. Large auto companies, for example, must depend on their myriad suppliers to deliver well-made parts and components on time. Both public and private means of diffusing technology are weak in the United States. It is uncommon for large U.S. manufacturers to lend technical assistance to their suppliers, something that is everyday practice in Japan. And there is little in this country to compare with the network of free, public technology extension services for small and medium-sized manufacturers that blankets Japan. Nor do we have anything like the apprenticeship system that trains more than half the young people in Germany and Sweden and produces a high level of skills in the work forcr--a key factor in diffusing new technologies throughout manufacturing in those countries. Throughout the past decade, Congress has taken actions aimed at transferring advanced technologies from lab to factory, bringing small and mediumsized firms up to date in best practice manufacturing technologies, and removing legal barriers that might interfere with technology advance in manufacturing. Some of these actions are well along. Others have just begun. Congress may wish to strengthen and expand some of the more promising existing programs and consider adding others. No one program, by itself, is likely to improve U.S. manufacturing performance dramatically or rapidly. Some may fail. But given time to prove themselves, several of these measures in combination could pay off in real contributions to competitiveness. 11 Technology extension has the potential to improve the manufacturing performance of small and medium-sized American fins. Defining industrial extension as one-on-one technical advice given to individual fins, 16 States had real extension programs (including field agents) in early 1991, and another 7 had technology demonstration or assistance centers. At a rough guess, total spending for 27 extension or demonstration programs in 23 States amounted to about $50 million. Most of these programs are new, although a handful, such as Georgia Techs Industrial Extension Service, have years of solid experience. Federal industrial exten-
PAGE 54
48 l Competing Economies: America, Europe, and the Pacific Rim sion is still smaller and newer. A program of Manufacturing Technology Centers was established in the 1988 Trade Act, and by 1991 included five centers, with another planned for 1992. Fiscal year 1991 funding was about $12 million, with nearly $14 million proposed for 1992. The Centers are required to get matching funds from the States or private sources and, by law, cannot receive Federal funds for more than 6 years. Because many States were in financial distress in 1991, interest in supporting new Centers had declined somewhat. 12 For perspective, compare these few and scattered programs with those in Japan. Besides the nationwide system of 185 technology extension centers, funded at about $500 million (half from the national government and half from prefectures), many Japanese cities, wards, and other localities support industrial halls that offer similar services. In addition to individual technical advice from field agents, these services include regular workshops on common manufacturing problems, use of specialized or costly equipment at low fees, demonstrations of new technologies, and referrals to expert consultants for advice on difficult problems. There is room for expansion of both State and Federal technology extension services in the United States. States, being closer to clients, may be more in touch with local needs. On the other hand, various kinds of manufacture tend to be regionally concentrated, and Federal agencies can more easily span State lines. Moreover, while some States do an excellent job of delivering services, some are less adept. A Federal effort could help to set an acceptable, consistent quality of service. Supposing Congress wished to support a wider network of technology extension centers, it might set a rninimum goal of providing services in some 120 centers to 24,000 small and medium-sized firms nationwide per year-i. e., about 7 percent of the nations 355,000 manufacturing firms with fewer than 500 employees. This would cost about $120 to $480 million per year, depending on the level of service. Some of the funds could come from State or private sources, though it may be unrealistic to demand that these sources take over all the funding within a few years (as Federal law now provides in the case of the Manufacturing Technology Centers). A program of this size is modest and might be overwhelmed with requests for assistance once the centers gained a good reputation. The State of Georgias highly regarded industrial extension service, run by Georgia-Tech, serves a similar proportion of its manufacturers; it does not advertise for fear of being swamped with requests. It is worth noting that the U.S. Agricultural Extension Service is funded at $1.2 billion per year (with about $370 million, or 30 percent, coming from the Federal Government), and has 9,650 county agents, 4,650 scientific and technical staff, and offices in nearly every county in 50 States. Agriculture accounts for 2 percent of the U.S. gross national product, manufacturing 19 percent. Government financial aid tied to improvements in technology can be an effective means for modernizing manufacturing. It has not been much employed in the United States (except to induce investments in pollution control equipment), but has been widely and ingeniously used in Japan. One option with at least two distinct advantages is a system that would allow manufacturers to lease modern production equipment, or buy it on the installment plan, at subsidized rates. This scheme would not only encourage manufacturers to use up-to-date equipment, such as computer numerically controlled (CNC) machine tools; if the system bought U.S.-made equipment it could also benefit U.S. builders of the machinery by offering a stable, assured market for part of their output. An equipment leasing system for CNC machine tools, for example, could start with modest government funding, probably about $3 million per year. 13 It might be open only to small manufacturing fins, or could be open to all, with lower rates for small fins. It is also possible to make government-backed financing to small manufacturing firms conditional on a technical assessment. However, that presupposes a nationwide, readily available industrial extension service, which does not yet exist in this country. Moreover; Federal programs of financial support for small business are not large enough to reach many fins. All the programs combined (including loans, loan guarantees, and investments in development corporations) amount to about $3.5 to $4 billion per year for every kind of small business, not just manufacturing fins. In Japan, the required technical assessment is common practice, and there it does matter. Although exact comparisons are not possible, we do know that Japanese loans and loan guarantees to small firms are at least 20 times greater than similar U.S. financial aid, and the level of subsidy is higher (some government loans for modernizing equipment are interest-free).
PAGE 55
Chapter 2--Policy Issues and Options l 49 As of 1988, nearly 11 percent of all outstanding loans to Japanese small businesses for plant and equipment investment were from government financial institutions, a big enough chunk to make required technical assessments significant. 14 Another option is tax incentives for investments in advanced manufacturing equipment. This is another of the many inducements the Japanese Government offers to businesses, especially small ones, to modernize. For example, Japanese tax law was changed in 1984 to allow very rapid depreciation by small firms of high-technology (mechatronic) equipment, including CNC machine tools. This set off a flurry of buying known as the NC-ization period. Commercialization of technology from Federal laboratories is a goal that Congress has pursued for more than a decade through laws and oversight. There has been progress, especially in the last year, but the goal is not yet fully realized. The U.S. Government spends about $23 billion per year for R&D in Federal laboratories; only a minor portion of this is of commercial interest. Much of the labs R&D is for advanced development of weapons, and a large share is for basic research quite far from the commercial arena. However, some lab results have proven useful to civilian industry, and companies have benefited from using specialized lab facilities (e.g., the Synchrotrons Light Source at Brookhaven National Laboratory and the Combustion Research Facility of Sandia National Laboratories). Two principal ways of bridging the gap between lab research and commercial manufacturing are: 1) granting firms some exclusive rights to inventions from the labs, so it is worthwhile to invest in commercializing the technologies; and, 2) creating cooperative R&D projects, in which industry puts up half the money and is involved in the planning from the beginning. Two obstacles on the government side have been too little money and too much red tape. With other missions taking priority, lab funding of technology transfer has been scanty. Bureaucratic hoops in the parent agency, especially delays for legal review, have often stalled technology licensing and the conclusion of cooperative agreements between the labs and private industry for many months, sometimes a year. 15 Congress has taken steps to cut the red tape and provide more funding for technology transfer from the labs. Since the passage of the National Competitiveness and Technology Transfer Act of 1989, technology commercialization has received more attention at the Department of Energy (DOE) and its labs. The high-level Technology Transfer Project Group has focused on streamlinin g approvals for both licensing and cooperative R&D. Moreover, DOE has extended to some labs more control over patents and data rights from lab research, permitting a faster track to licensing technologies with commercial appeal. l6 For its part, industry has become more aggressive at seeking out lab technology .17 The heightened attention to commercialization has yielded some results. Cooperative R&D has increased. 18 But despite the progress, funding is still limited. DOEs Defense Programs devoted only $20 million, or 0.7 percent of its $3 billion fiscal year 1991 R&D budget for technology transfer. Sandia National Laboratories direct spending on technology transfer does not exceed 0.5 percent, or $5 million, of its $1 billion budget, and many other labs have earmarked less or nothing at a11. 19 Congress might consider designating some of the labs R&D appropriation for promoting commercialization, perhaps mandating that a few percent of the budget be set aside for the purpose. Activities would include identification and marketing of promising technologies, patenting when appropriate, and participating in cooperative R&D projects. Bills being drafted for the 102d Congress would increase the labs role in promoting technology commercialization. 20 In addition, Congress could take action to remove some remaining barriers to commercialization. For example, the law might be changed so that software created by government employees can be copyrighted, which would sometimes make its commercialization more feasible. 21 At some point, however, major responsibility for energetic technology transfer must fall to the labs, their parent agencies (especially DOE and the Department of Defense), and private companies. Japanese technology is another lode that U.S. companies could mine if they begin to pursue new technologies from outside sources more aggressively. There are difficulties-the most obvious being the language barrier-in getting access to Japanese technologies. Also, most Japanese technology is developed by private industry and thus is less accessible to outsiders than technical knowledge
PAGE 56
50 Competing Economies: America, Europe, and the Pacific Rim that is freely available at universities and public institutions. A few universities have fellowship programs that send American scientists and engineers to Japanese companies and research institutions on long-term projects, and the National Science Foundation (NSF) has recently established similar programs, largely funded by grants from the Japanese Government. The NSF-Japan programs are not yet fully subscribed, but are likely to attract more applicants as they become better known (typical of university programs). Congress may wish to monitor the progress of the NSF programs, possibly adding funding if they become oversubscribed. Another option would be to establish a congressional U. S.Japanese Fellowship Program, taking advantage of the visibility and prestige conferred by the sponsorship of Congress (perhaps especially in Japanese eyes). In addition, Congress might wish to encourage sabbaticals in Japan for researchers working in Federal labs. A longer term, more fundamental option is to promote Japanese language instruction both in the public schools and in universities, especially for scientists and engineers. Improved protection of intellectual property and modest changes in antitrust law might bolster the competitive position of some U.S. manufacturing industries. Better intellectual property protection could start at home with speedier enforcement. Patent cases that go to trial take an average of 21/2 years before a decision. Congress might help by designating special patent judges with the technical experience to move cases through expeditiously (similar in principle to the tax courts). As for foreign markets, the Japanese patent system is a particular problem. It is slower than the U.S. system in issuing and enforcing patents, and it strongly favors licensing of patents, which U.S. companies do not always wish to do. The U.S. Government is negotiating with the Japanese on these and other problems. In addition, Congress may wish to establish a program in the Patent Office or elsewhere in the Commerce Department to provide information to U.S. companies about the Japanese patent system, which most firms do not understand. Antitrust law and enforcement have been relaxed in the past decade, but some cooperative endeavors that could help U.S. firms may be dampened by fear of antitrust action. Changes in antitrust law and enforcement should be approached with caution; the laws have served the country well for many years in maintaining g competition. However, the laws complexity and vagueness, together with stiff penalties, may discourage some useful cooperation. Congress amended the law in 1984 22 to make joint R&D easier, chiefly by reducing the maximum penalty from treble damages to single damages for publicly registered projects. Other options might include extending the 1984 Act to cover joint manufacturing (as proposed by the Administration in the 102d Congress), and case-by-case review and advance certification by the Justice Department that particular joint projects do not violate the law. Another possibility is to establish safe harbor market shares, so that cooperating firms with combined market shares below a certain percentage would not be in violation. Perhaps most important, Congress could instruct the Justice Department and the courts to evaluate possible mergers and joint ventures in light of a key fact of modern international competition: that foreign firms with low U.S. market share may rapidly increase that share and even become dominant, if they possess strategic technology, large world market share, sufficient financing, or other advantages. If this fact is accepted, then under standard antitrust analysis mergers or joint ventures between U.S. firms should sometimes be permitted to avoid eventual dominance by a foreign firm. Similarly, some buyouts of U.S. firms by a foreign firm would sometimes be stopped, in order to avoid eventual dominance by the foreign firm. These arguments have often been met with skepticism; Congress could urge that they be taken seriously. Information and exhortation to American manufacturers on how to make things better, given under U.S. Government auspices, might not seem a very promising strategy. But a Department of Commerce program (the Malcolm Baldrige National Quality Award) does that very thing, and has proven astonishingly effective in just 3 years. It costs the taxpayers almost nothing. Even the administrative expenses are covered by private contributions and applications fees from companies vying for the award. The award was created by Congress in 1987. It is given to companies or organizations that have substantially benefited the economic and social well-being of the United States through improvements in the quality of their goods or services from
PAGE 57
Chapter 2--Policy Issues and Options 51 Box 2-AThe Baldrige National Quality Award The manager of the Baldrige Award, the National Institute of Standards and Technology (NIST) of the U.S. Department of Commerce, sets out seven major criteria that contestants must address in their applications for the award. NISTs Application Guidelines list the maximum points applicants can win in each of the categories, adding to a total potential score of 1,000 points. l Leadership: The senior executives success in creating and sustaining a quality culture. 100 points. Information and Analysis: The effectiveness of the companys collection and analysis of information for quality improvement and planning. 70 points. Planning: The effectiveness of integration of quality requirements into the companys business plans. 60 points. Human Resource Utilization: The success of the companys efforts to utilize the full potential of the work force for quality. 150 points. Quality Assurance: The effectiveness of the companys systems for assuring quality control of all operations. 140 points. Quality Assurance Results: The companys results in quality achievement and quality improvement, demonstrated through quantitative measures. 180 points. Customer Satisfaction: The effectiveness of the companys systems to determine customer requirements and demonstrated success in meeting them. 300 points. Every year, six awards may be given, two each in three categories, Manufacturing, Service, and Small Business. Awards need not be given if no one qualifies. The winners are shown in the table below. Winners of the Baldrige Award Manufacturing Service Small business 1988 Motorola, Inc. Globe Metallurgical Commercial Nuclear Fuel Division, Westinghouse Electric Corp. 1989 Milliken & Co. Xerox Corp., Business Products & Systems 1990 Cadillac Motor Car Division Federal Express Wallace Co., Inc. IBM Rochester Iu.s. Dep~@ of Cor.m.neme, National Institute of Standards and lkchnology, Malcolm Baldrige National Quality Award, P~P~et (Ga.ithersburg, MD: NIS~ 1991). effective practice of quality management. 23 Six to comment, and how it follows up with customers awards may be presented each yeat--two each for manufacturing, services, and small business-but none need be given if no one qualifies. In the first 3 years of the program (1988-90), there were 203 applications and 9 awards given, 6 to manufacturing companies, 2 to small business, and 1 to a service company (see box 2-A). The award has been an excellent means of technology diffusion. Just filling out an application, following NISTs 42-page booklet of Application Guidelines, can be an eye-opener. For example, the company must satisfactorily relate how it gives customers easy access to assistance or opportunities to see if they are satisfied with products and services. These are just 2 of 33 detailed areas that companies must cover in their applications (50 pages for small companies, 75 pages for large ones). According to Jerry Junkins, CEO of Texas Instruments, If you measure your-self against the criteria laid out by the Baldrige award, you have a blueprint for a better company. 24 All applicants, win or lose, receive reports from ex aminers outlining their good and bad points, and this is enough to improve some companies quality efforts. Some companies do not enter but use the Baldrige criteria as a company standard. One winner, Motorola, demanded that 3,600 of its
PAGE 58
52 Competing Economies: America, Europe, and the Pacific Rim larger suppliers prepare to compete or face being dropped. 25 Winners must share with others details of what they did to win the award; company representatives give hundreds of speeches a year and hold large briefing sessions for executives of other companies, including their competitors. One manager who attended a winners presentation came away amazed at the level of detail, which he described as everything but the financial data. After hearing the presentation, he reckoned that his own company, considered a leader in its high-technology field, would not be able to qualify for the award for at least 6 years. Despite a few criticisms of the Baldrige award (e.g., some consultants consider the criteria are too directive), its good effects appear to far outweigh any adverse ones. In the general area of quality standards, however, there may be room for further government action. Suppliers trying to meet formal quality requirements for large customer fins, particularly the Big Three auto assemblers, are frustrated by slightly varying requirements and separate certification. If Fords Q-1 and GMs Mark of Excellence certifications were made consistent and interchangeable, suppliers believe they would benefit. Through oversight, Congress might encourage NIST to look into the question and work out a constructive solution. TRADE POLICY TOOLS Since World War II, U.S. policy has promoted free trade by keeping its own market generally open and urging other countries to do the same. For two or three decades this policy served U.S. interests well enough. U.S. manufacturing technology was so far ahead of most other countries, and the U.S. market so much larger than others, that manufacturers flourished even in cases where the free trade was mostly one wayfree for U.S. imports but quite encumbered for exports. At the same time, opening American markets to foreign goods served the U.S. policy of promoting economic development in poorer countries and cementing loyalty among Cold War allies. Today, free trade is not invariably to the United States advantage. Advanced countries have rebuilt their war-damaged economies, several poorer countries have begun to industrialize, and the United States is no longer predominant. In some industries and technologies it has fallen to second place. This is a particular concern in key high-reward industries such as electronics, including parts of the semiconductor and computer industries. These are industries that can help make a country rich, because markets are rapidly expanding, producers can capture substantial economies of scale and learning, and the industry has technology spillover benefits for the rest of the economy. Most of the United States decline was selfinflicted. The U.S. Government has never formed the partnerships with industry that helped its best competitors advance. Many of its industries have stuck with outmoded management and technologies while others passed them by. Part of the trouble, however, was that foreign market barriers and export drives deprived key U.S. industries of needed revenues and experience, sapping their strength and even undermining their existence. The neardestruction of the U.S. consumer electronics industry was caused partly by closed foreign markets combined with dumping (selling at prices below fair value) of foreign goods in the U.S. market. U.S. trade law and policy are supposed to prevent events like these. If foreign markets are closed to U.S. goods, the U.S. Government tries to get the barriers removed. If imports are subsidized by foreign governments, or if foreign subsidies or market barriers result in dumping, the U.S. Government can in some circumstances levy a duty on the imports intended to compensate for the foreign advantage. However, foreign market barriers may take years to remove. U.S. law regarding subsidies and dumping has been at best slow and incomplete in compensating for the advantages foreign firms receive from their governments. Moreover, tariffs are inherently inadequate to compensate for another countrys domestic policies; when foreign governments help to give their industries a competitive edge, it takes more than trade policy for U.S. industries to catch up. U.S. trade policy could more effectively promote U.S. competitiveness if it were part of an overall strategy. A redirected trade policy might have three operating principles. First, when a critical industry is in trouble, the primary government response would be a domestic one. Measures might include R&D support, tax breaks and incentives for R&D & capital investment, support for technology diffusion, and support for education and training. Trade
PAGE 59
Chapter 2--Policy Issues and Options l 53 measures-trying to open foreign markets, and protecting the U.S. market--could be used when necessary, but in a subordinate role. Second, efforts to open markets would focus on areas of the greatest strategic importance. Third, when opting to use protection, the basis would not be legalistic criteria, but rather the industrys need and place in the economy. The important industries are not difficult to identify. Several government and private reports in the United States, Japan, and Europe have listed high-priority technologies that drive competitive industrial performance, and their lists are similar. They include electronic components, information technologies, materials and associated processing, manufacturing process technologies, and propulsion and powertrain technologies. 26 Ideally, the important industries would be identified by an institution responsible for a comprehensive government approach to trade and competitiveness policy. Otherwise, individual government agencies involved could draw up their own lists. This leaves the question of how to fashion assistance to threatened U.S. industries. There is the danger that government assistance will be squandered on industries unable to compete, or that assistance will remove industries incentive to improve their competitiveness. However, other countries experiences provide significant guideposts showing, for example, that industry should take the lead in proposing joint government-industry R&D programs and shoulder much of the cost; that policies should conform with market forces as much as possible; that U.S. industry must compete with the best in the world; and that industry must make its own efforts to improve its competitiveness and outgrow the need for assistance. Ideally, a government institution with overall responsibility for trade and competitiveness would coordinate strategic assistance. But in the absence of such an institution, individual agencies could act on their own according to these principles. For example, the Department of Energy, in its pilot program for cost-shared R&Din high-temperature superconductivity, participates only in projects that industry proposes and for which industry pays about half the cost. 27 Finally, these options come with an important caveat. This report is addressed to Congress, and presents options that Congress may wish to adopt. Yet there is little that Congress can do on its own. It can state goals, allocate resources, and give guidelines, but strategic policies to improve U.S. competitiveness policies require initiative and judgment by the executive branch. The sympathies and energy of the executive branch are needed to get the job done. Responses to Foreign Market Barriers Foreign market barriers can hurt U.S. industries. 28 While GATT has reduced quotas (quantitative restrictions on imports) and tariffs (taxes on imports, also called duties), there are many other barriers to imports. These include burdensome customs procedures; preferential government procurement; discriminatory standards and regulations; and companies agreements or practices, sometimes tolerated or encouraged by the government, not to buy foreign products. Normally, the U.S. response, if any, is to negotiate to eliminate barriers. The Office of the United States Trade Representative (USTR) leads the negotiations, with assistance from other agencies. The USTRs lean staff can negotiate only a limited number of cases, and may be so overburdened as to affect the quality of negotiations. And the fast turnover of many senior (political level) negotiators and policymakers has sometimes hindered the United States from pressing its position consistently over the years. High ranking government officials have frequently left government to represent foreign manufacturers or their U.S. importers, giving those interests both access to top government decisionmakers and the savvy to exploit any weaknesses on the U.S. side. Congress could address these problems in various ways. It could expand the USTR staff, enabling it to take on a wider range of foreign market barriers and match other nations, Japan in particular, in preparation and negotiating strength. Congress could also reduce the number of political appointments and add more high-level career civil service slots dealing with trade policy to the USTR, the Commerce Department, and other agencies; making long-term service more attractive would improve institutional memory and facilitate the steady pursuit of goals over the years. Finally, Congress could prohibit senior trade officials from representing foreign interests for several years after they leave government service. Other reasons why foreign markets can take many years to open are harder to address. Barriers take
PAGE 60
54 Competing Economies: America, Europe, and the Pacific Rim time to identify; other countries can stall negotiations; countries hedge on promises to remove barriers; and when one barrier is removed, another can take its place. Japanese trade barriers especially are often like an onion, with layer after layer to peel away. And some hindrances are ingrained in a countrys social norms and domestic policy, making their removal difficult even if the foreign government is willing. Success often depends on the leverage the United States has over other countries. The two major sources of leverage are limited. First, when the country and subject matter are covered by GATT, the United States can invoke GATT dispute resolution procedures. This process is slow and uncertain, since the other country can veto decisions adverse to its interests. Recent developments in the GATT Uruguay Round have improved the process; further improvements may yield a more reliable means for eliminating market barriers. Another approach would be to create a new, multilateral trading system, consisting of only those countries that are truly willing to abide by the dictates of free and open trade, and to negotiate bilateral deals with nations outside the system. 29 This kind of trade management far exceeds anything we have done in the postwar period, and creating it would require much time and patience. The new trade regime probably would need a strong enforcement mechanism, perhaps along the lines of the International Trade Organization (ITO), which was originally proposed in the late 1940s but was blocked by the United States and therefore never formed. 30 Congress might wish to forma task force to investigate what such a revised system could entail. The second source of leverage is to threaten retaliation under Section 301 and related sections of the Trade Act of 1974, as amended, 31 which allow the United States to impose punitive barriers to imports from another country in retaliation for that countrys unfair trade practices. However, the U.S. Government is often reluctant to retaliate under Section 301. The retaliation could violate GATT so publicly as to make the United States vulnerable to criticism (including formal complaint and investigation under GATT rules) or further retaliation. Punitive barriers on imports would not solve the U.S. industrys problem and could create problems for downstream industries in the United States. And even the start of a Section 301 investigation angers foreign countries, for whom Section 301 is a symbol that the United States considers its national laws superior to GATTs international law. 32 Because of these problems, negotiations to open foreign markets are in many cases slow and ineffective. Barriers that cause particular damage--such as Japanese barriers to the sale of semiconductors and supercomputers (chs. 4 and 6)--often persist the longest. While market opening is a worthwhile long-term goal, it often provides little or no relief in the short term, during which time the affected U.S. industry can suffer serious damage. Accordingly, it might make sense to change the U.S. response to foreign market barriers. The response could be primarily domestic programs to aid the affected industry. Often these programs would be enough. The U.S. market, still the worlds largest for many products, can often support the industry. In such cases, the United States could still negotiate to open foreign markets, but in a low-key manner, to encourage change without angering the countries involved. It would also make sense to allocate the governments limited resources according to an industrys strategic importance. However, domestic programs might not always be enough. This might be the case with semiconductors. Even though the U.S. Government is contributing $500 million in R&D assistance to the industry over 5 years through Sematech, that is insufficient to arrest the U.S. industrys competitive decline vis--vis Japan. The semiconductor industry is particularly dependent on economies of scale and learning, and without access to the Japanese market the largest and most discriminating in the world the U.S. semiconductor industry will likely keep slipping. In such a situation-a key industry in danger if a foreign market remains closedthe United States could consider an aggressive program to open the foreign market, if necessary by a prompt threat of retaliation under Section 301. These cases could merit cabinet level or even Presidential involvement. Semiconductors apparently is a case in which aggressive action was deferred for too long. Despite negotiations begun in 1972, Japans semiconductor market remained largely closed to U.S. products. 33 The United States commenced a Section 301 investigation only in 1985, leading to an agreement in 1986 that was broken by Japan, followed by U.S. retalia-
PAGE 61
Chapter 2-Policy Issues and Options l 55 tion in 1987, and finally, starting in 1989, a modest increase in U.S. firms share of the Japanese market. While it might occasionally make sense for the United States to pull out all the stops in trying to open a foreign market, such cases must be chosen carefully. As discussed above and in chapter 4, the threat of retaliation-and execution of this threat is likely to anger foreign countries and could have other serious adverse consequences. It is thus important to rank foreign market barriers, identifying those industries that are most important and in which barriers have a pronounced effect that domestic programs cannot alleviate. In general, the USTR has not made strategic priorities. For example, in the early 1980s, the U.S. Government successfully pressured Japan to buy a certain amount of U.S.caught fish, but would not do the same with semiconductors. 34 This lack of strategic planning was evident and surprising to Japan. Clyde Prestowitz, a key Japan negotiator in the Commerce Department for most of the 1980s, recalls: Once an official of the [Japan] Economic Planning Agency asked how we created our agendas. His agency, he said, had carefully studied the competitiveness of various U.S. industries, and in his opinion there were much better issues for the U.S. to pursue than the ones on its list. 35 Use of Protection U.S. industries normally receive trade protection only when imports are dumped or subsidized. 36 In this case, an extra duty can be assessed in an amount that in theory will precisely counteract the subsidy or dumping. In principle, this extra duty does not distort trade from what would occur with free trade and a free market, but rather corrects a distortion already present. The goal is to put U.S. industry back on an equal footing. Such extra duties are in principle consistent with GATT. However, the U.S. law and practice regarding subsidies and dumping by and large fail to compensate for the advantages foreign governments create for their firms. Reasons for this include delay, difficulty in proving subsidies or dumping, the laws ignoring or devaluing certain subsidies, difficulty in proving the required injury, and the high expense of legal proceedings. A further problem is that the effects of government assistance can increase over time rather than dissipate as the law assumes. To some extent, the laws limited effectiveness stems from adherence to GATT requirements. These U.S. laws, however imperfect, are now often the only line of defense for key industries facing stiff foreign competition; it could be harmful to weaken them further as many GATT members are seeking to do in the Uruguay Round. However, strengthening these laws might lead to a flood of cases, anger trading partners, provoke legal challenges under GATT, and result in imposing extra duties that might not be in the countrys best interest. The debate between those who would strengthen and those who would weaken the subsidy and dumping laws might be resolved by shifting the policy focus of import protection. First, for beleaguered U.S. industries, the primary response would be domestic programs to help the industry compete or rationalize. (To rationalize means to shrink, but in a planned fashion that also seeks to improve technology in potentially competitive subsectors.) This could include R&D support, tax breaks, and other measures designed either to help an industry regain technical parity with foreign competitors or to ameliorate the shock of downsizing. Where domestic measures alone might not suffice, protection could be used, lasting only as long as strictly necessary. Criteria for awarding protection would include the industrys need, its merit (including whether the industry was making reasonable efforts on its own and showed promise of effectively competing on its own), and the importance of the industry in the U.S. economy. While the presence of subsidies or dumping might be relevant to the decision to grant relief from imports-for example, to show that an industry is being beaten only because of the intervention of foreign governments subsidies or dumping would be among many factors to consider. Likely adverse effects on downstream industries would also be considered, though the government could take measures to ameliorate them (e.g., special tax breaks). The protection component of such a reoriented policy in principle could be consistent with GATT. Such an approach could build on Section 201 and the following sections of the Trade Act of 1974, as amended, under which the government can impose temporary relief (up to 8 years) when increased imports cause or threaten serious injury to an existing U.S. industry .37 Section 201 follows GATTs so-called escape clause, which permits import
PAGE 62
56 l Competing Economies: America, Europe, and the Pacific Rim barriers in this situation but requires the payment of compensation (normally reduced tariffs on other items) to affected countries (see ch. 4). Under Section 201, the U.S. International Trade Commission (ITC) determin es whether the injury requirement is satisfied and, if so, recommends relief. The final decision on relief is up to the President, who has great flexibility in choosing the nature of relief. Forms of relief include quotas, higher tariff rates, and negotiations with foreign governments to address the underlying cause of the increase in imports or to otherwise alleviate the injury. 38 The President must consider the efforts that the domestic industry is making on its own to improve its performance; every 2 years thereafter, while the protection lasts, ITC must report to the President on industrys continuing efforts to improve, and the President may modify or te rminate relief if he finds those efforts insufficient. However, as currently administered, Section 201 would have limited usefulness as a vehicle for strategic use of protection. The serious injury requirement is hard to satisfy, and partly for that reason Section 201 has been little used in recent years. 39 The law contains the requirement, not specified by GATT, that the increased imports be at least as great a cause of injury as any other cause. 40 The injury requirement is not often satisfied before serious damage is done. While the threat of serious injury is in principle enough to satisfy the statute, the threatened injury must be imminent, and a sufficient threat is rarely found. Also, following the language of GATTs escape clause, Section 201 by its terms can be used to protect industries already producing goods, not those still fo rming. 41 Congress could make Section 201 more serviceable by eliminating the requirement that the increased imports be at least as great a cause of injury as any other cause. Congress could also specify that the injury test can be satisfied even when an industry is doing well, if the imports have impeded the industrys development or import and industry trends point to eventual foreign dominance. Such a provision might be consistent with GATT, although other countries could argue otherwise. 42 GATTs escape clause requires that the increased imports and consequent injury be due to unforeseen developments. 43 Section 201 does not. However, if the United States attempted to use Section 201 to protect a new (infant) industry struggling to compete against a well-established foreign industry, foreign countries could complain that imports and the consequent injury to U.S. companies were hardly unforeseen. Uruguay Round negotiations might eliminate this requirement, which has rarely been used or invoked. All told, Section 201 is far from an ideal tool for providing GATT-compatible protection. However, Congress could empower the President to grant protection apart from Section 201. Ideally, a government agency responsible for coordinating competitiveness strategies would recommended protection when needed. Instead of justifying protection under GATTs escape clause, the United States could justify it under GATT Article XXVIII. Under that Article, the United States could negotiate compensation, typically in the form of reduced tariffs on certain other products, in exchange for which other countries would accept increased U.S. tariffs on the products at issue. The President could be empowered to offer such compensation, as recommended by an agency with overall competitiveness responsibility. While compensation could adversely affect other U.S. industries, the government might, in some cases, be able to mitigate these effects with tax breaks or other programs that the same agency could recommend and the President could be empowered to grant, subject to a congressional override. 44 If negotiations failed and other GATT-consistent means of protection could not be found, as a last resort the United States might impose protection anyway. Other countries have at times protected industries they consider crucial, regardless of GATT. Such a course might provoke a GATT dispute, and could possibly lead to a ruling requiring the United States to pay compensation. 45 This approach would not necessarily signal U.S. abandonment of its loyalty to GATT and the free trade ideals it represents. Rather, it would mean that the United States, like other countries, can depart from free trade ideals when necessary. Under both Section 201 and this new approach, it would be desirable to condition protection on a performance requirement or showing of progress by the industry receiving protection (which has happened to some extent under Section 201). Openended trade protection with no strings attached has been a recipe for third-rate performance for industries in many countries (e.g., the national champion computer firms in Europe-see ch. 5). On the other
PAGE 63
Chapter 2-Policy Issues and Options l 57 hand, the discipline Korea imposed on its developing industries+. g., requiring companies to succeed in exporting finished goods as a condition for permission to import production equipment, parts and materials-was an ingredient in the success of its export-led development, The specific policy tools Korea used are no model for the highly developed and much freer U.S. economy, but some kind of standard or gauge of serious effort on the part of the industry and progress toward competitive performance could be employed. Domestic Content Requirements The issue of domestic content is sure to come up if the United States protects industries against fair but damaging imports. One way of getting around protective tariffs and quotas is to build an assembly plant in the protected country; Japanese automobile production began in North America in large part to avoid current and prospective protection. The fact that foreign direct investment has grown faster than trade in the past couple of decades reflects, in part, companies desires to avoid or avert national protection. While foreign investment is often welcomed and occasionally sought, some nations have not been content to let foreign companies substitute domestically assembled merchandise for imported goods, because often the domestic assembly adds only a small proportion of value to what is basically an imported good. The United States has done little other than jawbone to increase the domestic content of foreign companies producing here, but many European nations and the European Community have moved more decisively. For example, the British Government eagerly pursued Japanese automobile investment but included the proviso that 60 percent of the content of the autos must be European at the time of startup and 80 percent within a few years. Such high levels of local content require that the body, the major mechanical components, and either the engine or the transmission be fully manufactured in Europe (see ch. 5); the current level is much greater than the local content of cars made by any Japanese transplant in North America, even after years of operation. Domestic content requirements are also prominent in Europe for electronics products, including office equipment, consumer electronics items, and semiconductors. Government procurement in most EC member nations strongly favors domestically made goods. Moreover, for 29 Z-Q89 O 91 +3 QL;3 nearly a decade, the EC has pursued vigorous antidumping measures against Japanese and other Asian firms selling electronics goods in Europe; the penalties can be lifted if the firms include substantial amounts of European-made parts and materials in their products, and many firms have relocated production to Europe in response. Domestic content regulations, like tariffs and quotas, can benefit a nation. Almost certainly, they mean more jobs in the protected industry than otherwise, at least in the short run. But, like other forms of protection, they nearly always mean some short-term sacrifice by consumers in the form of higher prices. Whether national productivity or competitiveness improve with domestic content requirements is uncertain and may depend on whether they coincide with or are tied to measures to improve the competitiveness of the domestic firms or industries. Experience with domestic content requirements in industrialized countries is recent and limited. There is some evidence that Japanese auto assembly transplants in North America have helped to improve the quality and productivity of U.S. suppliers; the more recent Japanese transplants in Europe are expected to do the same for European suppliers. Also, in North America the demonstration effect of the Japanese assemblers spurred the Big Three domestic automakers to improve their own and their suppliers quality and productivity. It is not yet clear that requiring high domestic content of Japanese producers in electronics products will help the European electronics industry. It maybe improving the performance of some European suppliers, but if so that improvement has not yet spilled over into greater competitiveness of the European systems manufacturers who make computers, other office machines, and consumer electronics items; all these European companies are in trouble. It is also unclear to what degree good performance by Japaneseowned firms will contribute to an elevated level of technology and higher standards of living in the host country. It may be that the leading edge of innovation will remain in the home country of the foreign investor, but that the host country could still benefit from demonstration and direct teaching of superior manufacturing practice. European policies reflect uncertainties about benefits investment. The EC seems the ambiguities and from foreign direct to have adopted a
PAGE 64
58 Competing Economies: America, Europe, and the Pacific Rim principle of tolerating or encouraging foreign investment in some critical sectors if the foreign firm agrees to a high level of local content. Yet there are contradictions. The recent EC agreement with the Japanese Government on automobiles would allow only a gradual increase in the Japanese share of the EC market, both transplant production and imports, despite the high domestic content levels that the transplants have already agreed to and are trying to achieve (see ch. 6). 46 Overall, it seems that the EC Commission and the member governments are still trying to decide whether foreign direct investment is a net benefit. Japan, on the other hand, has an unambiguous stand. Japan has often barred or severely constrained foreign direct investment, even when it was far behind other advanced industrial nations and trying to catch up. Even today, Japan remains one of the most difficult of industrialized nations in which to open a branch, partly because of the expense but also because of the red tape involved. Export Promotion Export promotionhelping firms take advanages of opportunities to sell abroadis another policy tool that could help U.S. manufacturing competitiveness. To export, companies must pass many hurdles: analyzing foreign markets; identifying and contacting potential customers; learning foreign ways of doing business; creating new labels or otherwise adapting the product for foreign use; getting financing; and arranging for shipping, insurance, and customs clearance, to name a few. Even when a U.S. firm has a good, well-priced product, it can easily miss an export sale. Government programs can provide information and contacts. They can inform U.S. firms about markets, potential customers, foreign regulations and procedures, shipping, and so on; they can inform foreign firms about U.S. firms and products; and they can arrange trade shows, interviews, and other contacts between U.S. and foreign fins. For manufactured goods, this assistance at the Federal level is provided primarily by the U.S. Foreign and Commercial Service (USFCS) of the Department of Commerce. (The Department of Agriculture provides export services for agricultural products.) Government programs can also assist with export financing. Often credit terms play an important part in export sales, especially to developing countries. Government help in providing easier credit terms can take the forms of: 1. 2. 3. The insuring the exporter against the customers default, if the exporter allows the customer deferred payment; guaranteeing a bank against the customers default, if the bank lends the customer money for the purchase; and, lending money itself, either directly to the customer or indirectly through a bank. Export-Import Bank of the United States (Eximbank), an independent agency of the U.S. Government, provides this kind of assistance. In both information and financing, U.S. export promotion programs are weaker than those of our major competitors, primarily because much less is spent on them (ch. 4). The difference is due in part to limits on spending by the U.S. Government, after years of enormous budget deficits. It also stems from a fundamental uncertainty among U.S. policymakers as to whether export promotion is something the government should be doing. If Congress and the Administration decide that export promotion is a legitimate government function, there are straightforward ways to make it more effective. Congress could increase funding for USFCS to pay for additional commercial officers posted abroad. These are the people who gather information about foreign market opportunities and help U.S. firms find foreign contacts. The s ummer 1991 level of about 200 foreign commercial officers, 47 while up somewhat from about 150 in 1980 and 1988, 48 is still low. For example, as of August 1990, there were only 15 commercial officers in Japan, plus 44 Japanese nationals assisting with export promotion, while 83 professional employees of the Japanese Government, all Japanese citizens, were working in the United States to promote exports. 49 Congress might wish to ask the Commerce Department whether other funding increases could provide improvements in service. In the late 1980s, budgets were very tight; sometimes USFCS officers even lacked funds to return phone calls to the United States. While funding has improved somewhat, given this history and the low level of spending compared to important trading partners, it is likely that additional funding could have a healthy payoff. The level of service depends on attitude as well as money. Congress might wish to make a policy
PAGE 65
Chapter 2-Policy Issues and Options 59 statement that export promotion should be a priority not only for commercial officers abroad but for the whole diplomatic staff. Cabinet-level involvement in promotion activities, such as Commerce Secretary Mossbachers presence in Tokyo in April 1991 to kickoff the Japan Corporate Program (ch. 4), could provide a boost. Export promotion efforts could have a more strategic focus. While USFCS emphasizes industry sectors whose fundamental competitiveness suggests substantial export potential, it does not consider which industries are strategic to overall U.S. competitiveness. While Eximbank reports to Congress on the amount of financing meeting certain strategic priorities (such as industries with high value added, or industries that particularly benefit downstream industries), it is not clear how much Eximbank takes them into account in its decisions. Ideally, an agency with overall competitiveness responsibility would coordinate strategic priorities. Strategic priorities for Eximbank could also help to solve the problem of excessive paperwork. Financing assistance by Eximbank must be justified on a case-by-case basis. Congress might consider adopting the approach used by Japan and many European countries, which determine in broad policy terms what exports to assist, and then assist all creditworthy exports within the guidelines. U.S. manufacturers are at a disadvantage, compared with foreign competitors, because of tied aid. This is a scheme by which a country gives foreign aid on condition that the recipient use the money to buy products from the donor. The United States ties some of its aid, but U.S. nondefense aid focuses on agriculture, health, nutrition, and education. American farmers may reap large benefits, but manufacturers seldom do. Other countries concentrate aid much more on heavy construction projects, such as power generation and transportation, so that their tied aid involves manufactured goods, especially capital equipment. Tied aid is often combined with export financing in a package of so-called mixed credits. For example, the exporting country might offer to pay outright 40 percent of the cost of a power plant, and finance the other 60 percent, provided that national companies of the exporting country get the contracts for construction and equipment sales. The U.S. effort to strengthen international agreements aimed at limiting tied aid are worth continuing. However, it is uncertain how successful that effort will be. Congress could expand the so-called War Chest for matching foreign tied aid offers to make it more effective in discouraging foreign tied aid. The War Chest was funded at $150 million in grants for FY 1991, which would result in about $500 million in loans, though as of July 1991 only $58 million in grants had been used, resulting in $131 million in loans. In contrast, Japan, France, and Germany use tied aid to make loans of billions of dollars per year. Congress could also expand the Trade and Development Program (TDP), funded at $31 million for FY 1990 and $35 million for FY 1991. TDP helps to pay for feasibility studies or other plannin g assistance performed by. U.S. firms for capital projects. Participation in the planning phase has often helped U.S. firms win contracts for the actual project. So far, $161 million in program funds have led to documented U.S. sales totaling $3.2 billion, with an estimated $18 billion more sales expected as projects mature. 50 In expanding TDP or otherwise increasing the emphasis on capital projects, 51 care should be taken to avoid adverse environmental and social effects, which in the 1970s turned the United States away from such projects. Export Controls The export of dual-use items, those having both military and civilian use, is regulated by the Export Administration Act of 1979, as amended (EAA)~ 2 which requires U.S. firms to get a license to export certain items to certain destinations. The intent is to deny strategically important goods and technology to potential military adversaries. In 1990, perhaps $90 billion worth of U.S. exports of manufactured goods required a license, or about 28 percent of the $316 billion in manufactured exports (see ch. 4). It is generally agreed that some export controls on dual-use items are needed. However, there is an emerging consensus that export controls have unduly hindered U.S. high-technology firms in competition with foreign manufacturers. For example, U.S. controls limited exports of personal computers based on Intels 80386 processor chip until mid1990. Yet the same computers were widely available from foreign fins. The controls merely diverted business to foreign fins.
PAGE 66
60 l Competing Economies: America, Europe, and the Pacific Rim The U.S. export control regime has been considerably liberalized in the last few years. The political changes in Eastern Europe and the Soviet Union reduced those countries military threat, lessening the need for controls, and in fact making desirable the export of technology to help those nations become open, economically viable societies. This changing political climate intensified the dissatisfaction of many allies with the United States stricter position on export controls; allies brought pressure for change in the international export control regime. Finally, concern has increased over the continuing decline of U.S. manufacturing competitiveness, prompting closer scrutiny of whether the security benefits of particular controls are worth the competitiveness costs. Reform has focused on East-West controls, traditionally the most common type, which are meant to deny militarily strategic technology to former Communist countries (see ch. 4 and box 4-C). Controls on many items have been removed or reduced, tracking a major reduction of controls at the international level, in CoCom (Coordinating Committee for Multilateral Export Controls, a group of nations cooperating in export controls). In principle, controls should be eliminated if they are not also imposed by other CoCom members, or if the item is available from unrestricted sources. In addition, delays in processing export license applications have been shortened. However, the reform is incomplete. The reason is the tug of war that exists-and to some degree must exist-between two important governmental goals: military security, whose champion is the Department of Defense; and improved competitiveness, championed by the Department of Commerce. Although the competitiveness interest has received increasing support from the President and his closest associates, it still has not achieved equal consideration with military concerns. Congress cannot change this on its own: where possible damage to military security is at stake, the executive branch must exercise wide discretion. If Congress were to create a new agency charged with promoting industrial competitiveness, many aspects of the export control function might reside there. It could coordinate export controls as the Commerce Department does now, but with a greater ability to serve the national interest. Other functions might include: evaluating the economic importance of different industries and the importance of exports to a given industry; expediting control reforms and license approvals in key industries; coordinating export policy with other policies (e.g., offering some compensating benefit to the affected industry when cumbersome export controls were deemed appropriate); and achieving enough prestige that its views on export control policy would carry equal weight to DoDs, within the Administration. Without such an agency, and lacking a commitment in the Administration to advance commercial competitiveness, there are still measures that Congress could take if it wished to give competitiveness a higher priority in export control policy. However, the EAA is already a very complicated statute, and even some analysts sympathetic to competitiveness concerns believe that it contains excessive micromanagement. On the other hand, the statute leaves the the Administration an out to do what it believes is truly necessary in most cases. Where the statute leaves no flexibility, sometimes the Administration has disobeyed it. This occurred with the 1988 provision eliminating reexport controls on U.S. goods and technology that are incorporated abroad into finished products, provided the controlled U.S. content is at most 25 percent of the products total value. 53 The Bureau of Export Administration (BXA) considered that provision dangerous; for example, it would remove controls from avionics equipment incorporated abroad into airplanes. BXA implemented this clause only incompletely and almost a year late (ch. 4). Congress could strengthen the role of the Commerce Department vis--vis the Defense Department and other agencies. Congress moved in this direction in the Export Enhancement Act of 1988 (1988 Act), 54 for example, by limiting to 40 days the time during which the Defense Department can block a license approval recommended by the Secretary of Commerce. 55 Further amendments in this direction are under consideration in S. 320, already passed by the Senate. 56 For example, S. 320 would give the Commerce Department permanent representation at CoCom and direct the State Department to forward to CoCom within 7 days certain Commerce Department actions requiring CoCom approval. 57
PAGE 67
Chapter 2-Policy Issues and Options 61 The Commerce Department has sometimes been bypassed when dual-use items have been put on the State Departments Munitions List, which is in principle limited to items with only military use. Items on the Munitions List face a stricter control regime, without the safeguards to protect commercial competitiveness that apply to dual-use items. Therefore, placing dual-use items on the Munitions List can reverse recent reforms and impede pending ones. Congress could discourage this practice by, for example, giving the Secretary of Commerce the right to force a quick Presidential decision regarding whether an item also has nonmilitary use. Congress could also issue a strong policy statement that the Munitions List is not to contain dual-use items. Congress might enact additional provisions to address problems identified in chapter 4. For example, Congress could impose stricter time limits for processing license applications, especially those requiring interagency review; mandate prompt continuing review of the Control List (list of controlled items); encourage license-free trade within CoCom countries; encourage use of industry advisory committees; and encourage the use of indexing, by which technical thresholds of what is controlled would be automatically adjusted over time unless the need to forego the adjustment were specifically justified. The 1988 Act and S. 320 address these concerns. Other provisions in S. 320 also address competitiveness concerns. This bill contains policy statements favoring approval of exports designated for reformed Eastern European countries, or needed to aid Soviet and Eastern European economic development. Another policy statement favors temporary exports for trade shows. However, these are by necessity only guidelines, leaving final discretion with the Administration. S. 320 provides for court review to force compliance with mandated deadlines, which have often been missed. 58 The EAA already provides for court suits to enforce statutory license processing deadlines; 59 this provision could be extended to cover other deadlines, such as for review of the Control List and for decontrol of items not multilaterally controlled. The review would be purely procedural; courts would not second-guess the substance of any decisions. 60 There are swifter, more severe ways of enforcing deadlines. One approach is to make a missed deadline act as an acquiescence by the Administration. The 1988 Act did this for determin ations of foreign availability requested by firms; if BXA missed the deadline, foreign availability would be assumed and the item decontrolled. 61 This provision was effective in speeding up those determinations. However, BXA reports that the provision strained its resources, and took effort away from other investigations of foreign availability that, while not requested, were perhaps more important to industry as a whole. BXA also states that that provision has the potential to force U.S. decontrol before it is authorized by CoCom, thereby undercutting the multilateral export control system that the United States is trying to strengthen. Another option is to make decontrol self-executing-that is, when decontrol is mandated by law (e.g., on items not controlled by other nations) the decontrol would take effect even if BXA has not published implementing regulations. Congress considered such a provision in 1990. 62 However, the absence of regulations could cause confusion, resulting in exports detrimental to national security. One practical, nontechnical way to facilitate timely adjustment of controls would be to increase the staffing in the Commerce Departments Office of Foreign Availability (OFA). OFA determines when foreign availability of items makes U.S. controls ineffective. These determinations are crucial to minimizing the drag on competitiveness, and they require difficult fact gathering and complex technical analysis. OFA now has about two dozen people. Another option would be to encourage political appointees with technical background. Technically knowledgeable senior BXA staff might be better able to argue their positions with other agencies. Finally, there is an emerging problem of NorthSouth or foreign policy controls, issued under Section 6 of the EAA (see box 4-C). (The term North-South is a convenient shorthand to distinguish the orientation of these controls from that of East-West controls. However, the use of this term is not meant to imply that all or most developing countries give cause for concern.) Some foreign policy controls aim to prevent proliferation of weapons of mass destruction, including missiles and chemical and biological weapons. 63 Other foreign policy controls, such as sanctions against countries that abuse human rights, are meant instead to make a political statement. Still other controls, such as sanctions against nations that use terrorism, appear to do both.
PAGE 68
62 Competing Economies: America, Europe, and the Pacific Rim The Gulf War heightened concern over proliferation, and prompted the Administrations Enhanced Proliferation Control Initiative (EPCI). Although this concern is justified, the Administrations use of foreign policy controls takes little account of competitiveness concerns. The Administration must report and just@ foreign policy controls annually to Congress, but the safeguards against unnecessary interference with commercial exports (e.g., the prohibition on controlling items with foreign availability and the prohibition on unilateral controls) do not apply to foreign policy controls. Thus, changes already adopted for export controls related to East-West national security matters (under Section 5) do not extend to controls directed to foreign policy purposes (Section 6). This made it possible for the Administration to impose unilateral controls in March 1991 on certain chemicals and manufacturing equipment that could be used to make chemical weapons. In August 1991 the Administration issued regulations that could be interpreted to require virtually all firms exporting any items to any countries to set up a monitoring and control system to guard against diversion of any products to chemical or biological weapons plants (ch. 4). If Congress believes that competitiveness concerns should, when possible, apply in the same way to foreign policy controls, a first step might be to separate foreign policy controls with military objectives from those with political objectives. To achieve political objectives, unilateral controls or controls on items available elsewhere could be appropriate and effective. Congress could include a strongly worded statement of policy that export controls for political reasons should be issued only after careful consideration of the effect on commercial exports. Congress might put foreign policy controls with military objectives under more or less the same discipline as national security controls. This is not a simple matter. It would not make sense, for example, to simply state that proliferation controls will henceforth be treated under Section 5 rather than Section 6, because controls under Section 5 are meant to keep items from former Communist bloc nations, and the law is written so as to coordinate controls through CoCom. However, control of the technologies for nuclear weapons, missiles, chemical weapons, and biological weapons is broader than an East-West issue. The Soviet Union and the Peoples Republic of China possess much of the important technology, and must join in controls for them to be effective. Similarly, to deny conventional weapons to countries sponsoring terrorism would also require the cooperation of the Soviet Union and China. Therefore, Congress might wish to treat foreign policy controls with the objective of denying military technology in a separate section of the law, which imposes discipline regarding, e.g., foreign availability, but recognizes their special international position. GOVERNMENT-INDUSTRY PARTNERSHIPS One thing that stands out in the story of nations that successfully use trade policy to promote the development of particular industries is that trade policy alone is not enough. Even when used aggressively, it is combined with promotion of technology development and diffusion, with risksharing between government and industry, and with support for adoption of new technologies and industrial success. Strategic Technology Policy The least intrusive and least expensive of several possible risk-sharing options between government and industry is what OTA has described as strategic technology policy, an R&D partnership for developing new technologies of commercial interest. 64 The potential benefits to society of such ventures are great, but the likely payoff to individual firms is too small to make it worth their taking all the risk. Therefore, the argument for industry/government risk-sharing takes on special force. Traditionally, U.S. policy has been to limit R&D support to basic science or else to the governments own needsprimarily, military security. There are notable exceptions: agriculture and civilian aircraft manufacture have had longtime steady support for technology development, based on a frank recognition that they were important to the nations economic welfare. With the dawning awareness that U.S. industries really are in competitive trouble, a consensus seems to be growing for a Federal role in commercially promising R&D. Congress took a first step in the 1988 Trade Act, which launched a small program for R&D partnerships, the Advanced Technology Program (ATP) located in the Commerce Departments National Institute for Standards and Technology (NIST). The Programs purpose is to help U.S. business rapidly
PAGE 69
Chapter 2-Policy Issues and Options l 63 commercialize new scientific discoveries and apply research results toward refining manufacturing technology. Through the ATP, NIST can assist private R&D ventures with technical advice or can actually participate in them: it can provide start-up funding; put up a minority share of the cost; or lend equipment, facilities, and people. Congress has consistently taken the lead with ATP, providing its first funding of $10 million in fiscal year 1990 and raising the ante to $36 million in fiscal year 1991. The Administration, for its part, has moved toward support of cooperative R&D for commercial ventures. The Presidents 1992 Budget proposal said: The Administration believes that appropriate Federal investments in applied civilian R&D can result in high payoff to the economy . 65 The Budget document went on to say that the principal strategy for Federal applied civilian R&D is to support agency mission requirements. But some of this R&D has broad applications in the private sector. In such cases, the government would support generic or enabling technologies at the precompetitive stage of R&D. These terms were defined as follows: generic or enabling technologies have the potential to be applied to a broad range of products or processes across many fins; precompetitive R&D is the stage of the R&D process where the results can be shared widely within and between industrial sectors, without reducing the incentive for individual firms to develop and market commercial products and processes based on the results. The intention of such definitions is to avoid favoring particular firms or industries, putting government money into technologies with broad applications across firms and sectors. This principle ties in with the idea that technologies with many spillovers and applications are of most potential benefit to society. It is not always possible, however, to distinguish so neatly between technologies and industries. For example, flat panel displays are a generic technology, having myriad applications from home television sets to engineering work stations to airplane cockpits. But government support for developing the technology benefits the electronics industries and firms that produce it. Congress has already indicated its interest in an industry-government partnership for applied commercial technology development that is not necessarily dependent on fallout from other government missions. If this program is to take a proactive, coherent approach, rather than responding to crisis calls from industries under competitive siege or to a wave of enthusiasm for the latest technology, it would need an agency in charge and a set of guiding principles. These issues are discussed briefly below. 66 A Civilian Technology Agency (CTA) would be needed to guide government-industry cooperative R&D whether or not the idea of broader, integrated competitiveness policies in support of selected critical industries wins acceptance. NISTs ATP might in time become a fill-fledged CTA, although it was not created with that explicit mission. Bills to establish more formally an Advanced Civilian Technology Agency in a new Department of Industry and Technology (which would replace the Department of Commerce) were introduced in the 100th and 101st Congresses. 67 These proposals defined the agencys mission as contributing to U.S. competitiveness by supporting long-term, high-risk projects likely to yield important benefits to the Nation but that lack adequate private support. A bill that passed the Senate in 1989 would have given the ATP a similar mission and authorized substantially increased funding, up to $100 million per year. Any CTA would have to start small, as the ATP has, and need never grow very large. A possible model is the Defense Advanced Research Projects Agency (DARPA), a small free-wheeling Department of Defense (DoD) agency supporting risky long-term R&D that often has commercial as well as military value. DARPA has 150 employees, about half of them scientific and technical, and some $1.5 billion a year for its research projects. Its relatively small size is one factor in a nonbureaucratic culture that gives staff members a great deal of freedom to exercise their own good judgment. At the same time, a research budget of $1 to $2 billion a year seems large enough to attract a critical mass of competent staff and fund a healthy portfolio of technologies. Where in the Federal bureaucracy a CTA is placed may not matter too much. The prestigious National Science Foundation is an independent agency. DARPA is smaller than NSF and is a tiny part of the huge, hierarchical Department of Defense, but it too has won renown for its competence and dedication. If the small, experimental ATP develops into a mature CTA, the question may answer itself, since
PAGE 70
64 Competing Economies: America, Europe, and the Pacific Rim the Program is already started life in the Department of Commerce. Like DARPA, a CTA might attract excellent staff by combining freedom from bureaucratic rules with great responsibility and the opportunity to serve ones country. Unlike DARPA, a CTA would not serve a defined mission and customerthe military. Instead, technologies supported by a CTA would have to prove their worth in the market. Even allowing for inevitable failures-and an agency charged with supporting high-risk technologies would not be worth its salt if it had no failures along with its successesit is much tougher to choose technologies that can make it commercially than ones with some credible military use. Collaboration with industry is essential in choosing technologies for support. If private companies are not interested enough to take some of the risk and do some of the work, then the chances of commercial success are probably remote. Joint funding helps the government escape pressure from special interests in selecting technologies for support, and enlists market forces in picking the best prospects. At the same time, a CTA would need a set of guiding principles to delineate broad areas appropriate for government-industry collaboration. One obvious principle is preference for technologies with wide applications in many products and industries. Another is knowledge-intensiveness, which means not only technologies important to industries that are clearly knowledge-intensive in themselves (e.g., computers) but also projects that could deepen the knowledge-intensiveness of traditional industries (e.g., precision machining). Another principle is potentially large markets. The importance of collaboration with industry in selecting commercially interesting projects is underscored by NASAs (National Aeronautics and Space Administration) experiences over the years. NASAs annual spending for aeronautical R&D amounts to about $800 million, and probably 90 percent of that is, or could be, relevant to the commercial aircraft industry, as well as to the military. This is a huge amount for the United States. Of course the military connection is a leading reason for this level of spending; even so, the research is available to commercial producers. Yet it would be mistake to think that most of NASAs R&D, or even the major share, gives U.S. aircraft manufacturers a competitive advantage. Some of it is basic research, not applicable to commercial production except possibly in the very long term. Much of the advanced technology development is quite freely available to the world, and some has been used first by Airbus in Europe, not by U.S. producers (see ch. 8). NASAs greatest contributions to competitiveness of American producers are in two areas: its facilities (e.g., wind tunnels and the Numerical Aerodynamic Simulator, a supercomputer complex), which are either preferentially or solely available to U.S. companies; and technology development projects in which the U.S. companies were close collaborators (as in the E 3 program for aircraft jet engines, described in vol. 2, to improve fuel efficiency and reduce emissions and noise). Coordination between new government regulations and collaborative technology development projects to help comply with the regulations can give an extra boost to competitiveness; this was the case with Federal regulations to reduce aircraft noise. There are at least two pluses to industry involvement in NASAs technology development projects. First, the projects are more likely to reflect genuine commercial concerns and possibilities; and second, the company engineers gain an intimate knowledge of the new technologies that outsiders cannot acquire simply by reading published research reports. These advantages are just as valid outside NASA and the commercial aircraft industry. The few government-industry technology partnerships that already exist for manufacturing industries (apart from aircraft) follow the pattern of at least 50percent funding by industry. Sematech, the largest of these ventures, gets $100 million per year both from the U.S. Government and from a consortium of industry members. The ATP follows a similar rule, with more than half the cost of cooperative projects paid by sponsoring firms. Although Sematech has its own facilities, the ATP-funded R&D takes place in members labs. There are opportunities to do more cooperative work in Federal labs, especially in the Department of Energys well-equipped multiprogram national labs. However, judging by NASAs experience, these ventures will be of more use to companies if they involve participation by the firms own researchers, not just a financial contribution. So far, the U.S. Governments offers of collaborative R&D projects have been snapped up by private
PAGE 71
Chapter 2-Policy Issues and Options l 65 companies. Sematech was, in fact, proposed by the semiconductor industry, which lobbied hard for it. Three small pilot projects ($5 to $6 million per year total) in the national labs for commercializing high-temperature superconductivity had more willing partners from industry than the labs could fund, and the same was true of DARPAs $30 million project for cooperative R&D on high resolution display technologies. ATPs first batch of grants for cooperative projects amounted to about $9 million, initiating R&D programs that are expected to cost $100 million (including private funds) over 5 years. The 11 winning grants were selected from 249 proposals requesting a total of $150 million (box 2-B provides details). Added together, the dollar amounts in these few cooperative programs are minute in a Federal budget of more than $1 trillion. They are tiny compared to the more than $90 billion per year that U.S. manufacturers spend for R&D. It is noteworthy, however, that such very modest programs have drawn responses from so many companies, large and small. Among ATPs first 11 grantees were industry giants such as Du Pent, AT&T Bell Laboratories, and two prominent industry consortia, the Microelectronics and Computer Technology Corp. and the National Center for Manufacturing Sciences. Although these companies and consortia have big R&D budgets of their own, each one funds a great many projects, and there are always promising but risky ones that do not make the corporate cut. However, such projects may look more attractive with cooperative funding, including government money. Furthermore, the government backing may lend a certain prestige to the undertaking. Or companies may fear missing out if their competition joins in and they do not. So far, Congress has taken a gradual approach to expanding government partnerships with industry on commercially relevant R&D. This approach makes sense, considering that the U.S. Government has little experience with industrial partnerships; that the governments institutional ability to manage R&D partnerships is at an early stage; and that relations between government and industry in this country have traditionally been adversarial. If Congress wishes to continue this measured expansion, with the ultimate goal of having an agency about equal in size and importance to DARPA or the National Science Foundation, it may opt for a modestly increasing ATP budget for several years. It might also wish to consider writing into law a more formal statement of goals for the agency. Participation by foreign firms in cooperative R&D programs that receive government funding is a new and somewhat unsettled question. Part of the problem is in defining just what a foreign firm is (see ch. 3). In appropriating funds for the ATP for fiscal year 1991, Congress took on the problem in an innovative way. It set standards that apply to U.S.-owned as well as foreign-owned firms, thus bypassing ownership as the central criterion for deciding whether a firm can participate in ATP projecta. 68 The Secretary of Commerce is authorized to decide whether firms are eligible, using the standards to determine that their participation would be in the economic interest of the United States. The standards applying to all firms call for investments within the United States in research, development, and manufacturing, including the manufacture of major components or subassemblies (thus insisting that investments go beyond assembly plants that add little value or knowledge-intensiveness); a significant contribution to employment in the United States; and agreement to promote U.S. manufacture of products resulting from ATPassisted technology projects. Further conditions apply to foreign-owned fins: they may participate if the Secretary finds that their home country offers U.S.-owned firms comparable opportunities to take part in joint ventures for technology development, allows U.S. companies to invest on equal terms with other countries, and affords adequate protection of the intellectual property rights of U.S. companies. 69 These provisions offer guidance but also give great latitude to the Secretary of Commerce in determining the eligibility of firms, both foreign and domestic, for ATP projects. Congress may wish to exercise substantial oversight for a time on how these novel provisions are carried out. Financial Risk-Sharing A strategic technology policy, worthwhile as it may be, goes only so far. Government partnership in technology development stops short of commercialization. After that, it is up to industry to make the much larger investments in the product design, manufacturing equipment and tooling, worker training, and acquisition of know-how by managers and production engineers that are necessary for the commercial manufacture of new or improved prod-
PAGE 72
66 l Competing Economies: America, Europe, and the Pacific Rim Box 2-BThe Advanced Technology Programs First Round The U.S. Department of Commerce announced in March 1991 the first awards under the innovative Advanced Technology Program (ATP). The ATP was created by Congress in 1988 to help American business turn research results into new commercial products and improved manufacturing technologies. It established a governmentindustry partnership, in which the ATP could lend technical assistance, equipment, and people to cooperative research projects and could contribute a minority share of the funding. Awards of about $9 million went to 11 grantees and were first installments in R&D programs expected to cost $100 million over 5 years (with more than half the money coming from private industry). l The 11 winners were chosen from 249 proposals requesting $150 million in first year grants, Several hundred volunteer reviewers from both government and industry helped to make the choices. Of the 11 grants went 5 to joint ventures or consortia-an indication, according to Commerce Department officials, that the program is fostering a greater spirit of cooperation among highly individualistic companies for precompetitive R&D. 2 The program encourages joint efforts and rewards them with grant money that covers some indirect as well as direct costs. Most of the technologies were related to microelectronics and computers, including optical recording and computer hardware. Others were in the fields of high-temperature superconductivity, machine tool control, and novel laser designs. The grantees, their projects, and the grants they requested, are described below. 3 Printed Wiring Board Interconnect Systems National Center for Manufacturing Sciences, Inc. (NCMS) Printed wiring boards are ubiquitous in electronic products, from radios to computers. The U.S. share of a $25 billion world market has dropped from 42 to 29 percent in 3 years. Current technology is approaching fundamental limits in the materials and processes now used. Four members of the NCMS consortium, AT&T, Texas Instruments, Digital Equipment Corp. and Hamilton Standard Interconnect will work with Sandia National Laboratories to develop new materials, better processes, and improved technical understanding. First year request: $2,370,000; total 5-year request: $13,783,000; matching funds: $14,674,000 Volume Holographic Mass Storage Subsystem Microelectronics & Computer Technology Corp. (MCC) MCC proposes to build on a basic concept it has already demonstrated for a radically new form of dense, ultra-fast computer memory storage, potentially replacing slow disk drives and magnetic tape (about 100,000 times slower than the typical microprocessor). The new system would respond in microseconds, and would store data as holographic images in photorefractive crystals. First year request: $823,000; total 5-year request: $10,331,000; matching funds: $12,700,000 Advanced Manufacturing Technology for Low Cost Flat Panel Displays Advanced Display Manufacturers of America Research Consortium Advances are needed in testing and repair equipment, as well as in interconnection and packaging technologies, to successfully commercialize high-quality, low-cost flat panel displays. Three relatively small companies will lead a consortium in a linked series of research programs; they are Optical Imaging Systems (Troy, NY), Photonics Imaging (Norwood, OH), and Planar Systems, Inc. (Beaverton, OR). Seven other companies are participating. First year request: $1,251,000; total 5-year request; $7,305,000; matching funds: $7,604,000 Short Wavelength Sources for Optical Recording National Storage Industry Consortium Data storage devices are a $50 billion-per-year industry; two-thirds of the market is now controlled by U.S. firms, one-third by Japanese. The most promising technology in the field is optical recording, as used in compact disks. The program would develop an integrated short wave length laser source, with solid state components, for read/write heads of fast, small, rugged optical memory devices. Members of the consortium, which is not yet fully 1~ ~ome ~5e5, pm of tie Cornpuy contribution is in kind (e.g., labomtory equipment), so that tie gov~ent gr~t r~est~ maY be larger that the cash outlay proposed by the company. 2~~AdvaWd ~~oIo= BOP CIWS Another Hurdle, New Technology Week, Mu. 11, 1991. 3~e des~ptio~ ~e men ~m UOS. Department of Commerce, office of the SeCretary, Comerce News, press rekase dated Mar, 5, 1991 and First Winners in the Advanced Technology Program, New Technology Week, Mar. 11, 1991.
PAGE 73
Chapter 2--Policy Issues and Options l 67 formed, include Applied Magnetics, Bernoulli Optical Systems, Eastman Kodak IBM, Maxoptix Corp., and the University of Arizona. An initial grant of $50,000 is contingent on further development of the joint venture. First year request: $50,000; total 5-plus years request: $5,421,000; matching funds: $9,200,000 Fabrication and Testing of Precision Optics for Soft X-Ray Projection Lithography AT&T Bell Laboratories X-ray lithography is a key technology for new generations of dense microelectronic circuits. This program will attack a key problem limiting projection X-ray lithography: the manufacture, testing, and assembly of relatively large scale X-ray mirrors. It will develop technology to test, fabricate, assemble and align aspherical X-ray mirrors. Three-quarters of the grant will go to small business subcontractors. First year request: $955,000; total 5-plus years request: $2,000,000; matching funds: $3,525,000 Solid State Laser Technology for Point Source X-Ray Lithography Hampshire Instruments, Inc. and McDonnell Douglas Electronic Systems Co. These small companies will lead a joint venture to exploit recent advances in laser materials for use in low cost, high-performance X-ray lithography. First year (total) request: $1,090,000; matching funds: $1,094,000 Nonvolatile Magnetoresistive Semiconductor Technology Nonvolatile Electronics, Inc. Computer memory is volatile--the data vanish when the power is shut off. This program aims to develop a fast, dense, nonvolatile memory, basing the technology on a magnetoresistive memory (MRAM) patented by Honeywell and intended for space and avionics applications. The company has licensed the technology for nonaerospace applications, and means to develop it as a competitor to conventional dynamic random access memory (DRAM) semiconductors. First year request: $599,000; total 3-year request: $1,738,000; matching funds: $869,000 New User-Interface for Computers Based on On-Line Recognition of Natural Handwriting Communication Intelligence Corp. (CIC) CIC plans a robust natural handwriting recognition system that does not require training the computer to recognize each individuals handwriting (a key limitation of most current systems). Cursive handwriting input for computers could be a revolutionary advance, especially for notebook and laptop machines. First year request: $671,000; total 2-year request: $1,264,000; matching funds: $912,000 Advanced Thallium Superconductor Technology E.I. du Pent de Nemours & Co. The proposal is to develop thin-film fabrication techniques for a new, proprietary high-temperature superconductor invented by Du Pent. Project includes developing fabrication techniques and creating representative superconducting electronic devices to demonstrate feasibility. First year request: $370,000; total 3-year request: $1,590,000; matching funds: $784,000 Tunable Deep UV and VUV Solid State Laser Source Light Age, Inc. The company will apply recent developments in laser technology to produce high-average power ultraviolet lasers that are cheaper, safer, more reliable and easier to use than current products. Potential applications are in medical and scientific instruments and materials processing. It could be particularly significant in photolithography for the semiconductor industry, challenging the dominan t Japanese suppliers of semiconductor production equipment. First year request: $627,000; total 1.5-year request: $701,000; matching funds: $254,000 Advanced Compensation Techniques for Enhancing Machine Tool Accuracy Saginaw Machine Systems, Inc. The program seeks a general, economic solution to the problem of correcting for errors in machining caused by thermal expansion and contraction of the machine tool; thermal errors contribute to about half of the errors in machining. Working with the University of Michigan, Saginaw will develop a mathematical model of thermal errors and a sensor and computer control system that can help machine tool builders greatly improve the accuracy of their machines at reasonable cost. First year request: $266,000; total 2-year request: $540,000; matching funds: $168,000
PAGE 74
68 l Competing Economies: America, Europe, and the Pacific Rim ucts or the adoption of new manufacturing process technology. It is normal and expected for private industry to make these investments, take the risks and then, if all goes well, reap the rewards. Sometimes, however, in some critical sectors, private investment is inadequate from the standpoint of social benefits. Take semiconductors. One important reason why U.S. companies have lost out to Japanese firms over the last decade is that Japanese rates of investment were higher. From 1982 until 1991, Japanese producers invested a larger share of their sales of integrated circuits in plants and equipment than did U.S. producers, and since 1984 have out invested their U.S. rivals in absolute amounts .70 This pattern holds true in all industries throughout both economies; Japanese investment in machinery and equipment as a share of gross domestic product has consistently run at double the U.S. rate since the mid1970s, and in the late 1980s the discrepancy widened. This bodes ill for U.S. competitiveness generally, but in the technologydependent semiconductor industry, where new generations of products are introduced every 3 years or so, the disadvantage can be crippling. The relatively meager U.S. investments in new production equipment reflect high capital costs and a financial environment that discourages or fails to support long-term investment (see the summary discussion in ch. 1 and more detailed discussion in OTAs 1990 report, Making Things Better). Macroeconomic policies that lower interest rates and provide stability are probably the most important things government can do to encourage greater investment in technology development and deployment by U.S. manufacturing companies. There are other options as well, however. Besides the generally unsupportive financial environment in the United States, American semiconductor companies have the added handicap of facing Japanese competitors that are much larger; are vertically integrated, from semiconductor chips through computers; make a much greater variety of end products (VCRs and compact disk players as well as computers); and have ample retained earnings for new investments. Yet formidable as they are today, Japanese companies were not always preeminent. In the 1970s, they had a catch-up job to do that was greater than the one facing U.S. companies today. They did it not only by hard work and effective management but also with government policies that gave them protected domestic markets, tight controls over foreign investment, guaranteed sales to important gov ernment customers, gov ernmentindustry R&D partnerships, and a variety of measures assuring plenty of low-cost capital at a time when companies financial resources were much more limited (see ch. 6). Government policies to share financial risks with industry can take the form of subsidies or loans on advantageous terms, or they can take the more indirect form of tax breaks (i.e., tax expenditures). Both put burdens on government resources and, under the Federal budget agreement adopted in 1990, Congress cannot opt for either without compensating tax rises or spending cuts in other domestic programs. Of the two forms of financial risk-sharing, tax expenditures are more within U.S. traditions and experience. U.S. companies in the past have received accelerated depreciation and tax credits for capital investments, and they currently get a tax credit for R&D (although Congress has not made this a permanent feature of the tax system but instead has renewed it from year to year). Although certain kinds of investment (e.g., real estate) have been singled out for specially favored tax treatment, Congress has not in the past designed these tax incentives to improve the competitiveness of particular industries. 71 If Congress wishes to target tax breaks to selected industries because of their importance to the U.S. economy, the best way to do it is make the tax measures part of a comprehensive strategy that also includes such things as R&D partnerships and trade policy. Many governments have supported selected industries with more direct financial aid, in addition to tax incentives. Japan, for example, offers companies hojokin (success dependent loans) for risky enterprises in selected industries; payments can wait for a positive cash flow. Thus companies are insulated against catastrophic losses. European governments have done much the same for Airbus. The U.S. Government, by contrast, has had little experience in giving direct financial aid to selected industries for strategic competitive purposes. There have been some well-publicized government bailouts of failing individual companies, notably Lockheed and Chrysler. The Synfuels program of the late 1970s did have the purpose of energy independence; that program is generally considered a failure. The broadest U.S.
PAGE 75
Chapter 2-Policy Issues and Options 69 experience with direct loans to industry (leaving aside special loan programs for small business) was the Reconstruction Finance Corp. (RFC) in the post World War II years. 72 RFC was created in 1932 to shore up banks in the credit deflation of the Great Depression, and it went on to procurement of strategic materials for national defense during the war. The postwar RFC had no such clearly defined purpose. In practice, it spent most of its very ample resources (close to $1 billion per year in 1946-47, translated into 1990 dollars) in bailing out large, prominent but floundering companies, such as the Kaiser-Fraser automobile company. The one RFC project that seemed to have any focus other than keeping a big firm afloat was a series of nine loans to the Lustron Corp., which boasted a new technologymanufactured steel-frame houses. The project failed, and accounted for RFCs biggest loss. RFCs failure in the postwar years was surely due in large part to its lack of any strategic purpose or guidance. However, its failure also suggests some dangers inherent in governments giving large sums of money to companies that cannot get funding from private sources. This cautionary lesson has been learned so well that now even suggestions for direct financial aid from government to industry are few and far between. In its 1989 report to the President and Congress, the National Committee on Semiconductors (NACS) did propose an attenuated form of government loan guarantees to U.S. companies trying tore-enter the advanced consumer electronic business. 73 The idea was that the semiconductor industry is handicapped by the lack of a U.S. consumer electronics sector, considering the remarkable convergence of technologies in consumer electronics and other electronics goods (e.g., computers). NACS proposed a private corporation, to be backed by pledges of support from Federal, State and local governments, that would provide low-cost patient capital to startup consumer electronics companies with U.S. ownership. The suggestion was not repeated in the second NACS report in 1991. Some committee members privately offered the explanation that members could not agree on technical details for the proposed corporation, and said that the committee may take up the issue again in its third and final report. The likelihood of a revived consumer electronics sector in the United States under U.S. ownership is virtually nil without infant industry assistance from the government; this might include trade protection and technology partnership as well as financial aid. However, whether such a revival is necessary or important to the U.S. economy is no simple question. It includes the issue of whether foreign-owned companies producing in the United States (Sony, Philips North America, Thomson) might provide the same benefits. Moreover, reviving a consumer electronics industry is a big, expensive job, and could be quite a burden to the taxpayers even if government took only a minor share of the risk. If Congress does wish to consider direct financial aid to this or other selected industries, it may want to start with a program of very modest scale and increase it slowly. Having a large pot of money available for such ventures invites imprudence, even for an experienced institution. And the United States does not now have an institution with the experience or capacity to fit financial aid into a coordinated strategic competitiveness policy. Government Purchases Government procurement was a vital factor in the birth and early growth of several important U.S. industries: semiconductors, computers, aircraft, and aircraft engines. As might be inspected, the big buyer in each case was the Department of Defense (DoD). For semiconductors, for example, the amounts DoD spent in the early years were tiny compared to the sums spent today, yet the DoD then was the launch customer for a product and technology in its infancy, and bought nearly 100 percent of the industrys output. Today, the Federal Government might still be a valuable first customer for untested products that combine public benefits with the potential for competitive success. It might also be an important customer for existing products important to U.S. competitiveness. However, the opportunities are somewhat limited. Ninety percent of Federal purchases of goods are for defense and DoD does buy large amounts of certain important products. The share of aircraft and aircraft engine production for defense is about 43 percent; for radio and TV communications equipment, 36 percent; and for electronic components, 23 percent. But much of this equipment is so highly specialized for military use that technological spillovers to the commercial side are limited; in fact, because of long lead times for developing weapons, some commercial technologies are far ahead of military applications. Technological spillovers from the military appear to be diminishing, although the evidence is
PAGE 76
70 l Competing Economies: America, Europe, and the Pacific Rim mixed. What is certain is that restrictive laws and DoD regulations have made it increasingly difficult for companies to take advantage of whatever technological synergies may exist. Nondefense purchases comprise less than 10 percent of the total Federal purchases of goods, only $10.3 billion in 1990 out of $107 billion total. The United States has no national telecommunications service, no national railway, no national airline, no national health service, and no national university system. In both Japan and Europe, government bodies of these kinds have been important buyers of such products as semiconductors, computers, telecommunications equipment, aircraft, rolling stock, and medical equipment, and have used their purchases to support domestic industries. The Federal Governments track record in improving manufacturing competitiveness, whether as a launch customer for new products or an important customer for established products, is weak. This is partly because competitiveness is not a goal of U.S. procurement policy. Preferences for U.S. goods are not motivated by strategic economic considerations; any benefit to commercial manufacturing competitiveness is usually just a happy coincidence. In contrast, the United States most important trading partners do use procurement to promote certain manufacturing industries. Key U.S. trading partners have much less open procurement than does the U.S. Government. The United States has been trying to make foreign procurement markets more accessible. First, the United States is negotiating to expand the scope of the GATT Procurement Code. Under the Code, the United States and the other signatories, which include the United States major trading partners, grant reciprocal access to each others procurement markets for covered purchases. However, the Code currently covers only a modest amount of those purchases. 74 Second, the United States has begun renegotiating several of the 19 Memoranda of Understanding (MOUs) with allies regarding defense procurement. These MOUs have substantially opened the U.S. defense procurement market but have not had the same effect on allies defense procurement markets. Third, the United States has negotiated under Title VII of the Omnibus Trade and Competitiveness Act of 1988. 75 Under Title VII, the U.S. Trade Representative is to identify cases of procurement discrimination, including important cases not yet covered by the GATT Procurement Code, and to negotiate improved market access. If negotiations are unsuccessful, the United States can retaliate by discriminating in its own procurement against goods from the country in question (see ch. 4). The U.S. Government might try other tactics to improve access to foreign procurement markets. If some Code signatories appear more willing than other to open their procurement markets, the United States might then consider creating a kind of Gold Club Membership GATT Code with countries that agree to abide by very strict rules and enforcement procedures. Members would have broad reciprocal access to cosignatories public contracts. There could even be a Gold Club Federal Contracts Journal in which participants advertise their procurements no later than they are announced elsewhere. However, based on experience with other types of market barriers, a strategy based primarily on opening foreign procurement is likely to bring disappointing results. If closed foreign procurement is hurting important U.S. industries, the United States could also use domestic measures, such as R&D support and tax breaks, to promote the industries in question. In addition, the United States could use its own procurement strategically to develop important technologies and industries. Some measures could be taken consistent with U.S. obligations under the Current GATT Procurement Code and MOUs. For example, current U.S. law restricts defense purchases of supercomputers to U.S. machines. This has been an important help to the U.S. supercomputer industry, and a change in that policy would increase the already substantial risk that the United States will lose dominance to Japan (see ch. 6). As another example, the GATT Procurement Code does not now cover the U.S. Postal Service. The Postal Service would be an ideal launch customer to develop an electric vehicle industry, since the limitations of electric vehicles under current technology-short range and need for frequent recharging-would not be a problem for the vehicles used to deliver local mail. Procurement of U.S.-made electric vehicles by the Postal Service could provide the United States a sharp advantage in a new industry for which the United States, Europe, and Japan are all competing to develop the technology. If the United States considered an electric
PAGE 77
Chapter 2--Policy Issues and Options 71 vehicle industry worth promoting, it might want to keep the Postal Service outside the Procurement Code until the industry was well on its way. Of course, the United States might still decide to subject the Postal Service to Procurement Code discipline in exchange for concessions by other countries. The important thing is that such opportunities not be bargained away thoughtlessly. It is not clear how many such opportunities exist. Much Federal procurement is subject to neither the GATT Procurement Code or the defense MOUs. Even the best estimates are rough. In 1990, Federal procurement of goods and services amounted to $229.6 billion, of which $181.5 billion was for defense. 76 Of the $48.1 billion in nondefense spending, perhaps roughly $4.6 billion was covered by the Procurement Code, 77 leaving $43.5 billion uncovered. Of the defense procurement, probably at most $109.2 billion was covered by MOUs and/or the Procurement Code, 78 leaving at least $72.3 billion uncovered. While some of these uncovered amounts are for services (beyond the scope of this report), probably at least $5.7 billion in uncovered nondefense procurement is for goods, 79 as might be a substantial portion (perhaps $30 billion or more) of the uncovered defense procurement. 80 Probably only a small part of the uncovered procurement could be of strategic importance. However, Congress might wish to mandate that these strategic opportunities be assessed. Even when purchases are covered by the GATT Procurement Code or other international agreement, the U.S. Government could help industry without breaking international rules. This code does not cover R&D contracts, so the government could award such contracts to U.S. firms to enhance their position to bid on covered purchases. For example, if the Postal Service were to be covered by the Procurement Code, it could still award R&D contracts relating to electric cars to U.S. firms before soliciting bids for a fleet. There might also be a competitiveness bonus in standards development. The GATT Procurement Code permits countries to require national standards in government contracts; companies at the forefront in developing both standards and corresponding products would have a head start in winnin g the contracts. R&D funding can be applied to develop national standards that are then included in product specifications, as with the Research and Development in Advanced Communications in Europe Program (RACE) of the European Community (see ch. 5). In addition, all government contract specifications might be shaped to conform as closely as possible to commercial products made by U.S. industry, or to planned commercial products. For example, the specifications on a purchase of electric cars by the Postal Service might be written to make production of those vehicles a stepping stone toward U.S. fins planned commercial production. 81 Such a purchase might also be timed to fit the product cycle of U.S. fins. With direction from Congress and the President, each agency could seek out opportunities such as these where Federal procurement and associated R&D spending can help competitiveness. However, if one agency had overall responsibility for competitiveness, it might work with other agencies to find and coordinate such opportunities. That agency could do the same with State and local governments, advising them on how working their procurements could help U.S. competitiveness. While some State and local governments have some preferences for U.S. goods, they are not coordinated into any national policy or strategy. The agency could act as a clearinghouse to help State and local governments find U.S. suppliers. State and local procurement could be a powerful tool for competitiveness. One reason is size. In 1990, State and local government spending on goods was $87 billion, compared with the Federal Governments spending of $107 billion (all but $10 billion for defense). For certain items, State and local purchases are a significant part of the market. For example, in 1990 States and localities spent an estimated $8 billion on cars and trucks, 82 amounting to 6 percent of U.S. motor vehicle sales that year. Also, State and local governments are not now governed by the GATT Procurement Code. 83 Thus, State and local governments have more freedom than the Federal Government to grant preferences for U.S. firms. INSTITUTIONS FOR A STRATEGIC COMPETITIVENESS POLICY In each of the policy areas discussed above, the Federal Government could take many constructive actions. Government efforts to encourage invest-
PAGE 78
72 l Competing Economies: America, Europe, and the Pacific Rim ment and savings, enhance human resources, encourage commercial technology development and adoption, and take a more proactive stance in trade policy could help many industries become more competitive. Yet a host of individual actions, undertaken piecemeal, do not add up to a coherent strategy. At present, the Federal Government does not have an institutional structure capable of a strategic, integrated approach to competitiveness and trade policy. As part of a more proactive approach, the Federal Government might choose to focus various government policies on assistance to critical commercial industries. One element in this strategy is development of criteria that would keep the list of eligible industries short and highly focused. Those on the short list might qualify for priority technology or financial assistance offered by Federal agencies or possibly for special consideration in U.S. trade policy. If Congress wishes to proceed with this kind of initiative, stronger institutional capacity in the Federal Government would be needed to provide the careful analysis of trade and competitiveness issues that would lead to rational choice of strategies and industries. At the very least, the Government would need the analytic capability to identify candidate industries, to develop action-oriented strategies tailored to specific industries, and to delineate specific government actions, programs and policies. The strategy would need to be industry led, with eligible industries extensively involved in shaping feasible approaches and selecting candidate industries, and in making substantial investments in whatever government-industry partnerships are undertaken. Government decisionmakers would need to make sure that the choices are consistent with the overall goals of the program and criteria for selection. A coordinated strategy to support critical industries might not get much attention if it were assigned to an existing line agency already saddled with numerous trade and industry responsibilities. Thus, Congress might establish a new organization in the executive branch to develop strategic competitiveness policies. The office could be small, since its role would be largely analytical and catalytic. But it would have to be well and prominently positioned (e.g., in the Executive Office of the President) to effectively leverage or influence actions of key trade and commerce agencies, and it would need Administration support to have clout. A strategic competitiveness policy will accomplish little unless it prompts action on the part of the many Federal agencies with responsibilities in such diverse areas as foreign trade, research and development, antitrust, and taxation. Recognition is growing among Federal agencies that there are many links between policies affecting domestic industry and foreign trade. However, the two are seldom carried out in concert. Thus, a strategic competitiveness policy would require a coordinated response by Federal agencies. A certain amount of restructuring and reorganizing of current Federal functions could help. For many years, Congress has been debating whether to rearrange the wide array of Federal trade and commercial programs into new departments or agencies. A purpose underlying most of the proposed reorganizations is to create a more coherent organizational structure for U.S. Government decisions on international trade. If Congress does establish such a department, strategic competitiveness policy could be one of its responsibilities. At the end of this section, two organizational options are discussed in greater detail, in light of the criteria and objectives discussed immediately below. The two options are not exclusive of each other. The more modest option, establishing a critical commercial industries office in the executive branch to formulate and coordinate implementation of strategy in support of critical industries, could well be part of a more far reaching departmental and trade agency reorganization that has as one its goals furthering strategic competitiveness policy. Congress might also use advisory committees as a first step in identifying industries and recommending actions for subsequent adoption by the executive branch or Congress. For example, the newly launched Competitiveness Policy Council might undertake this function, 84 while the National Advisory Committee on Semiconductors (see box 2-C) could serve as a model for developing strategies for specific industry sectors. Although advisory committees can be useful in identifying problems and needs, only agencies and departments have the authority to implement policies or coordinate Federal responses. Hence, legislation or additional executive action would still be required.
PAGE 79
Chapter 2--Policy Issues and Options 73 Box 2-C-National Advisory Committee on Semiconductors The National Advisory Committee on Semiconductors (NACS) is an example of a joint effort by industry and government to address the problems of a specific industry sector--in this case the troubled American semiconductor industry. Congress established the committee in the 1988 Omnibus Trade and Competitiveness Act, calling on NACS to devise and promulgate a national semiconductor strategy. Its final report, scheduled for late 1991, is expected to lay out an overall strategy, including possible trade policy approaches. In the meantime, NACS has issued two interim reports outlining the problems of the American semiconductor industry and making specific recommendations for action by Congress and the administration. The committee is structured to assure extensive interaction between industry and government in developing the strategy. NACS is technically an independent advisory body in the executive branch. Eight prominent industry executives serve on the committee (the president of AT&T Bell Labs serves as chairman), as do five high-ranking government officials with key responsibilities for research and development or technology policy. l (The agencies include Defense, Commerce, the Energy Department, the National Science Foundation and the Office of Science and Technology Policy (OSTP).) A Defense Advanced Research Projects Agency official saves as Executive Director, under agreement with OSTP. The committee, which meets every other month, also has setup several working groups to address specific issues. The working groups typically include some committee members and outside experts. NACSs first two reports, issued in late 1989 and early 1991, recommended several government actions as initial steps toward achieving a national strategy. Some of the recommendations were quite general, such as improving the education and training system, and creating a favorable business environment for all industries. Others were specific to the semiconductor industry, emphasizing, for example, research and development actions Federal agencies could take to further semiconductor technology. It is probably premature to talk about NACSs overall impact on executive branch policy, Congress, or the industry itself, since it has yet to issue its strategy. Because Federal officials serve on NACS, its suggestions may have influenced some agency R&D actions. It also has helped elevate visibility of semiconductor issues in the Administration. (Committee members met with President Bushs chief of staff, John Sununu, as well as Office of Management and Budget Director Richard Darman and Michael Boskin, who chairs the Presidents Council of Economic Advisors.) NACS's interim reports have also helped define the terms of the debate about semiconductor problem areas. INonf~~ COmmiti rn~bra are appointed by the lhsiden~ through the Office of Science imd lkdmoIosY Poliq. n ~w s~tes thatfourof thenonfederalmembers were to be selected from the semiconductor industry; the other four weretobe eminent in technology, defense and economic development. A proactive strategy does not mean that governreason for protection is simply that an industry has ment would produce a blue print or plan for industry. the political leverage to gain it. Instead, the critical industries organization could champion competitiveness, and other domestic policies. It could encourage agencies to give priority for some forms of government assistance (e.g., technology help) to critical industries. It could also encourage agencies to take competitiveness concerns into account when making regulatory decisions that affect a critical industry. Selecting Industries There are many reasons why governments adopt protective stances toward specific industries, ranging from national defense, to economic security, to a desire to mitigate the impact of import competition on communities, firms, and workers. Often, the Implicit in the concept of strategic competitiveness policy is the idea that certain critical industries are important for national economic security. Doubtless, the most promising candidates would come from manufacturing industries in the technological forefront for developing new products and processes. Such industries have a high proportion of technology-oriented workers and spend proportionately more on R&D. Examples include computers and software, electronic components, communication equipment, advanced materials, precision machining equipment, robotics, biotechnology, and aerospace. Some critical industries would coalesce around new technologies that could in time provide dispro-
PAGE 80
74 Competing Economies: America, Europe, and the Pacific Rim portionate returns to the firms or countries with the largest market share. There is not much disagreement about what technologies are critical to national economic prosperity and national security. Lists developed by several U.S. Government and business organizations in 1990-91 are remarkably similar (table 2-1). 85 And the conclusion is widely shared that U.S. industrys position in these technologies has weakened significantly in the past 10 years, and continues to weaken. The benefits from commercializing these technologies could be large, as suggested by Department of Commerces 1990 estimate that 13 emerging technologies might yield $356 billion in annual product sales in the U.S. market alone by the year 2000. Yet if current trends continue, the United States would lag Japan in most of these technologies and the European Communities in several. 86 The element of criticality suggests not only technology intensity but also the potential to provide good new jobs and make large contributions to the economy. Thus, a critical industry might be an emerging or developing high-technology industry with large potential markets. Or a critical industry might produce technology or services that enable other industries to make dramatic advances in productivity and quality. The classic examples are machine tools and semiconductors. A rationale for a Federal role in developing a U.S. high-definition television (HDTV) industry is that it could drive technologies critical for other parts of the electronics industry. Consumer demand for HDTV could be sizable. This is one of several emerging industries in which Japan is ahead. It is important to recognize that not every industry with a claim to leading edge technologies, growing markets, good jobs, and technology spillovers could be selected for support. Public and private resources, energy, and talents are limited. The judgment of industry leaders and their willingness to put up their own money in cooperative ventures are an invaluable guide to the selection of promising industries. This implies the necessity to develop new relationships of trust and shared visions of truly national interest between government and industry. What Kind of Support? The range of policy tools that might be used to support a strategic competitiveness policy includes those discussed in previous sections, such as technology partnerships, financial incentives, and trade policy. Some of the relevant programs already exist, or at least are on the books. To round out the array of policy tools, Congress could authorize some new ones. Options for expanding the list of these programs are touched on only briefly here, as they were discussed in more detail earlier in the chapter. As noted above, financing long-term investment has been a particular problem for American industry, reflecting the high costs of capital and pressures to realize short-term profits. Federal policy affecting investment might be tailored to meet specific needs of a designated critical industry. For example, across-the-board loosening of depreciation rules for all American industry would be very expensive, but there might be merit to loosening the rules for specific critical industries. The public costs might still be considerable but the stimulus would at least be focused on the specific needs of industries found critical to the national interest. The National Advisory Committee on Semiconductors, for example, estimates that changing current depreciation rules for new investments in semiconductor manufacturing equipment from 5 to 3 years would cost the U.S. Treasury $180 million in lost tax receipts, but could prompt $450 million in capital investment each year by this industry .87 Almost by definition, technological advance will be a key factor in the competitiveness of critical industries. The creation of a CTA along the lines discussed earlier, or a major expansion of NISTs ATP, could help. While a coherent policy in support of commercially important technologies does not yet exist in this country, the Federal Government does take part in technology development that is useful to some commercial industries, notably NASAs aeronautics R&D program and the Sematech project. Generally, however, such support has not been part of a systematic effort to further U.S. competitiveness. Sematech, for example, was created ad hoc, in response to strong industry pressure and the argument that a competitive U.S. semiconductor industry is essential to national defense. As an organizing concept, promotion of critical industries could give direction to future government support of commercial technology. Whatever the institution, adequate funding will be critical. ATP is a tiny program in a diverse agency. The overall NIST budget is the same in real terms today as it was two decades ago. Even if NIST's budget doubles in the
PAGE 81
Chapter 2-Policy Issues and Options l 75 Table 2-lComparison of National Critical Technologies with Department of Commerce Emerging Technologies and Department of Defense Critical Technologies National critical technologies Materials l Materials synthesis and processing l Electronic and photonic materials l Ceramics l Composites l High-performance metals and alloys Manufacturing l Flexible computer integrated manufacturing l Intelligent processing equipment l Microand nanofabrication l Systems management technologies Information and Communications l l l l l l l Software Microelectronics and optoelectronics High-performance computing and networking High-definition imaging and displays Sensors and signal processing Data storage and peripherals Computer simulation and modeling Biotechnology and Life Sciences l Applied molecular biology Medical technology Aeronautics and Surface Transportation l Aeronautics l Surface transportation technologies Energy and Environment l Energy technologies l Pollution minimization, remediation, and waste management Commercial emerging technologies l Advanced materials l Advanced semimconductor devices Superconductors l Advanced materials l Flexible computer integrated manufacturing l Artificial intelligence l l l l l l l l High-performance computing Advanced semiconductor devises Optoelectronics High-performance computing Digital imaging Sensor technology High-density data storage High performance computing Biotechnology l Medical devises and diagnostics Defense critical technologies l Composite materials l Semimconductor materials and microelectronic circuits l Composite materials l Machine intelligence and robotics l l l l l l l l l l l l l Software producibility Semiconductor materials and microelectronic circuits Photonics Parallei computer architectures Data fusion Data fusion Signal processing Passive sensors Sensitive radars Machine intelligence and robotics Photonics Simulation and modellng Computational fluid dynamics l Biotechnology materials and processes l Air-breathing propulsion l No National Critical Technoiogies counterpart: High energy density materials, Hypervelocity projectiles, Pulsed power, signature control, Weapon system environment SOURCE: U.S. Department of Commerce, Emerging 7bchnologies:A Survey of 7bchnica/and Economic Opportunities, Spring 1990; and U.S. Department of Defense: Critica/ T4whno/ogies P/an, 15 March, 1990; as cited in The National Critical Technologies Panel, Report of the Nationa/ Critica/ TAno/ogies Pane/ (Washington, DC: U.S. Government Printing Office, March 1991), table 2. next 5 years, as President Bush has proposed, ATP year (see earlier discussion), though it would take may not get much of the increase. At its present size time for the agency to develop the staff and ($36 million in fiscal year 1991), ATP is only a experience to manage a program of that size. beginning. It is not yet adequate for a governmentindustry technology partnership that is big and broad If Congress wishes to authorize a critical indusenough to make a difference to the whole economy. tries program, it could also take action to give Whether seen as an outgrowth or eventual replacecritical industries special priority in other areas of ment for ATP, a CTA might well begin small. But a government decision making. For example, in the mature program of technology support would probaantitrust area, firms in a critical industry might be bly need to be budgeted at $1 billion to $2 billion per extended the same kind of protection for joint
PAGE 82
76 Competing Economies: America, Europe, and the Pacific Rim manufacturing or joint production activities that are now given to cooperative R&D activities under the National Cooperative Research and Development Act of 1984. (Special action for critical industries would not be needed if Congress decides to amend the 1984 law along these lines for all industries. Several bills to accomplish this have been proposed in recent Congresses, including S.479 as introduced in the 102d Congress.) In the same vein, Congress could require Federal agencies to prepare competitiveness impact evaluations before promulgating policies likely to have a major adverse impact on the international competitiveness of a designated critical industry. By focusing only on critical industries, the sea of paperwork that might be created by broader impact statement requirements might be avoided. As noted, strategic competitiveness policy would allow occasional departures from the prevailing free trade philosophy. Such departures would not be frequent, but the ability to take the initiative in matters of trade policy would be a necessary part of the overall strategy. Once critical industries are identified, a special interagency group could be set up to consider trade policy actions that would promote the competitiveness of the industry; the critical industries organization would be responsible for coordinating trade policies with other strategies in support of the selected industries. As mentioned earlier, trade negotiators could be directed to give critical commercial industries top priority in dealings with other countries on fair trade and market access issues. The presence of a competitiveness champion within U.S. Government would strengthen the hand of U.S. negotiators in such dealings. The most important job for a government body responsible for support of critical commercial industries would be to tie together the policy strands in a reasonably coherent whole. Of course, no government agency can ever establish a neat, wholly coherent policy on any broad national issue, whether it be fiscal policy, health, education, environment, or competitiveness. There will always be some messiness, conflicts and overlap between agencies, struggles between special interests and the national interest, compromise and disarray. Because the U.S. Government and the American people have relatively little experience in government-industry partnerships, it would be overoptimistic to expect the creation of a coherent strategic competitiveness policy to be easy or rapid, even assuming a consensus in support of such a strategy. A modest start and evolutionary growth is a more reasonable expectation. But the aim of coherence would nevertheless have to be steadily pursued. Otherwise, it is too easy to be sidetracked into spreading available resources too thinly, or hijacked into serving special interests. Institutional Alternatives Institutional arrangements for developing and implementing a critical commercial industries approach could take many forms. Two possibilities are discussed below. A Critical Industries Office A small office in the executive branch with an elite staff could serve as the lead agency in developing and implementing strategic competitiveness policy. The office might be placed in a department, or, in a willing Administration, the Executive Office of the President. In either case, championing critical industry strategies and serving as a catalyst for action by Federal agencies could be part of the offices mandate. Proposals in recent Congresses to create an office of competitive analysis might be a starting point. As proposed in H.R. 1274, a trade reorganization proposal introduced but not acted on in the 101st Congress, the office would report each year on the competitive prospects of American industries, and could empanel temporary industry councils to advise on needed changes in Federal policy with respect to specific industries. Even if Congress stopped short of trade reorganization, it could direct the Administration to create a critical industries office within an existing department. The Department of Commerce, which now administers several trade and technology programs, might be a logical place. Locating the office in the Executive Office of the President would be a good option in an Administraion that is supportive of the critical industries approach. The Executive Office can bring high visibility and government-wide perspective to issues. However, such a location is likely to be ineffective in an Administration hostile to the concept. While private industry input would be indispensable, a critical industries office would need a strong
PAGE 83
Chapter 2-Policy Issues and Options l 77 staff. First, the process of identifying critical industries would demand highly competent personnel. An even more demanding task for the staff would be to encourage Federal actions and policies that make a real difference in the competitiveness of critical industries. If Congress were to direct the Administration to set up such an office, it might consider exempting the office from the normal civil service guidelines on hiring and dismissal of employees. To assure fresh thinking, Congress could direct that no more than half of the initial staff could be drawn from existing agency personnel. Congress might also make sure that the agency had the resources to actively recruit from industry, universities, and research institutions. Government Reorganization and Strategic Industry and Trade Policy A major change in current government organization for trade and competitiveness policy is not a prerequisite for a critical industries approach. However, the way the government organizes its functions does affect policy outcomes, and competitiveness policy is no exception. If Congress wishes to promote a more proactive approach to trade and competitiveness issues, then reorganization could produce a more focused policymaking apparatus and more direct lines of authority for carrying out the policy. The current structure for trade decisionmaking within the Federal Government is diffuse, with dozens of agencies having roles to play and a bewildering array of interagency task forces playing coordination functions. No fewer than 10 departments, 2 independent agencies, and 4 executive office agencies take part in trade policy formulation, and the actions of many other agencies and departments can affect international trade. The Commerce, State, Treasury, and Agriculture Departments all powerfully influence trade policy, as do some independent agencies (the International Trade Commission). The U.S. Trade Representative (USTR) has statutory responsibility for trade policy coordination and negotiation, but (as noted) its staff is stretched thin by the issues of the day (e.g., securing a Uruguay Round agreement, launching negotiations for a Mexican Free Trade Agreement). Responsibility for other government functions affecting the competitiveness of American industry is similarly diffuse. Dozens of agencies have regulatory responsibilities that, in large and small ways, can influence industrial competitiveness. Several agencies-commerce, Defense, Energy, the National Science Foundation-have R&D responsibilities relevant to industry. Efforts to coordinate functions among agencies is predictably sporadic; the White House Office of Science and Technology Policy (OSTP) in the Bush Administration, for example, has breathed life into interagency coordinating committees that a few years earlier had been all but abandoned. Sprawling, decentralized policy structures may have adequately served the overall strategic interests of the United States throughout most of the post World War II period. Yet the absence of a powerful voice within government for the international competitiveness of American industry almost assures that other objectives (e.g., foreign policy, national security) begin with the stronger hand when disputes arise, whatever the substance of the matter. Moreover, the lack of central focus and direction, far from underpinning g the U.S. Government commitment to free trade, has resulted in trade policy with quite a few contradictions and apparent exceptions. Some in Congress have proposed government reorganization as a means to improve the focus and effectiveness of Federal trade and commercial policy functions. Some bills would establish a new Department of International Trade and Industry, assuming many functions now carried out by the Office of the USTR, the Commerce Department, and some export financing agencies. 88 Other proposals have called for a Department of International Trade and Investment or Department of International Commercial Policy, consolidating USTR and the trade policy units of several existing Departments into one agency, and establishing a cabinet committee to coordinate international economic policy .89 Another set of proposals have called for creation of a Department of Industry and Technology, building on existing Commerce Department authority for export promotion and creating a Civilian Technology Agency. 90 Some proposals to set up an industry and technology department would also create an independent U.S. Trade Administration, comprised of the Office of the USTR and Commerce Department agencies responsible for trade administration and international economic policy .91 (The USTR would continue to serve as a cabinet rank official).
PAGE 84
78 l Competing Economies: America, Europe, and the Pacific Rim In the end, the specific bureaucratic arrangements outlined in these proposals are less important than the substantive goals they try to achieve. Nor should any of these arrangements be expected to eliminate coordination problems. As a practical matter, not all the Federal agency functions related to competitiveness could ever be consolidated into one department. Many key financial, trade, and technology policy functions would continue to be carried out elsewhere. Moreover, executive office coordination of these functions would still be needed. It is hard to conceive, however, that a coherent competitiveness policy can succeed without a strong agency heading up the effort. 1 Two earlier reports in OTAs assessment of Technology, Innovation and U.S. Trade consider many of these issues in detail. Paying the Bill: Manufacturing and Americans Trade Deficit, OTA-ITE390 (Springfield, VA: National Technical Information Service, 1988) discusses the place of manufacturing in international trade and the U.S. economy. Making Things Better: Competing in Manufmturing, OIYdlT!443 (Washington DC: U.S. Government Printing Office, February 1990) considers how technology can improve U.S. manufacturing competitiveness, including four possible areas for government action: the financial environment, human resources, diffusion of technology, and strategic technology policy. Other recent, closely related OTA reports are Worher Training: Competing in the New International Economy, OTA-ITE-457 (Washington, DC: U.S. Government Printing Office, September 1990) and Commercializing High Temperature Superconductivity, OTA-ITE-388 (Washington DC: U.S. Government Printing Office, August 1988). 2 U.S. Congress, Office of Technology Assessment Paying the Bill, op. cit., passim, and Making Things Better, ibid., passim; MIT Commission on Productivity, Made in America: Regaining the Productive Edge (Cambridge, MA: The MIT Press, 1989) passim. 3 The agreement was the 1986 U.S.-Japan Semiconductor trade agreement, renewed with substantial modifications in 1991. U.S. producers access to the Japanese market (which was informally part of the agreement) has been lower than U.S. companies and negotiators deem acceptable, although U.S. companies Japanese market shares did improve after the agreement took effect. While there is dispute about whether the agreement caused higher prices for memory chips, the price hikes that followed the STA also hurt the computer industry. For further explanation of this agreement, see ch. 4. 4 Chalmers Johnson MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925-1975 (Stanford, CA: Stanford University Press, 1982). 5 The discussion below is based on options detailed in Making Things Better, op. cit., ch. 2, and Worker Training, op. cit., ch. 2. 6 From 1977 to 1988, the U.S. cost of capital for investments in machinery and equipment with a physical life of 20 years averaged 3.4 percentage points higher than similar costs in Japan and 4.9 percentage points higher for investments in a factory with a physical life of 40 years. (Robert N. McCauley and Steven A. Zimmer, Explaining International Differences in Capital Costs, Federal Reserve Bank of New York Quarterly Review, summer 1989, pp. 7-28.) Although some international differences in capital costs are due to risks related to currency fluctuations, sustained differences of the magnitudes shown by McCauley and Zimmer are unlikely under free market conditions. Despite liberalization in the 1980s, the Japanese financial market is not yet as open as that in the United States. Also, tax incentives and exemptions are widely used in Japan to lower capital costs and promote investment, often in specific industries or technologies. Furthermore, the Japanese system of stable and mutual shareholding means that only a small portion of companies shares are traded on financial markets, thus relieving pressure on management to show short-term profits. See Making Things Better, op. cit., ch. 3, for a discussion of the costs of capital and conditions of availability in Japan and the United States. 7 The 1990 budget agreement between Congress and the Administration, which partitions spending into defense, domestic nondefense, and foreign nondefense categories, lasts through three fiscal years, from 1991 through 1993. Afterwards, the three categories may again be merged so that it would be possible, say, to pay for tax expenditures in the non-defense part of the budget with spending cuts in defense. 8 The Joint Committee on Taxation estimated the cost in lost revenues of the investment tax credit at $13 billion to $37 billion per year in the 1980s, and $8 billion to $64 billion per year for the Accelerated Cost Recovery System. Both programs were ended in the tax reform law of 1986, in favor of an overall reduction in the corporate tax rate (although certain classes of assets are still granted fast depreciation). 9 These policy options for training of active workers are drawn from the OTA report Worker Training, op. cit., ch. 2. 10 A bill to setup such a program, Section 404 of S. 2114, was introduced in the 101st Congress. 11 For a detailed discussion of these options, see Making Things Better, op. cit., pp. 53-71. Descriptive and analytic material on technology diffusion in the United States and other countries is in chapters 6 and 7 of that report. 12 Fewer States applied for the Centers opening in 1991 than for those beginning 2 years earlier (20 proposals compared with 36). Commerce Selects Two Manufacturing Centers, New Technology Week, Mar. 11, 1991. A smattering of other Federal programs offer some technology extension services. The largest of these is Trade Adjustment Assistance for firms, which has been funded at about $9 million per year in recent years. 13 Making Things Better, op. cit., p. 58. 14 This is direct government loans only and does not count loan guarantees, which are about twice as great. In 1987, direct loans to small and medium enterprises from the three main government financing institutions amounted to Y3.8 trillion ($30 billion at Y125 to the dollar); loans guaranteed by the 52 nationwide credit associations amounted to Y7.8 trillion ($62 billion). D. H. Whittaker, New Technology Acquisition in Small Japanese Enterprises: Government Assistance and Private Initiative, contract report to the Office of Technology Assessment, May 1989. 15 See Making Things Better, op. cit., pp. 192-194. 16 On Mar. 1, 1991, DOE signed an agreement with Martin Marietta Energy Systems, which runs Oak Ridge National Laboratory, for a class waiver on inventions proposed by Martin Marietta for commercialization (subject to national security exceptions). Class waivers had previously been granted only to non-profit operators of DOEs national labs; this was the first such agreement with a for-profit operator. Paul Kemezis, Oak Ridge Lab Receives Approval to Sign Deals With U.S. Industry, New Technology Week, Mar. 11, 1991. 17 For example, the National Center for Manufacturing Sciences organizes shopping trips to the labs on behalf of its over 100 industry members, seeking to pair lab expertise with industry needs. This complements a longer established effort from the labs side, the Federal Laboratory Consortium, which operates a small core staff and 300 volunteers from the labs to match inquiries from firms with the appropriate lab researcher. 18 See McGraw-Hills Tech Transfer Report, February 1991, for several recent examples. 19 Ibid. 20 Bill Loveless & David Kramer, Expanded Missions Eyed for DOEs National Labs, McGraw-Hills Tech Transfer Report, February 1991, p. 13.
PAGE 85
Chapter 2--Policy Issues and Options l 79 21 Software written in contractor-operated Federal labs can be copyrighted and then licensed to firms. An exclusive license is sometimes a necessary incentive for a firm to commercialize the software. However, software written in government-operated labs is written by government employees and cannot now be copyrighted, making it difficult or impossible to grant firms exclusive rights. 22 The National Cooperative Research Act of 1984 lessened the legal risks of joint R&D. See P.L. 98-462, 15 U.S.C. 43014305. 23 Malcolm Baldrige National QuaIity Improvement Act of 1987, Public Law 100-107, Sec. 3. 24 Jeremy Main How to Win the Baldrige Award, Fortune, Apr. 23, 1990, p. 101. 25 Ibid. 26 These categories are drawn from the report of a private group, the Council on Competitiveness; see Gaining New Ground: Technology Priorities for Americans Future (Washington, DC: The Council, 1991). The government reports include U.S. Department of Commerce, Technology Administration, Emerging Technologies: A Survey of Technical and Economic Opportunities (Washington, DC: Department of Commerce, 1990); U.S. Department of Defense, Critical Technologies Plan, report for the Committees on Armed Services, U.S. Congress, AD-A234-900 (May 1, 1991); Government of Japan, Ministry of International Trade and Industry, Trade and Future Tusks in Industrial Technology (Sangyo Gijutsu no Doko to Kadai) (Tokyo: 1988); Commission of the European Community, First Report on the State of Science and Technology in Europe (Brussels: 1988). 27 The program began in October 1988. U.S. Congress, Office of Technology Assessment, Making Things Better, op. cit., p. 188, box 7-B. 28 Foreign market barriers, their effect on U.S. industry, and the U.S. response are discussed in more detail in ch. 4. In addition, barriers imposed by the EC, Japam Korea and Taiwan are discussed in chs. 5-7. 29 Clyde V. Prestowitz, Jr., Alan Tonelsom and Robert W. Jerome, The Last Gasp of GATTism, Harvard Business Review, March-April 1991, p. 137. 30 See box 4-A. In the GATT Uruguay Round, Canada and the EC have proposed transforming GATT into an ITO; consideration of this proposal has been deferred until after the Round is completed. 31 19 U.S.C. 2411-2420. 32 This is the pot calling the kettle black. Many countries consider their laws superior to GATT when they deem it necessary. 33 This case is discussed in detail inch. 4. 34 Clyde V. Prestowitz, Jr., Trading Places: How We Allowed Japan To Take The Lead (New York NY: Basic Books, 1988), pp. 283, 50-52. 35 Ibid., pp. 276-277. 36 Protection can be granted in certain other cases-for example, under Section 201, as described below. Protection has also at times been instituted by means of bilateral agreements limiting the quantity of imports of certain products from particular countries. Examples are textiles, machine tools, and automobiles. Except for textiles, in which the bilateral agreements come under an overall arrangement sanctioned by GATT these bilateral agreements at least arguably violate GATT Protection against imports is discussed in detail in ch. 4. 37 19 U.S.C. 2251-2254. 38 19 U.S.C. 2253(a)(3). 39 From 1980 through 1990, U.S. industry petitioned for relief under Section 201 20 times. The required injury was found in seven cases (though in two only for some of the products at issue), and the U.S. industry received some protection in five (see table 4-8). 40 19 U.S.C. 2252(b)(l)(B). 41 GATTs escape clause (Article XIX) requires injury to domestic producers of like or directly competitive products, and U.S. law tracks this language (19 U.S.C. 2251). In contrast, GATTs provisions regarding subsidies and dumping (Article VI) permit a showing that the dumping or subsidization retard[s] materially the establishment of a domestic industry, and U.S. law follows that language (19 U.S.C. 1671(a)(2), 1673(2)). 42 The language of GATTs escape clause is general, requiring only that the increased imports cause or threaten serious injury to domestic producers. of like or directly competitive products. There has been little interpretation of this language by GATT dispute resolution panels. Negotiations underway in the Uruguay Round could clarify the escape clauses scope. The United States seeks to restrict rather than expand this scope. 43 GATT Article XIX, par. l(a). 44 At times protection could also be justified under GATTs national security exception (Article XXXVIII). This could be the case, for example, with protection of the domestic semiconductor industry, which furnishes the U.S. military with essential components of weapons. However, the United States and other countries have been cautious about using that exception for fear of setting precedents that would make it into an all-purpose tool to justify a great deal of protection. 45 In the time it takes to reach this point the U.S. industrys condition might improve enough so that it could get along without protection. 46 The terms of the EC proposal to the Japanese on automobiles had not been made public at this writing; this description is based on information from trade sources. 47 U.S. Commerce Department International Trade Adrministration, Office of Management personal communication, Aug. 2, 1991. 48 U.S. General Accounting Office, Export Promotion: Problems in Commerces Programs, NSIAD-89-44 (Gaithersburg, MD: U.S. General Accounting Office, 1989), p. 17. 49 The Japanese effort included 3 officers at the Japanese embassy and 78 employees of the Japan External Trade Organization (JETRO), an agency of MITI. Those figures are from Akemi Yoshida, Attach in Management and Coordination, Japanese Embassy, Washington DC, Aug. 21, 1991; and Hiro Sate, Director of Research JETRO (New York Office), Aug. 21, 1991. The figures for U.S. staffing in Japan are from an interview with Ed Stumpf, Director, East Asia and Pacific, USFCS, Aug. 21, 1991. 50 United States Trade and Development Program, 1990 Annual Report, pp. 27, 29, 55-61. 51 H.R. 2508 and S. 1435, in conferences of August 1991, would do both. 52 50 U.S.C. App. 2401-2420. This Acts authorization lapsed as of Oct. 1, 1990, but President Bush continued its provisions in force by invoking the International Emergency Economic Powers Act 50 U.S.C. 1702. See executive order 12730-Continuation of Export Control Regulations, Sept. 30, 1990. 53 Public Law 100-418, sec. 2414, codified at 50 U.S.C. App. 2504(a)(5). 54 Title II of the Omnibus Trade and Competitiveness Act of 1988, Public Law 100-418. 55 50 U.S.C. App. 2409(g). 56 S. 320 passed the Senate on Feb. 20, 1991, and was referred jointly to the House Committee on Foreign Affairs and the House Committee on the Judiciary on Feb. 26, 1991. This bill resembles the conference version of H.R. 101-4653 and S. 101-2927, as passed by the House and Senate in 1990, but omits certain provisions that President Bush found objectionable in vetoing that bill. See President George Bush, Memorandum of Disapproval for the Omnibus Export Amendments Act of 1990, Nov. 16, 1990. 57 S. 320, Sec. 117. 58 A 1990 conference report on this same provision commented
PAGE 86
80 l Competing Economies: America, Europe, and the Pacific Rim that [s]ome provisions of the 1988 amendment to the EAA were not implemented or were implemented far beyond statutory deadlines. H. Rept, 101-944, p. 75. 59 50 U.S.C. App. 2409(j). 60 S. 320, Sec. 124. 61 50 U.S.C. App. 2402(f)(3)(B). 62 This was proposed in H.R. 4653, Sec. 120, but deleted in conference. See House Conference Report 101-944 (on H. 4653), p. 75. 63 Dual-use nuclear technology comes p rimarily under another statute, Section 309(c) of the Nuclear Non-Proliferation Act of 1978, codified at 42 U.S.C. 2139a(b). See Section 17(d) of the EAA, at 50 U.S.C. App. 2416(d). 64 See Making Things Better, op. cit., especially pp. 71-89. 65 Budget of the United States Government: Fiscal Year 1992, part two, p. 47. 66 For more detailed discussion see Making Things Better, op. cit., pp. 71-79. 67 One of the bills, S. 1233 in the 100th Cong., was reported out of the Senate Committee on Governmental Affairs, attached to the 1988 trade act, but was then dropped. Two similar bills, H.R. 3838 and S. 1978 were introduced in the 10lst Cong. 68 U.S.-owned companies are defined as those that have majority ownership or control by citizens of the United States. 69 U.S. Congress, House of Representatives, Conference Report [to accompany H.R. 5021] making appropriations for the Departments of Commerce, Justice and State, the Judiciary, and related agencies for the fiscal year ending Sept. 30, 1991, and for other purposes, report 101-909, Oct. 20, 1990, pp. 25-26. 70 These comparisons are for merchant producers, who sell semiconductors on the open market. Companies that produce only for their own use (captive production) are excluded; IBM the worlds biggest producer, is in this category and therefore excluded. All major Japanese producers are included. 71 In practice, capital-intensive and R&D-intensive industries tend to get more benefits than others from the tax incentives. 72 A succinct description and evaluation of the postwar RFC is in Arthur T. Denzau and Clifford M. Hardin, A National Development Bank: Ghost of the RFC Past, formal publication number 62, Center for the Study of American Business, June 1984. 73 A Strategic Indusrtry at Risk, a report to the President and the Congress from the National Advisory Committee on Semiconductors (Washington DC: The Committee, 1989), p. 20. See also Making Things Better, op. cit., pp. 80-89. 74 Recent data have been classified by the U.S. Government. Estimates based on detailed data are publicly available only for 1981, the Codes first year of operation. In that year, the Code was estimated to have covered only $18 billion in U.S. procurement and $4.2 billion in all foreign procurement combined. (For comparison in 1985 U.S. procurement, and procurement by all other Code signatories, each totaled in the range of $200 billion.) The amount of actual trade under the Code in 1981 appeared to be much smaller+ 3.3 billion for U.S. imports and $270 million in U.S. exports. And the amount of this trade that would not have occurred without the Code appeared to be at most $210 million for U.S. imports and perhaps substantially less than $270 million for U.S. exports. (See ch. 4.) 75 Public Law 100-418, Title VII. 76 U.S. Department of Commerce, Bureau of Economic Analysis. BEA breaks this down as follows: defense goods, $96.7 billion defense services, $84.8 billion nondefense goods, $10.3 billion nondefense services, $37.8 billion. (Procurement of services does not include compensation paid to employees.) 77 To get this result, proportion estimates that area few years old are applied to the 1990 figures. Based on testimony in 1989, the Code covers about 10 percent of total Federal procurement. W. Douglas Newkirk, Assistant U.S. Trade Representative for GATT Affairs, Office of the United States Trade Representative, testimony at hearings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27, 1989, p. 6 (p. 56 of hearing print). Based on GAO testimony in 1989, describing procurement during 1985 through 1987, Code-covered procurement is split roughly 80 percent for DoD and 20 percent for other agencies. Adam Cowled, Senior Evaluator, U.S. General Accounting Office, personal communication, July 24, 1990, interpreting U.S. Congress, House Committee on Government Operations, International Procurement and Waivers of the Buy America Act: U.S. Business at a Disadvantage, H. Rept. 101-989, Nov. 29, 1990, p. 7 fn. 6, and Allan I. Mendelowitz Director, Trade, Energy and Financial Issues, United States General Accounting Office, testimony at hearings before the House Subcommi ttee on Legislation and National Security, Sept. 27, 1989, p. 30 of Committee print. Thus, in 1990 about 8 percent of $229.6 billion, or $18.4 billion, is Code-covered defense procurement, while about 2 percent of $229.6 billion, or $4.6 billion, is Code-covered nondefense procurement. 78 Probably roughly 50 percent of defense procurement, or $90.8 billion, is covered by MOUs (ch. 4). Perhaps about $18,4 billion of defense spending is covered under the Procurement Code (see previous footnote). Assuming that there is no overlap between Code and MOU coverage (which yields the highest possible figure), the combined coverage is $90.8 billion plus $18.4 billion, or $109.2 billion. 79 The Procurement Code covers only goods and services incidental to the purchase of goods. Even if all $4,6 billion of the Code-covered nondefense procurement is for goods, that still leaves $5.7 billion in non-Code-covered nondefense procurement, since the total nondefense procurement of goods is $10.3 billion (see first footnote in this paragraph). 80 Defense procurement for 1990 consisted of $96.7 billion for goods and $84.8 billion for services (see first footnote in this paragraph). While the $18.4 billion of Code-covered defense procurement is likely almost all goods (see previous footnote), it is hard to tell how the $90.8 of MOU-covered procurement is divided between goods and services. If it were divided proportionately to the total spent on defense goods versus services, $48.4 billion would be for goods. This yields a total of at most (assuming no overlap between Code and MOU coverage) $18.4 billion @ IS $48.4 billion or $66.8 billion in defense goods covered by the Code or MOUs, leaving at least $29.9 billion of uncovered defense goods. 81 Other countries could argue that such actions violate the GATT Procurement Code, Article IV, which states that [technical specifications. shall not be prepared, adopted or applied with a view to creating obstacles to international trade nor have the effect of creating unnecessary obstacles to international trade. However, the point is debatable, especially if the specifications are shaped merely to be convenient for U.S. firms rather than to deliberately be inconvenient for foreign firms. 82 This estimate is based on Commerce Department data for State and local purchases of durable goods in 1990 ($33 billion) and the percentage of such purchases going for motor vehicles in 1982 (24 percent at that time-no more recent figure is available). 83 However, the EC has taken the position that if the Procurement Codes coverage is expanded State and local government procurements in the United States should be covered. 84 The Competitiveness Policy Council (CPC) was established by the 1988 trade act (Public Law 100-418) but did not get underway until early 1991. Its members are appointed by the President and by Senate and House leaders from both parties; it is to report annually to the President and Congress. It should not be confused with the nonstatutory Council on Competitiveness in the office of Vice President Quayle, which evaluates the impact of Federal regulation on competitiveness, or the private Council on Competitiveness. Congress will need to act if the CPC is to continue; the Bush Admini strations proposed budget says that the Council will complete its work in 1991 and disband in 1992, with Vice President Quayles competitiveness council as suming responsibil-
PAGE 87
Chapter 2-Policy Issues and Options 81 ity for regulatory and other competitiveness issues. To date, the Vice Presidents council has not shown evidence of a willingness to address the broad range of issues assigned to the CPC. 85 Also broadly similar are lists of critical technologies developed by the Japanese Ministry of International Trade and Industry and the Commission of the European Community. 86 Technology Administration U.S. Department of Commerce, Emerging Technologies: A Survey of Technical and Economic Opportunities (Washington DC: U.S. Department of Commerce, spring 1990). 87 NationaI Advisory Committee on Semiconductors, Toward A National Semiconductor Strategy, 2d annual report, N.A.C.S., Arlington, VA, February,1991, p. 15. 88 Bills to establish a Department of International Trade and Industry include, among others, S.121 as reported by the Senate Committee on Governmental Affairs in the 98th Cong.; S. 1365 as introduced in the 99th Cong.; H.R. 1338 and H.R. 2135 as introduced in the 100th Cong., and H.R. 1274 as introduced in the 101st Cong. 89 For an analysis of this approach see Stephen D. Cohen The Making of United States International Economic Policy: Principles, Problems and Proposals for Reform (New York NY: Praeger, 1988), pp. 248-271 90 See S.1978, the proposed Trade and Technology Promotion Act of 1989, as introduced in the 101st Cong. and S.1233, the proposed Economic Competitiveness, International Trade and Technology Development Act of 1987, as reported by the Senate Committee on Governmental Affairs on June 23, 1987. (Senate Report 100-82) 91 S.1233 (100th Cong.) as reported by the Senate Committee on Governmental Affairs on June 23, 1987. See S. Rept. 100-82 for discussion.
PAGE 88
Chapter 3 American Firms, Foreign Firms: Contributions to the Nation
PAGE 89
CONTENTS Page SUMMARY AND FINDINGS . . . . . . . . . . . . . . 85 INTRODUCTION . . . . . . . . . . . . . . . . . 87 THE BASICS: WHO INVESTS WHERE AND IN WHAT . . . . . . . . 88 U.S. Multinationals Activities at Home and Abroad . . . . . . . . . 88 FDI in the United States. . . . . . . . . . . . . . . . 88 FDI AND U.S. MANUFACTURIN G COMPETITIVENESS . . . . . . . 90 Contributions to Knowledge and Technology . . . . . . +...,..., . . 95 Japanese FDI: A Special Case . . . . . . . . . . . . . . 97 EMPLOYMENT AND EMPLOYEE COMPENSATION . . . . . . . . 101 PROFITS . . . . . . . . . . . . . . . . . . . 104 INTERNATIONAL TRADE . . . . . . . . . . . . . . . 106 NATIONAL INTERESTS, BUSINESS INTERESTS . . . . . . . . . 108 The Policy Environment: How America Treats FDI . . . . . . . . . 109 Figures Figure Page 3-1. Foreign Direct Investment in the United States: Sales of Manufacturing Affiliates . 88 3-2. Foreign Direct Investment, 1977-88: Assets of Major Investors . . . . . . 89 3-3. Foreign Direct Investment Position in the United States, 1990 . . . . . . 89 3-4. Automobile Assembly Plant Defect% 1988-89.. . . . . . . . . . 94 3-5. Automobile Assembly Productivity, 1988.. . . . . . . . . . . 94 3-6. R&D Intensity, U.S. Manufacturers and U.S. Affiliates of Foreign Manufacturing Firms . +.....,. . . . . . . . . . . . . . . . 96 3-7. R&D/Manufacturing Employee FDIUS and U.S, Manufacturers . . . . . 97 3-8. Spending for New Plant and Equipment, Foreign Manufacturing Affiiliates and U.S. Manufacturers . . . . . . . . . . . . . . . . 97 3-9. Investment in Plant, Property and Equipment as a Percent of Sales of Manufacturing Affiliates . . . . . . . . . . . . . . . . . . 98 3-10. Employment in U.S. Affiliates, 1977-88 . . . . . . . . . . . 102 3-11. Earnings and Reinvested Earnings: Foreign Direct Investors in the United States . . 104 3-12. Earnings and Reinvested Earnings: Japanese Direct Investors in the United States . 105 3-13. Earnings and Reinvested Earnings: European Direct Investors in the United States . 105 3-14. Earnings and Reinvested Earnings: U.S. Direct Investment Abroad . . . . . 106 3-15. Merchandise Trade, FDIUS Affiliates . . . . . . . . . . . 107 Tables Table Page 3-1. Foreign Direct Investment in the United States: Manufacturing Sales, 1977 and 1988 . 91 3-2. Foreign Direct Investment in the United States: Assets of Manufacturing Affiliates, 1977 and 1988 . . . . . . . . . . . . . . . . . 91 3-3. European Direct Investment in the United States Manufacturing Sales, 1977 and 1988 . 92 3-4. Japanese Direct Investment in the United States: Manufacturing Sales, 1977 and 1988.... 92 3-5. Canadian Direct Investment in the United States: Manufacturing Sales, 1977and1988 . 93 3-6. Manufacturing Research and Development Intensity: Foreign Direct Investors and U.S. Firms, 1977-88 . . . . . . . . . . . . . . . 96 3-7. Training Hours . . . . . . . . . . .. .. .. .. .. ... ..+. . 103
PAGE 90
Chapter 3 American Firms, Foreign Firms: Contributions to the Nation SUMMARY AND FINDINGS Both American and foreign companies are becoming more international, investing in a variety of activities outside their home counties. This is especially true of foreign companies, whose investment abroad rose at a much faster rate than that of U.S. firms in the 1980swith a substantial share being made in the United States. l At the same time, most advanced nations are undertaking programs to promote competitiveness. Combined with the trend toward more international investment by private companies, this puts pressure on political systems to decide how or whether multinational firms can participate in competitiveness programs. Nominally, the OECD 2 nations subscribe to the principle of national treatment, which means no discrimination against or in favor of any firm based on the nationality of its owners. Exceptions are routinely made for national security reasons to assure that nations retain sovereignty and the ability to command military production in times of national emergency. However, the distinctions between national security and economic competitiveness are becoming blurred, as military needs increasingly depend on industries whose primary business is in the civilian sector. Especially in advanced nations, decisionmakers are increasingly confronted with uncomfortable decisions on how to treat foreign firms and their affiliates. Less developed nations have long wrestled with policies towards foreign multinationals, but the issue was secondary for advanced nations until the last couple of decades. The United States is a newcomer to this realm of political decisionmaking, as it is facing rapidly rising foreign assets and control for the first time in its modem history. European nations have long had higher participation by multinationals but have not yet determined how to treat foreign affiliates, especially now that Japanese multinationals, with deep pockets, advanced technologies, and outstanding records of successful market penetration, are on the European scene. Japan, the outlier among industrialized countries in the degree to which multinationals are not participants in its economy, is being pressured by many other nations to open its markets to both imports and investment. There is some agreement among policy analysts that whatever principles govern the treatment of foreign affiliates (vis--vis their participation in programs to promote economic competitiveness), they should not be based narrowly on ownership. There is disagreement on what other principles should apply. One point of view is that national treatment should be the only principle, and that the standards for handling international investment should be the same as those that govern international trade in the GATT---i.e., openness and nondiscrimination based on nationality. Foreign affiliates, it is argued, behave very much like domestic fins, with only minor exceptions. Therefore, political intervention that treats foreign affiliates differently from domestic firms introduces distortions that decrease economic well-being for everyone. Another view is that reciprocity should be the governing principle. Reciprocity means that affiliates of foreign firms are given the same treatment in the host country as the host countrys firms are given in the nation they call headquarters. Reciprocity is already applied in a few instances in the United States, for example, in mineral leasing on public lands. A current, controversial approach focuses on performance standards. Both in Europe and the United States, there is serious talk of establishing standards that any firm must meet to qualify for government-funded or government-sponsored programs. Standards generally have to do with how much production, R&D, employment, and value added firms do in the host country, compared with domestic fins. Existing data give limited insight into how foreign affiliates behave in the United States. Foreign direct investment (FDI) in the United States is on the rise, especially in manufacturing, where foreign affiliates now account for over 10 percent of the sales of all U.S. manufacturing. These affiliates contribute in various ways to the U.S. economy and, although there are some distinctions between their behavior and that of U.S.-owned fins, they are similar in some important ways.
PAGE 91
84 l Competing Economies: America, Europe, and the Pacific Rim Foreign manufacturing affiliates are generally the equals if not the superiors of U.S. manufacturing firms in yearly investments in new plant and equipment. They do considerably less R&D, as a percentage of sales, than U.S. manufacturing fins. These two measures are the most direct aggregated data we have on how foreign affiliates contribute to U.S. technology, knowledge, and productivity. They are not adequate to make complete judgments of those contributions and will not resolve the ongoing debate over whether foreign affiliates contribute to U.S. technology on balance or make net transfers of technology and other economic benefits mostly to their home countries. In terms of employment practices, foreign affiliates are hard to distinguish from domestic companies. In manufacturing, they pay about the same compensation as domestic firms, and more than American firms overall, but this is due more to their disproportionate investment in high-wage service industries than to any propensity to pay more than comparable American establishments. Foreign affiliates are neither more nor less reliable employers than U.S. companies; affiliates are about as likely to lay off workers during economic downturns as U.S. companies. 3 Their qualitative contributions to the competence and knowledge of American employees is based on anecdotal evidence. Some foreign firms have made special contributions to American managers and workers knowledge and skills, as, for example, the New United Motor Manufacturing, Inc. (NUMMI) joint venture with Toyota did for General Motors managers, engineers, and shop-floor workers. In other cases, foreign control seems to have made little difference in the behavior or attitudes of managers or workers; and in a few cases, foreign control has been a source of strife. The most noticeable difference between foreign affiliates and U.S. firms is in their propensity to import. Firms invest abroad mainly to sell abroad; to differing degrees, that means selling products made at home. The overall trade deficit associated with foreign affiliates is sizable--a merchandise trade deficit in 1988 of $90 billion, compared with an overall U.S. merchandise trade deficit of $120 billion. Affiliates of U.S. firms in other countries have in the past generated substantial trade surpluses for the United States, but those surpluses are declining. In 1988, trade between U.S. parent companies and their foreign affiliates produced a merchandise trade surplus of about $8 billion. Japanese affiliates have by far the greatest propensity to import of any foreign affiliates, and most of what they import is made in Japan. European investors import more, per dollar of sales, than American firms or Canadian affiliates, and most of what they import is from other countries in Europe; their inclination to import from all of Europe is not as great as that of Japanese affiliates to import from Japan. It is simplistic, however, to hold imports by foreign direct investors responsible for the high U.S. trade deficits. The fundamental causes of our poor trade performance are the Nations anemic savings rate and declining overall competitiveness of its manufacturers, based on the ability to make highquality products at reasonable costs. 4 Economic theory also argues that imports in any particular sector do not affect the overall trade balance, but rather exert their effect on the value of the dollar. Macroeconomic factors-specifically, domestic savings (including government surpluses or deficits) and domestic investment-are considered the determining factors in the overall size and direction of the current account trade balance. If foreign investors imports persistently outweigh exports, and this makes for a greater U.S. trade deficit, then presumably the dollar will fall, which tends to promote U.S. exports and balance the current account. 5 This process is costly. A persistently weak dollar can enfeeble the U.S. economy and lower the standard of living. The furor over foreign direct investment (FDI) seems ironic to some, who point out that much of it is a natural response to nations dis criminatory trade policies. Firms invest abroad for many reasons; one is to continue to sell products abroad when exporting becomes difficult. Trade policy actions that limit Japanese exports are primary motivations for the heavy Japanese investments in the United States and Europe over the past decade or so. When the Japanese Government wished to protect Japanese firms from foreign competition in the postwar decades, it was obliged to limit both imports and direct investment (see ch. 6). Another complication is the growth in international strategic alliances of all types, only some of which can be classed as direct investment. Crosslicensing agreements, some joint ventures, and small equity investments do not show up in statistics on direct investment, but they do affect things that
PAGE 92
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation l 87 governments care about deeply. A current debate focuses on participation of foreign companies (or their offshore affiliates) in government-funded R&D consortia. Even when foreign firms or their affiliates are excluded, the web of technology development agreements between U.S. companies and foreign companies makes it likely that at least some of the knowledge generated in such programs will go abroad. It has never been possible for governments to control international dissemination of technology, even before firms had extensive international operations. However, with the proliferation of international activity of all types, the speed of technology diffusion is lessening and the control governments have over it are diminishing. In the United States and Europe, 6 the debate over how to treat foreign affiliates focuses heavily on Japanese affiliates. Japanese companies are the most feared, because they have reputations for manufacturing excellence and voracious appetites for technology, and also because they are perceived to behave in more nationalistic ways than firms from North America or Europe. Some anecdotal evidence bears out this perception. Besides differences in patterns and magnitude of trade, Japanese investors are more likely than other foreign investors to retain control of the operations by hiring Japanese, rather than host country, managers; more likely to equip their factories with Japanese machinery; and more likely to refer significant decisions to Japanese headquarters. These things are changing. Japanese firms in a recent MITI survey plan to give greater control and discretionary power to their foreign affiliates in the next 5 years, but the majority will still maintain control at home. The effect of Japanese management practices on the host country-good or badis not yet known. Nevertheless, differences in the behavior of Japanese investors, along with their formidable records in international competition, will continue to make Japanese direct investment a highly charged political focus. INTRODUCTION The interests of nations and firms are sometimes similar, sometimes not. Some of the things firms want-a stable business environment, productive workers, healthy profits and growth-are also attractive to governments, which is why so many governments try to improve the climate for private enterprise within their borders. But not all businesses are equally attractive to governments, and some of the things businesses want may go counter to some government interests. Nations and firms often have different stakes in the international transfer of weapons, particularly advanced weapons and the high-technology equipment needed to make them. Some enterprises contribute disproportionately to pollution and other public safety hazards; some consume large quantities of scarce natural resources. Thus, while governments at many levels may be assiduous in attracting firms from outside the country or region to locate there, they may also regulate the activities of firms in ways that discourage investment. The fit between corporate interests and governments objectives is growing more important to the United States. With many State and local governments actively courting foreign investment, and foreign firms and their attendant lobbies becoming more prominent players in American public policy, the Federal Government is more and more often forced to deal with issues of foreign ownership and control. The issue comes up in various ways. There is increasing concern over foreign investment in real estate and its effect on local real estate prices, especially with respect to Japan, and particularly in areas such as Los Angeles and Hawaii. 7 In the late 1970s and early 1980s, for example, Middle Eastern investment caused concern, and Japanese investment in American banking and finance is a current issue. National security is also a concern; in 1988 Congress authorized the President to suspend or prohibit any foreign acquisition of an American firm that is deemed to threaten defense production or the ability of domestic industries to meet national defense requirements. 8 There is a burgeoning debate on the effect of FDI on American manufacturing competitiveness. While all of these issues are important, only the latter is of concern in this study. Before going on, however, we note that different issues may call for different definitions of what FDI is, and how concerned we are over regulating it; we will address only one of the issues in this chapter, and the discussion does not apply to all other foreign-investment issues. Finally, we should consider whether multinational firms that are nominally American are so globalized that their interests and the Nations interests are not likely to have much in common. This is the Who Is Us? question. Robert Reich, who framed the question in these terms, contrasted two hypothetical companies: one with headquarters
PAGE 93
88 Competing Economies: America, Europe, and the Pacific Rim in the United States but with much of its R&D and most of its sales, assets, complex manufacturing, and jobs in foreign countries; the other, headquarters abroad but with much of its technology development and most of its jobs in the United States. 9 The question of what foreign direct investors contribute to U.S. technology, competitiveness, and jobs is complex, and the major subject of this chapter. Whether the typical U.S. multinational fits the hypothetical picture of a stateless, thoroughly globalized company is easier to answer. There may be a few such companies, and there maybe more in the future, but for now that is a false picture. U.S.-based multinationals do most of their business in the United States, and most of their jobs and technology development are here. Overall, they are identifiably American companies, and their competitive performance is linked to that of the Nation. THE BASICS: WHO INVESTS WHERE AND IN WHAT U.S. Multinationals Activities at Home and Abroad In 1988 (the latest year for which data are available), U.S. multinational companies had 78 percent of their total assets, 70 percent of their sales, and 74 percent of their employment in the United States. 10 All these percentages were higher than in 1977 (the first year in this data series). There was no consistent trend in the 1980s toward more footloose operations by U.S. multinationals, slightly to the contrary in fact. Although direct investment abroad by U.S. multinationals increased during those years, so did their investments at home. The data also indicate that U.S. multinationals are keeping good jobs and technology development at home. In 1988, compensation per employee in affiliated companies abroad was about 72 percent of that for employees of the parent company in the United States. Assets per employee in the affiliates were 77 percent of the figure for parent companies, which implies that the more productive jobs resided in the United States. ll The same applies to R&D. The latest government figures for the location of R&D spending by U.S. multinationals date from 1982. 12 At that time, spending for R&D by foreign affiliates was under 9 percent of the total for parents and affiliates; this compared with affiliates share of total sales, which in 1982 was 33 percent. R&D as Figure 3-1Foreign Direct Investment in the United States: Sales of Manufacturing Affiliates Billions of dollars 300 1 250 200 < 150 100 50 b T T 1977 78 79 80 81 82 83 84 85 86 87 88 Total FDIU S Europ e Japan SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table E-5, and subsequent series. a percent of sales was nearly three times higher among U.S. parent manufacturing companies than among manufacturing affiliates abroad (3.33 percent v. 1.15 percent). Some U.S. multinationals have important R&D facilities in other countries. For example, American automobile companies develop and sell unique products in Europe. It was two scientists in IBMs Zurich research laboratory who discovered hightemperature superconductivity in 1986; even so, IBM does 80 percent of its R&Din the United States, about 12 percent in Europe, and 8 percent in Asia. 13 The overall picture may change. The European Community (EC) proposes to allow companies that have filly integrated operations, that include R&D, manufacturing, and sales within Europe, to participate in EC-funded R&D programs. This may have the effect of shifting more of U.S. multinationals R&D, or more of their high value-added jobs, to Europe. As of now, however, much the greater part of these activities take place in the United States. FDI in the United States Direct investment in the United States rose from 16 percent of total world direct investment in 1980 to 25 percent in 1987, while the shares of Europe, Canada, Australia, and South Africa decreased. 14 In 1990, foreign fins total direct investments in the United States amounted to $404 billion, compared with direct investments of $421 billion by U.S. multinationals in foreign countries. The gross prod-
PAGE 94
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation l 89 Figure 3-2Foreign Direct Investment, 1977-88: Assets of Major Investors Billions of dollars 300 25 0 20 0 150 I J 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Franc e m Germany ~ Netherlands = U.K. = Japan ~ Canada SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table B-7, and subsequent series. uct of foreign affiliates in the United States accounted for 3.4 percent of GNP in 1987, up from 1.8 percent in 1977. 15 The rapidly increasing Japanese direct investment in manufacturing plants within the United States has made most of the news. There are three common measures of the importance of direct investment: sales, assets, and position. l6 Japanese affiliates share of the sales of all foreign manufacturing affiliates in the United States rose from 4 percent in 1977 to 9 percent in 1988 (figure 3-l). The Japanese share of the total sales of all foreign affiliates was 26 percent, the same at the beginning and the end of the period. But their investments in manufacturing during the period show up in their share of affiliates assets, which rose from 12 to 24 percent of the total. By the late 1980s, the assets of U.S. affiliates of Japanese direct investors stood at $275 billion, surpassing all the rest, including the United Kingdom, historically the largest foreign direct investor in the United States (figure 3-2). The total value of assets of U.K. affiliates stood at $194 billion in 1988. 17 However, the United Kingdom was still by far the leader in direct investment position, with investments valued at $108 billion by the end of 1990 (figure 3-3). Japan was second, having passed the Netherlands in 1988; its direct investment Figure 3-3-Foreign Direct Investment Position in the United States, 1990 Billions of dollars 140 120 100 80 60 40 20 0 108.1 83.5 64.3 27.7 27.8 w Canada France Germany Japan Netherlands U.K. SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Russell B. Scholl, The International Investment Position of the United States in 1990, Survey of Current Business, vol. 71, No. 6, June 1991, table 7, p. 32. position in the United States amounted to $70 billion in 1989. Foreign firms participation in U.S. manufacturing is greater than their overall participation in other sectors of the economy. Foreign affiliates accounted for 12.2 percent of the assets of U.S. manufacturing in 1987, compared with 8.9 percent of the total net worth of all nonfinancial corporations. 18 Manufac292-889 0 91 4 QL:3
PAGE 95
90 Competing Economies: America, Europe, and the Pacific Rim turing jobs were still more skewed to foreign affiliates, which accounted for 9.0 percent of U.S. manufacturing employment in 1988 but only 3.4 percent of all U.S. civilian employment. 19 The picture is different for Japanese direct investment. About 15 percent of the sales of Japanese affiliates in the United States is in manufacturing, compared to 38 percent of the sales of European affiliates. 20 Over 60 percent of Japanese investors sales comes from wholesaling affiliates, 21 a far higher percentage than for any other major direct investor. This implies that the Japanese interest in the American market is primarily in selling goods made in Japan. Of course, a main reason for any firm, from any nation, to invest in a foreign country is to sell more goods in that market. However, the heavy emphasis by Japanese investors on wholesaling suggests that Japanese fins, compared with those of other nations, are more interested in exporting and less interested in producing goods in the country where the goods are sold. Trade figures for affiliates support this observation. Japanese affiliates imports were significantly higher than those of European or Canadian affiliates throughout the period for which we have data (1977-88). In 1988, imports of Japanese affiliates in the United States were $75.9 billion, accounting for 51 percent of the imports of all affiliates of foreign investors, and the Japanese affiliates imports amounted to 34 percent of their sales, compared to 12 percent for European affiliates. 22 Moreover, a detailed survey in 1987 showed that 93 percent of the imports of Japanese affiliates were from Japan. To be sure, all affiliates import mostly from the home country (or in the case of Europe, the home region); 70 percent of the imports of European affiliates were from Europe, and 73 percent of the imports of Canadian affiliates were from Canada. 23 But the Japanese affiliates have by far the highest ratio of homecountry imports of all, as well as the highest imports relative to sales. Direct investment in U.S. manufacturing between 1977 and 1988 shows annual increases of 16 percent in sales and almost 19 percent in assets (tables 3-1 and 3-2). The most rapid growth in manufacturing sales by foreign affiliates was in transportation equipment, where affiliates sales increased at the rapid clip of 46.5 percent per year. Most of this was just where one would expect: in sales of motor vehicles from Japanese affiliates. If sales of motor vehicles from wholesaling affiliates are added in, the influence of Japanese affiliates is even more apparent. In 1988, the combined total of wholesale and manufacturing sales of motor vehicles, by all foreign affiliates, amounted to $80.9 billion. Two-thirds of this ($52.9 billion) was from Japanese affiliates, and most of it ($44.3 billion) was sales from wholesaling, not manufacturing, establishments. 24 Other industries with substantial manufacturing sales by foreign affiliates include chemicals and machinery (including electronic equipment); European affiliates are preeminent in both of these major sectors (table 3-3). About 29 percent of the sales of European affiliates is in chemicals and allied products, with Germany the leading foreign affiliate in the sector and the United Kingdom not far behind. Several European countries have large sales in machinery, a category that includes machine tools and various types of production equipment used in nearly every other industry as well as semiconductors, computers, and consumer electronic goods. The United Kingdom is a leader in nonelectrical machinery, while France, Germany, and the Netherlands all have important affiliates in the United States making electronic products (Thomson, Siemens, and Philips, respectively) .25 Although there are also Japanese affiliates making or assembling electronic products in this country, most of Americas huge purchases of Japanese electronic goods, from semiconductors to compact disk players, are imports. Japanese manufacturing affiliates sales are concentrated in primary and fabricated metals (steel), electrical and electronic equipment, and transportation equipment (table 3-4). Canada and the United States have long had substantial investment in each others markets as a result of shared language, proximity, and similar culture and business environments. Canadian direct investment in the United States is heaviest in chemicals, 26 followed by primary metals and electrical and electronic equipment (table 3-5). FDI AND U.S. MANUFACTURING COMPETITIVENESS The relationship between FDI and U.S. manufacturing competitiveness is anything but straightforward. In some cases, foreign investment seems to have stimulated American manufacturers to improve
PAGE 96
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 91 Table 3-lForeign Direct Investment in the United States: Manufacturing Sales, 1977 and 1988 Annual average Sales Sales growth rate 1977 1988 1977-88 Industry ($ millions) ($ millions) (percent) Manufacturing . . . . . . . . $50,489 $258,511 16.0% Food and kindred products . . . . . 6,983 32,995 15.2 Chemicals and allied products . . . . 16,303 63,245 13.1 Primary and fabricated metals . . . . 6,881 32,806 15.3 Primary metal industries . . . . . 5,545 20,476 12.6 Fabricated metal products . . . . . 1,336 12,330 22.4 Machinery . . . . . . . . . 9,838 45,933 15.0 Machinery, excluding electrical . . . 4,512 17,905 13.3 Electrical and electronic equipment. . . 5,326 28,029 16.3 Other manufacturing: Textiles and apparel . . . . . . 1,072 3,746 12.0 Paper and allied products . . . . . 1,803 8,033 14.5 Printing and publishing . . . . . 1,741 12,386 19.5 Rubber and plastics products . . . . 916 11,295 25.7 Stone, clay and glass products . . . 2,022 12,363 17.9 Transportation equipment . . . . . 279 18,649 46.5 NOTE: Individual industries do not add to total manufacturing. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table E-5, and Foreign Direct lnvestment i n the United States, Operations of U.S. Affiliates of Foreign Companies, Preliminary 1988 Estimates, 1990, table E-7. Table 3-2Foreign Direct Investment in United States: Assets of Manufacturing Affiliates, 1977 and 1988 Assets 1977 Industry ($ millions) Manufacturing . . . . . . . . $41,759 Food and kindred products . . . . . 4,373 Chemicals and allied products . . . . 15,258 Primary and fabricated metals . . . . 5,931 Primary metal industries . . . . . 4,670 Fabricated metal products . . . . . 1,261 Machinery . . . . . . . . . 7,508 Machinery, excluding electrical . . . 3,754 Electrical and electronic equipment . . 3,754 Other manufacturing Textiles and apparel . . . . . . 726 Paper and allied products . . . . . 1,416 Printing and publishing . . . . . 1,361 Rubber and plastics products . . . . 606 Stone, clay and glass products . . . 1,736 Transportation equipment . . . . . 587 Annual average Assets growth rate 1988 1977-88 ($ millions) (percent) $281,316 30,317 80,911 34,018 17,495 16,523 45,857 20,507 25,351 18.9% 19.2 16.4 17.2 12.8 26.4 17.9 16.7 19.0 4,132 17.1 7,015 15.7 15,075 24.4 10,164 29.2 21,113 25.5 9.666 29.0 NOTE: Individual industries do not add to total manufacturing. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U. S. Affiliates, 1977-80, 1985, tabie B-7, and Foreign Direct lnvestment in the United States, Operations of U.S. Affiliates of Foreign Companies, 1988, 1990, table B-5. their competitive performance. An example comes from 1981 to 1985, and probably also to neutralize from motor vehicles. the effects of other forms of protection that might be imposed in the future. Honda was the pioneer. 27 Japanese automakers began assembling motor With several years of experience making motorvehicles in the United States in the 1980s, partly to cycles in the United States, it built the first Japanesebypass the voluntary restraint agreement that limited owned assembly plant in Marysville, Ohio, and Japanese motor vehicle exports to the United States began producing cars there in 1982. Hondas entry
PAGE 97
92 l Competing Economies: America, Europe, and the Pacific Rim Table 3-3-European Direct Investment in the United States: Manufacturing Sales, 1977 and 1988 Sales Sales Percent of 1977 1988 total Industry ($ millions) ($ millions) in 1988 Manufacturing . . . . . . . . $36,754 $166,608 Food and kindred products . . . . . 4,594 25,547 15.3% Chemicals and allied products . . . . 15,330 47,421 ,28.5 Primary and fabricated metals . . . . 3,694 14,148 8.5 Primary metal industries . . . . . 2,886 6,187 3.7 Fabricated metal products . . . . . 808 7,961 4.8 Machinery . . . . . . . . . 6,701 28,385 17,0 Machinery, excluding electrical . . . 3,006 11,839 7.1 Electrical and electronic eqp. . . . . 3,695 17,546 10.5 Other manufacturing Textiles and apparel . . . . . . 780 2,303 1.4 Paper and allied products . . . . . NA a 5,156 3.1 Printing and publishing . . . . . 481 6,552 3.9 Rubber and plastics products . . . . 772 5,112 3.1 Stone, clay and glass products . . . 1,878 9,964 6.0 Transportation equipment. . . . . NA 8,923 5.4 a Data suppressed to avoid disclosure of information for individual companies. NOTE: Individual industries do not add to total manufacturing. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80,1985, table E-5, and Foreign Direct lnvestment in the United States, Operations of U.S. Affiliates of Foreign Companies, 1988, table E-8. Table 3-4-Japanese Direct Investment in the United States: Manufacturing Sales, 1977 and 1988 Percent of Sales Sales total Industry 1977 1988 in 1988 Manufacturing . . . . . . . . $2,255 $33,180 Food and kindred products . . . . . 275 1,055 3.2% Chemicals and allied products . . . . 121 2,060 6.2 Primary and fabricated metals . . . . 654 5,390 16.2 Primary metal industries . . . . . NA a 3,716 11.2 Fabricated metal products . . . . . NA 1,675 5.0 Machinery . . . . . . . . . 497 8,992 27.1 Machinery, excluding electrical . . . 317 3,276 9.9 Electrical and electronic equipment . . 180 5,716 17.2 Other manufacturing Textiles and apparel . . . . . . 67 346 1.0 Paper and allied products . . . . . NA 635 1.9 Printing and publishing . . . . . NA 1,094 3.3 Rubber and plastics products . . . . 17 2,842 8.6 Stone, clay and glass products . . . NA 1,043 3.1 Transportation equipment. . . . . NA 8,584 25.9 a Data suppressed to avoid disclosure of information about particular companies. NOTE: Individual industries do not add to total manufacturing. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U. S. Affiliates, 1977-80, 1985, table E-5, and Foreign Direct lnvestment in the United States, Operations of U.S. Affiliates of Foreign Companies, 1988, 1990, table E-8. into U.S. motor vehicle production was followed by In the 1970s and early 1980s, the prevailing view Nissan, Mazda, Subaru-Isuzu, Toyota, and Mitsuamong American automakers and their suppliers bishi (in a joint venture with Chrysler), By 1991, the was that the Japanese advantage in the American Japanese transplants (including the joint ventures) market stemmed mainly from low labor costs. It took are expected to be able to produce nearly 3 million firsthand demonstrations of Japanese manufacturing vehicles in North America. 28 prowess in America to convince them that the real
PAGE 98
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 93 Table 3-5-Canadian Direct Investment in the United States: Manufacturing Sales, 1977 and 1988 Sales Sales Percent of 1977 1988 total Industry ($ millions) ($ millions) in 1988 Manufacturing . . . . . . . . $11,650 $38,307 Food and kindred products . . . . . 1,972 3,740 9.80/o Chemicals and allied products . . . . 649 11,902 31.1 Primary and fabricated metals . . . . 2,068 5,842 15.3 Primary metal industries . . . . . 1,747 4,644 12.1 Fabricated metal products . . . . . 321 1,198 3.1 Machinery . . . . . . . . . 3,108 5,248 13.7 Machinery, exe. electrical . . . . . 1,489 1,343 3.5 Electrical and electronic equipment. . . 1,619 3,905 10.2 Other manufacturing Textiles and apparel . . . . . . NA a 668 1.7 Paper and allied products . . . . . NA 1,990 5.2 Printing and publishing . . . . . 1,171 3,463 9.0 Rubber and plastics products . . . . 125 2,024 b 5.3 Stone, clay and glass products . . . 815 690 1.8 Transportation equipment . . . . . NA 490 1.3 a Data supressed to avoid disclosure of information for individual companies. b lncludes only information for plastics; data for rubber products was suppressed to avoid disclosure of information for an individual company. NOTE: Individual industries do not add to total manufacturing. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Invesfrnenf in the United Sfates:Operations of U.S. Affiliates, 1977-80,1985, table E-5, and Foreign Direct lnvestment in the United states, Operations of U.S. Affiliates of Foreign Companies, 1988, 1990, table E-8. advantage of the Japanese manufacturers was their rigorous and ceaseless attention to quality and efficiency in manufacturing. The Japanese transplants have the best quality record (in terms of defects per 100 vehicles) and average productivity (in hours per auto) of any plants in North America, and the worst Japanese transplants have about the same productivity as the average for U.S. plants in North America (figures 3-4 and 3-5). 29 For American auto parts makers, the lesson was more than just a demonstration. The exacting standards of quality, price, and delivery time that the Japanese auto assemblers held suppliers to in Japan were, according to the Japanese fins, beyond what most American auto parts and components makers were accustomed to providing to Detroit. In addition, business practices of the American and Japanese auto assemblers were quite different: Japanese assemblers had (and have) many fewer suppliers than the American assemblers, and those suppliers are expected to deliver whole assemblies instead of individual parts. Japanese assemblers also expected collaboration in initial design and quick turnaround on design changes, which required in-house engineering ability that American components makers were unaccustomed to providing and often did not have. 30 Few were able to establish relations with Japanese assemblers in North America. By 1987, researchers from the MIT International Motor Vehicle Program estimated that the local parts content of the Japanese assemblers in North America was only 30 percent; this was forecast to increase to 50 percent by 1990. 31 American suppliers who were able to negotiate arrangements with Japanese transplants report difficulties in establishing the relationship, but those who succeeded also made positive changes. These include improving product quality and inventory management, increasing productivity, and expanding engineering, design, and R&D. 32 Many of the same kinds of changes are increasingly required by Ford, General Motors, and Chrysler in their own attempts to compete with Japanese imports and transplants. FDI can also enhance technology development. For example, American manufacturers have benefited from the patient capital or technology-oriented strategies of their foreign investors. When the West German chemical firm Hoechst purchased Celanese Corp., Hoechsts objective was to find a technologyintensive strategy for competing in the U.S. market. 33 Rather than expand its own U.S. operation, Hoechst purchased Celanese, an existing American chemical company with well-recognized products, competent R&D, and established customer relation-
PAGE 99
94 l Competing Economies: America, Europe, and the Pacific Rim Figure 3-4-Automobile Assembly Plant Defects, 1988-89 Hours/vehicle 100 8 0 60 4 0 20 0 50.9 49.0 Figure 3-5Automobile Assembly Productivity, 1988 Hours/vehicle 30r 25 20 1 16.8 78.5 15 10 5 n I Japan/Japan Japan/U.S. U.S./U.S. Japan/Japan = Japanese-owned Plants located in Japan Japan/U.S. = Japanese-owned plants and joint ventures in the United States U.S./U.S. = U.S.-owned plants in the United States SOURCE: John F. Krafcik, Training and the Automobile Industry: international Comparisons, contractor report prepared for the Office of Technology Assessment, February 1990, pp. 8-9. ships. Despite initial difficulties in reconciling the corporate R&D cultures of the two firms, Celaneses R&D spending increased by 10 percent annually after the acquisition. Moreover, Hoechst was more willing to engage in long-term research than Celaneses management had been, and less reluctant to make major commitments to projects with uncertain and distant payoff. While there are many examples of foreign investment that seem beneficial or at worst neutral in their impact on manufacturing competitiveness, there are worries as well. Most of the worry centers on Japan and several high-technology industrial sectors. In its simplest form, the fear is that Japanese investors, with their appetite for new technology, their deep pockets, and their perceived preferences for doing business with other Japanese companies, invest in American high-technology companies in order to gain access to new technologies, but that most of the benefits of such investment (jobs, economic growth, contributions to the national stock of technology) will end up in Japan. Another worry is that, since Japanese corporations investing abroad have commonly been followed by their Japanese suppliers, American businesses that could benefit from relationships with Japanese multinationals might be crowded out. Such accusations surfaced recently, when two former executives of Ardent Computer sued Kubota 20.5 25.0 Japan/Japan Japan/U.S. U.S.U.S. Japan/Japan = Japanese-owned plants located in Japan Japan/U.S. = Japanese-owned plants and joint ventures in the United States U.S./U.S. = U.S.-owned plants in the United States SOURCE: John F. Krafcik, Training and the Automobile industry: international Comparisons, contractor report prepared for the Office of Technology Assessment, February 1990, pp. 8-9. Ltd., a Japanese tractor company that had acquired a stake in Ardent in 1986. The two executives allege that Kubota forced Ardent to merge with Stellar Computer, and conspired to transfer the technology of the merged company (Stardent) to a Kubota subsidiary, Kubota Computers America. 34 Kubota denies the claims, saying instead that the U.S. executives were failed managers who demanded money and sued after they were denied payment. Fear of Japanese dominance also colored American reaction to the investment by some of Japans big electronics companies in R&D centers in the United States. 35 The companies are staffing these centers by hiring leading American computer scientists from American universities and corporate labs, with offers of high salaries, excellent equipment, and plenty of R&D money. Often, R&D investments within this country are used as a measure of the positive value of FDI. But some analysts see the hiring of Americas best computer brains by powerful Japanese companies as threatening one of the last, best competitive advantages of U.S. computer companies-basic research in computer science. These cases are part of the fear that some people have about Japanese firms. The danger, according to those who suspect Japanese investors of playing by nationalistic rules, is that American companies will end up depending on Japanese companies almost
PAGE 100
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 95 exclusively for key components and equipment, 36 and even that the next big creative advances in technology may be locked up in patents held by Japanese companies. This dependence, they fear, will be the downfall of the American companies. What does this have to do with FDI? According to this analysis, Japanese investment is simply another way (along with exports) for Japanese companies to dominate their industries. Such fears are made more plausible by the fact that most Japanese businesses large enough to deal internationally operate in keiretsu, or groups that hold each others stock and give preference to other group members in procuring supplies and services in ways that tend to exclude outsiders. Fear of Japanese dominance may be rooted in xenophobia, jealousy, experience, common sense, or some combination of all of these. Many people familiar with Japanese business practices regard Japanese investors as more likely than Europeans or Canadians to direct the benefits of their investments in America to Japan. Evidence on either side of this debate is sparse. What little there is does suggest that there are good reasons to keep an eye on foreign investment in general, and Japanese investment in particular, and monitor the effects on American competitiveness. What should we keep track of? More and more, developed nations are grappling with these issues, particularly as the strains on the postwar trade regime intensify. Most focus on the readily available measures of fins behaviorR&D, employment, worker compensation, value added, and the like. A common problem is that many of these measures gauge inputs, rather than the outputs that nations are interested in. It is easy to argue about the effects of FDI without proper measures of what foreign investors contribute to the nations stock of knowledge and overall well-being, since many of the arguments on all sides depend on unsatisfactory measures and anecdotes. In the United States, the reevaluation of policies toward international investment is a result of the tremendous growth in FDI during the past two decades. 37 In large part, the European Community seems committed to making the benefits of 1992 reforms open to anyone, but there are some important questions and exceptions. In a few critical industries, principally microelectronics and motor vehicles, it appears that firms wishing to sell their products in Europe under the same conditions as European firms will need to make substantial portions of the products in Europe. As this report is written, for example, European negotiators are discussing domestic content regulations ranging up to 85 percent on motor vehicles, and a transition period of 5 years, starting in 1993, during which Japanese automakers agree to limit exports to the European Community countries. 38 In electronics, the European Community has decided that the most significant part of semiconductor manufacture, diffusion in wafer fabrication, must be done in Europe for semiconductors to count as of local origin (and thus not be subject to the ECs 15 percent tariff). Some expect this decision to result in a boom in wafer fabrication in Europe, primarily on the part of American and Japanese companies. 39 Clearly, one thing the European Community is interested in is jobshence the emphasis on local content in big-ticket trade items like semiconductors and automobiles, which account for a large share of Europes imports. There is more to it than jobs though. Countries are concerned about the extent to which foreign investors are players in the political process, add to a nations stock of knowledge and technology, contribute to imports and exports, pay taxes, and enhance human resource development. All of these are things that can be expected to contribute to a nations well-being, and to the competitiveness of its fins. Contributions to Knowledge and Technology Firms add to nations technical and scientific knowledge in many ways, most of which are difficult to measure. One measure that is available is the amount spent on R&D. 40 R&D spending is often used as a proxy for all contributions foreign firms with domestic operations make to the nations technical knowledge. Spending on plant and equipment and reinvestment of earnings in domestic operations are also sometimes used to indicate levels of contributions to technology. It is important to keep in mind, however, that such figures are only proxies and may obscure the complexities of what goes on in the real world. The R&D Measure R&D spending is usually measured as a percentage of net sales, known as R&D intensity. Figure 3-6 and table 3-6 show that the manufacturing affiliates of foreign direct investors lag behind U.S. manufac-
PAGE 101
96 l Competing Economies: America, Europe, and the Pacific Rim Figure 3-6-R&D Intensity, U.S. Manufacturers and U.S. Affiliates of Foreign Manufacturing Firms 4 Percent c 3 1 2 [ 1 0 I 1 1 1 1 1 I 1 I 1977 78 79 80 81 82 83 84 85 86 87 88 ~ U.S. manufacturing ~ Total FDIUS SOURCE: Economic Report of the President, 1990 (Washington, DC: U.S. Government Printing Office, February 1990), table C-90; National Science Foundation, Division of Science Resources, unpublished data; and U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table H-3, and subsequent series. turers in R&D intensity .41 This means that domestic firms perform a greater share of their R&D here than foreign firms operating here do, which is just what might be expected. As noted, U.S. multinational fins R&D intensity is much higher in the United States than it is in their various foreign outposts. Data showing R&D of foreign affiliates by country and industry exist for only one year, 1987. In that year, R&D (by manufacturing fins) as a percent of manufacturing sales of European affiliates was 2.3 percent, close to the average for all foreign affiliates. However, the R&D intensity for Japanese manufacturing affiliates was much below average, 1.1 percent. 42 It is likely that Japanese manufacturing multinationals, like U.S. multinationals, were doing most of their R&D at home. There is good reason for R&D intensity to be higher in the home country of multinational companies than in their foreign affiliates. R&D comprises a variety of activities, some of which are not particular to specific markets. For example, much basic and applied research done by firms is not dependent on the eccentricities of different markets, and can probably be done most efficiently at a central R&D facility. Development and design work, on the other hand, might need to be apportioned to each of a companys major markets in order to be tailored to the tastes and specifications of local Table 3-6-Manufacturing Research and Development Intensity: Foreign Direct Investors and U.S. Firms, 1977-88 (In percent) Total U.S. manufacturers Company funded All FDIUS a Year (in percent) (in percent) (in percent) 1977 . . 2.2% 1978 . . 2.1 1979 . . 2.1 1980 . . 2.2 1981 . . 2.3 1982 . . 2.7 1983 . . 2.9 1984 . . 3.0 1985 . . 3.3 1986 . . 3.6 1987 . . 3.6 1988 . . 3.5 NA NA 1.470 1.5 1.6 1.9 2.0 2.1 2.3 2.5 2.4 2.2 1.570 1.5 1.3 1.9 1.9 2.2 2.2 2.3 2.4 2.5 2.5 2.4 a Manufacturing R&D as a percent of sales by manufacturing affiliate. NOTE: Total R&D intensity includes ail funding for industrial R&D supplied by companies, the Federal Government and other sources; whereas company-funded R&D includes all funded industrial R&D work performed within company facilities from all sources except the Federal Government. SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table H-3, and subsequent series; and National Science Foundation, Division of Science Resources Studies, unpublished data. consumers. For example, all the major Japanese auto companies do design work (mainly styling) in the United States. Another way of gauging R&D contributions is in terms of spending per employee. 43 By this measure, foreign direct investors appear quite similar to U.S. firms (figure 3-7). There are complications in comparing R&D figures across nations, since the U.S. Government and foreign governments contribute different amounts for different purposes (e.g., defense v. civilian) to industrial R&D. All in all, however, there appears to be rough parity in R&D spending per employee between U.S. firms and foreign direct investors. Why should this be so, when R&D intensity is substantially higher for American fins? Interpretation is rather risky, since detailed knowledge of the underlying factors is lacking. However, it seems likely that affiliates or subsidiaries of foreign fins, no matter how firmly entrenched in countries outside of headquarters, are not as fully integrated to include all line and staff functions of the company as is headquarters. The affiliates may have a full production and sales staff, but are less likely to include functions such as accounting, finance, strategy, and plannin g in the foreign location.
PAGE 102
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 97 Figure 3-7-R&D/Manufacturing Employee FDIUS and U.S. Manufacturers Thousands of dollars 5 c 4 3 2 ?6 1 1 1 1 1977 78 79 80 81 82 83 84 85 86 87 88 1 1 `I 1 1 1 I ~ U.S. manufacturing ~ Foreign mfg. affiliates SOURCE: U.S. Department of Labor, Bureau of Labor Statistics, Employment and Earnings, November 1988, table B-1; National Science Foundation, Division of Science Resources, unpublished data; and U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1905, tables F-1 and H-3, and subsequent series. Investment in Plant and Equipment Besides R&D, another possible measure of a firms contribution to the nations technological proficiency is its investment in plant and equipment. While plant and equipment are not all there is to technology, they do embody and contribute to technology. Advanced equipment and welldesigned plants, together with well-trained workers, can make a significant contribution to productivity and product quality. Investment in plant and equipment may make a less direct or certain contribution to technology than R&D. Investment in R&D is sometimes embodied in products or patents that can be widely diffused, while investments in plant and equipment may raise the productivity of only a few plants, and the lessons learned from such investments are difficult to transfer. Nonetheless, modern plant and equipment and intelligent use of workers and machinery do improve productivity and familiarity with modern methods of production: therefore plant and equipment investment is commonly associated with improved productivity and advancing technology. For the period 1977-88, foreign manufacturing affiliates have been at least the equal of U.S. manufacturing firms in their yearly spending for new plant and equipment (figure 3-8). They far surpassed the American firms in the late 1970s and early Figure3-8-Swndin~ for New Plant and Equipment, Foreign-Manufacturing Affiliates and U.S. Manufacturers Percent of manufacturing sales 10 ~ 8 1A nl 1 I 1 1 1 1 I 1 1 1 1977 78 79 80 81 82 83 84 85 86 87 88 ~ U.S. firms ~ Foreign affiliates SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985,. tables D-29 and E-1, and subsequent series; Economic Report of the President r 1990 (Washington, DC: U.S. Government Printing Office, February 1990), table C-90; and U.S. Department of Commerce, Bureau of the Census, Plant and Equipment Expenditures and Plans, Unit@ States Department of Commerce News, Sept. 13,1990, table 4. 1980s, fell behind for a few years in the mid-1980s, and more recently pulled ahead. Figures are available for 3 years (1980, 1987, and 1988) on property, plant, and equipment investment as a percentage of sales for manufacturing affiliates by country of the parent company. 44 These data show that the Japanese far outstripped their European counterparts, and foreign manufacturing affiliates in general, in such investments per dollar of sales (figure 3-9). One likely reason is the Japanese propensity to invest directly in new plants rather than to acquire or buy a share in existing ones, as the European investors are more prone to do. 45 Another is that the concentration of investment in capitalintensive industries--disproportionate investments in motor vehicles and electronics-means higher investment/sales ratios. Japanese FDI: A Special Case Japanese direct investment in the United States differs from that of other countries in more than one way. As noted, the very high proportion of sales from wholesale affiliates suggests that a dominant interest in Japanese investment is to sell goods made in Japan. This accords with the responses of firms surveyed by MITI in a recent survey. 46 Over 80 percent of respondents indicated that their motive
PAGE 103
98 l Competing Economies: America, Europe, and the Pacific Rim Figure 3-9investment in Plant, Property and Equipment as a Percent of Sales of Manufacturing Affiliates Percent 14 12 10 8 6 4 2 0 1980 1987 1988 = All foreign affiliates = European affiliates m Japanese affiliates SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, 1980, October 1983, tables D-9 and E-1, and subsequent series. for investing offshore was to increase sales in local markets. This motive ranked highest on the list of all reasons in all markets. 47 Other motives also strongly suggest that the main reason for investing was to increase sales; nearly 50 percent of respondents said they invested in the United States because of U.S. trade regulations (implying that they were substituting local production for exports); about 25 percent invested in the United States to obtain a good partner in the local market. 48 Another main interest (greater than that evident in other countries investments) seems to be gaining 49 For example, access to advanced technologies. contrary to the general Japanese characteristic of majority ownership in foreign affiliates, minority Japanese investments in small American hightechnology firms almost doubled from 1988 to 1989, rising from $176 million to $320 million. This can be seen as an effort by Japanese companies to gain access to U.S. technology and diversify into new businesses. It does not necessarily imply that Japanese firms are somehow siphoning technology from the United States, for the U.S. firms get something in return. Many find that FDI is a good entree to foreign markets, and a large Japanese partner may have distribution channels that small firms could not duplicate. 50 The advantages of a tie-up between a small American company and a big Japanese firm maybe illustrated by two recent agreements signed between Nicolet Instruments Corp. and Matsushita Communication Industrial Co. Ltd. The agreements grant Matsushita the right to sell a line of Nicolets oscilloscopes in Japan under Matsushitas name (something Matsushita doubtless has an advantage in doing), and provide for cooperation and exchange of technical information between the two companies in the development of electronic measurement products. According to the CEO of Nicolet, the alliance greatly enhances Nicolets market penetration for its new oscilloscope in Japan, while the technical exchanges between the two companies will allow them to develop a new line of instruments to be marketed worldwide. 51 While the agreement between Nicolet and Matsushita may or may not involve direct investment, the mechanics are familiar; many Japanese investments in American hightechnology companies involve similar arrangements. 52 Moreover, the different interests of the Japanese investors may prove a boon to American hightechnology companies as well. According to one source, Japanese firms are interested in one thing: long-term gain, which translates to a strong interest in R&D. Typically, Japanese middle managers have the authority to commit significant funds to R&D or joint development programs. U.S. investors in hightechnology firms-venture capitalists-want to cash in their gains in a relatively short time (in 7 or fewer years), a constraint the Japanese investors to not impose. 53 While many American companies find alliances with (and investment from) Japanese companies beneficial, there is still worry about the long-term consequences. For -example, many observers worry that, in alliances with large Japanese fins, small American companies may end up losing control of their technologies and products. Another concern is that Japanese investment in high-technology electronics firms and their suppliers will result in U.S. semiconductor and systems makers being overly dependent on Japanese firms for critical components. That dependence, in turn, could be used as a competitive or political weapon. For example, the recent decision of the Administration to allow the Japanese company Nippon Sanso to purchase the American firm Semi-Gas Systems, a supplier of high-quality gas equipment to semiconductor makers (and a participant in Sematech), could make the U.S. semiconductor industry vulnerable in several ways. The purchase will mean that Japanese companies will control over 40 percent of the world
PAGE 104
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation l 99 market and nearly half the U.S. market for high54 It fight, for instance, mean quality gas equipment. that Nippon Sanso could cut off supplies to Sematech or its member companies in response to trade policy decisions like the 1989 naming of Japan as an unfair trader under the Super 301 section of U.S. trade law. Or the pressure could be more in the realm of business operations than politics: Nippon Sanso could selectively favor its Japanese customers with new products, low prices, and quick deliveries. 55 Allegations of such discriminatory practices on the part of Japanese companies have surfaced in the past, but are difficult to pin down. 56 The situation in semiconductor manufacturing equipment as a whole is not so stark as that in high-quality gas equipment. According to the International Trade Administration of the Department of Commerce, Japanese fins investments accounted for 10 percent of the market in 1983, and had increased to 15 percent in 1988. VLSI Research estimated that the share would reach 26 percent in 1993. 57 These investments are a significant but not yet overwhelming share of the U.S. market, but combined with imports of semiconductor manufacturing equipment from Japan, they contribute to fears of Japanese dominance. Whether allegations of discriminatory practices are true or not, dependency is always worrisome, particularly in technologies or industries considered critical to a nations well-being. This category includes products for maintaining g national security and agricultural commodities in most advanced nations. Increasingly, dependence on foreign companies or countries for high-technology products has become a source of anxiety. The problem becomes acute when foreign competitors control the most advanced equipment, materials, and supplies needed to produce something as vital as semiconductor chips, especially when, as in the case of Japanese electronics companies, a few large firms are in a position to wield oligopolistic market power. Because of the particular nature of Japanese business, many fear that this control could be more costly than possible temporary disruptions in supplies or control of prices, in the end costing the United States the ability to produce advanced components and products at any price. There is also fear that Japanese investment could mean increased economic and political influence over American business and government. This concern is typical of nearly every country facing heavy FDI. Developing nations have long maintained strict controls over foreign fins, and sometimes nationalized them, to avoid foreign economic dominance or excessive interference in domestic affairs. In this respect too, however, many nations worry that Japanese corporate investment is somehow different from investment by firms of other nationalities. Another worry is that Japanese investment is shaped by the aims and goals of headquarters and is therefore unresponsive to local concerns. Even in developed nations, Japanese firms are unwilling to relinquish headquarters control of local operations. A symptom of this is the sparse representation of native managers in Japanese affiliates abroad. A recent report from the American Electronics Association showed that, while 71 percent of American electronics firms in Japan have Japanese CEOs, only 2 percent of Japanese electronics firms in the United States have American CEO S 58 Another report is that Japanese companies in America have difficulty recruiting qualified American management, because many U.S. executives believe that Japanese companies will keep non-Japanese staff out of important decisions, 59 or force them to check with the Japanese headquarters for all decisions of consequence. This phenomenon is not limited to the United States. A study of 62 multinational companies doing business in Australia (42 American or European and 20 Japanese) showed the same pattern. 60 Most American or European operations in Australia were managed by Australians, but only one of the Japanese operations was wholly Australian-run, and even when Australian managers were used, their discretionary power was curtailed by Japanese advisors. 61 Some Americans working for Japanese companies complain that they have few opportunities for advancement, and fear that Japanese companies are more likely than companies of other nationalities to keep high-paying jobs at home. 62 These complaints are consistent with the responses of Japanese companies to a recent MITI survey of their international operations. As of 1989, over 93 percent of the respondents managed their international operations from Japan, either by letting each functional division (e.g., marketing, manufacturing, or administration) manage both domestic and international business, or through an international business division that controls all overseas activi-
PAGE 105
100 Competing Economies: America, Europe, and the Pacific Rim ties. 63 Less than 6 percent of respondents have established a supervising corporation overseas to control manufacturing and sales in local markets, and less than 2 percent have overseas managing and supervising corporations that control all activities in local markets. Five years ago, however, less than 3 percent controlled any offshore operations from the local market. Nearly a third of respondents claimed that they would establish overseas managing and supervising corporations to control some or all of local operations by the mid1990s. Another hallmark of Japanese investment abroad is its pattern of purchasing capital equipment from the home country. The study of multinational companies doing business in Australia reported that the American and European companies equipment showed no national purchasing pattern; their equipment was made in America, Japan, and several European countries. Japanese companies, on the other hand, bought the overwhelming preponderance of their equipment from Japan. The studys author said: When an American or European company buys machinery to set up a plant, they take competitive bids. But the Japanese go directly to Japan. 64 Finally, for all industrialized countries, there is a fundamental asymmetry between Japanese investment abroad and foreign investment in Japan. The United States is just waking up to a reality that some nations in the industrialized world has long faced the presence of strong foreign commercial interests in the domestic market-and is reacting to it in what is probably typical fashion. Some people welcome the foreign investors, and see the increase in FDI as a response to market forces that will benefit U.S. consumers. Some react with heightened concern, even xenophobia, and regard any foreign influence as potentially suspicious. Opinions come in every shade between these two extremes. If Americans just awakening to these issues, Japan still slumbers. It is still the exception among industrial nations in the degree to which foreign investors are constrained from participating in its economy. While Japanese investors have aggressively stepped up investment in America, and recently in Europe, investment in Japan is still restricted. In 1985, assets of Japanese affiliates in the United States and those of U.S. affiliates in Japan were equal, both standing at $64 billion. In the next 3 years, assets of Japanese affiliates in this country multiplied more than fourfold, rising to $275 billion, while U.S. affiliates assets in Japan rose only to $129 billion. Leaving aside financial institutions, assets of Japanese affiliates in the United States rose 106 percent while U.S. affiliates assets in Japan grew 70 percent. American companies investing in Japan, particularly in preceding decades, were often obliged to license technologies, take a Japanese firm as a partner, or promise to limit market shares in return for permission to invest there; Japanese companies, in contrast, have been relatively free to invest in America, though the ambient political atmosphere surrounding their recent investments influenced their decisions. The asymmetry is not just bilateral. In 1986, FDI in manufacturing accounted for 10 percent of U.S. sales, 7 percent of employment, and 9 percent of assets in the manufacturing sector. Corresponding figures for major European economies are: for France, 27 percent of sales and 21 percent of employment; for Germany, 18 percent of sales, 13 percent of employment, and 17 percent of assets; and for the United Kingdom, 20 percent of sales, 14 percent of employment, and 14 percent of assets. In Japan, in stark contrast, FDI in manufacturing accounted for 1 percent of sales, employment, and assets. 65 Between 1960 and 1987, direct investment in Japan increased from 0.6 percent of the world total to 0.8 percent. During the same period, inward investment in the United States increased from 9.4 percent of all inward investment in the world to 25.2 percent, and Europes share increased from 29.8 to 37.6 percent. 66 While the United States and Europe debate the merits of foreign investment and how it contributes to national well-being, the Japanese seem to have made a clear choice: domestic firms are preferred to foreign firms in Japan. It is hard to say whether differences in Japanese investment behavior at home and abroad are a problem. Japanese firms are quite effective at selling in the markets they invest in. They are more likely than other foreign firms to build new plants and retain greater control over their affiliates (although their investment position is lower overall than those of European or Canadian investors), which could work to either the benefit or the detriment of the nation (or be neutral). For example, if Japanese affiliates invest heavily in training, R&D, and capital equipment, as Japanese parent firms generally do, then America may stand to gain more from Japanese investment than from investments by firms
PAGE 106
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 101 of other nationalities. In at least one industry, motor vehicles, Japanese affiliates have a good record of investment in training (see the discussion below), and the few figures available on the point indicate that their manufacturing affiliates invest more in plant and equipment, per dollar of sales, than foreign affiliates generally. R&D data by country of affiliate are even less adequate, but the scanty available figures suggest that Japanese affiliates R&D intensity is low. European firms fall closer to the average in all areas, but this is partly because European investment is so large that it has a greater effect on the average. In 1988, assets of European affiliates were 48 percent of the assets of all affiliates of foreign firms in the United States, while Japanese affiliates accounted for less than 24 percent. In manufacturing, where most of the R&D spending and capital investments take place, European affiliates had 63 percent of assets, and Japanese affiliates 10 percent. Japanese documents acknowledge that the international operations of Japanese companies have been mostly aimed at exporting, but there are indications that new strategies are emerging. As noted, MITIs survey showed that many more firms are planning to transfer significant control to foreign affiliates. Perhaps more significantly, increasing numbers of Japanese firms are also planning to spread R&D to offshore locations. At the time of the survey, 55 percent of respondents had R&D overseas only to provide technical support for sales and post-sale service; another 33 percent maintained foreign R&D bases to support local manufacturing. Another 10.5 percent maintained local R&D to design products specifically for local markets, while less than 2 percent maintained foreign R&D that was not directly in support of local market needs. 67 By the mid-1990s, however, 11.6 percent of the respondents planned to have foreign R&D aimed at general, rather than strictly local, corporate objectives. In addition, MITIs white paper maintains that Japanese corporations are in the first stage of global investment, and as investments mature, the expectation is that Japans trade surplus will diminish. Such plans, if implemented, will make Japanese investment both more acceptable and more beneficial to host nations. Some American observers also expect Japans trade surplus to dwindle as planned foreign investments are made. One member of MITs International Motor Vehicle Program maintains that exports of Japanese automobiles will eventually be replaced largely by production in overseas markets, partly because Japanese firms are becoming more confident about their ability to manage overseas facilities and partly because overseas markets will be increasingly unwilling to sustain large auto trade deficits with Japan. 68 A caution is in order, however. The plans of Japanese corporations are in line with the demands of foreign markets and governments, but it is not clear how much responsibility or R&D will be transferred offshore without such pressure, or whether plans to replace Japanese exports with offshore production will materialize absent the increasing trade friction of recent years : Selling abroad is not, of course, the only reason for international investment, and many Japanese firms will invest abroad even without pressure from foreign governments. But if there is a lessening of trade tensions (at this point, that does not appear likely) there may also be less change in the behavior or Japans offshore affiliates than the MITI survey suggests. Finally, it is always well to bear in mind that plans and expectations are often different from reality; unforeseen circumstances could well cause the respondents to MITIs survey to change their plans. Perhaps the best way of viewing the issue right now is summ ed up in a quote: The fears of some Americans, that the Japanese industrial presence in the United States is a mixed blessing, are not irrational. Japanese firms are not simply responding to trade friction by building an industrial presence in the U. S., but are pursuing a long-term strategy of creating an infrastructure which will enable them to sustain their market share above present levels, insulated from currency fluctuations and the vagaries of protectionist sentiment. 69 EMPLOYMENT AND EMPLOYEE COMPENSATION If it comes to a choice, most nations prefer foreign investment to imports, because foreign investment provides jobs, while the connection of imports with jobs is indirect at best. For example, one of the interests of the European Community, France and Italy in particular, in easing the adjustment to more open trade in motor vehicles is the employment they stand to lose if cars exported from Japan are allowed free access. The insistence of some in the EC on high levels of domestic content for foreign cars built in member countries. is additional evidence of the
PAGE 107
102 l Competing Economies: America, Europe, and the Pacific Rim concern for domestic workers and jobs. The effect of this demand will be either that of guaranteeing a market to domestic firms, or of ensuring that domestic workers get some of the benefit of foreign firms sales in the consuming country. Affiliates of foreign parents accounted for 1.7 million manufacturing employees in the United States in 1988; this was 8.9 percent of all U.S. manufacturing employment, up from 3.5 percent in 1977. While foreign corporations are, therefore, a growing force in U.S. manufacturing employment (figure 3-10), they are not necessarily creating jobs at the same rate that their employment has grown. Not all the jobs held by workers in companies called affiliates of foreign companies are jobs that the foreign company created through investment. DuPont is a good case in point; it is a foreign affiliate because a Canadian family owns 23 percent of the stock, yet it is difficult to believe that even 23 percent of DuPonts workers in the United States owe their jobs to the Canadian investment. On the other hand, the greenfield auto plants of the major Japanese auto companies, such as Nissan, Honda, Mazda, probably do represent net additions of jobs that would not exist otherwise. How many of the jobs these greenfield investments added is unknown; we do not know whether, in the absence of Japanese investment, other domestic producers would have made more cars, or if imports would have increased, 70 or what tradeoffs between production and price would have been made. Even in the case of these wholly new investments, the number of jobs involved almost certainly exceeds the upper bound of what could reasonably be called job creation resulting from FDI. 71 Although FDI does not create as many jobs as there are employees in affiliated establishments, it does affect employment in qualitative as well as quantitative ways. Where the foreign investors influence is significant, and therefore results in different training or a different business culture and management style, there are effects. Whether they are positive or negative is another matter. Training has been most carefully studied in Japanese-owned automobile assembly plants in the United States. They have a distinct edge over U.S.-owned assembly plants in the training they provide their employeesnot only shop-floor production workers, but also supervisors and manufacturing engineers (table 3-7). For newly hired producFigure 3-10-Employment in U.S. Affiliates, 1977-88 Employees (millions) 4 I 3 2 / 0~ ~ 1977 78 79 80 81 82 83 84 85 86 87 88 + Total ~ Manufacturing SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80,1985, table F-4, and subsequent series. tion workers, the difference is enormous (279 hours v. 46 hours), but there are substantial differences in every category. Note, too, that trainin g for most categories of employees is still greater in Japaneseowned plants in Japan than in the Japanese transplants in America. In some cases, Americans working for foreignowned companies speak glowingly about the lessons they have learned from their foreign parents. Managers at New United Motor Manufacturing Inc. (NUMMI), a joint venture of General Motors and Toyota, attribute much of the turnaround in the plants performance to their newly learned Japanese way of doing business. NUMMI operates from a plant in Fremont, California. The plant was formerly owned and run by General Motors and was often described as one of GMs worst plants. In 1982, it was shut down, presumably for good. It reopened 2 years later as NUMMI, with a crew of senior managers from Toyota and. a work force largely drawn from former UAW employees of the Fremont plant (80 percent of the workforce of NUMMI worked for GM Fremont). By 1986, when NUMMI was running at full capacity, its quality record matched that of Toyotas Takaoka assembly plant in Japan, while its productivity record was somewhat worse (19 assembly hours per car, compared with 16 at Takaoka). Its parts inventory averaged 2 days, compared with Takaokas 2 hours, but this was still substantially better than GMs Framingham plant, 72 where inventory averaged 2 weeks. GM says it is busy trying to pass the lessons it learned at NUMMI
PAGE 108
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation l 103 Table 3-7Training Hours Japan/Japan Japan/U.S. U.S./U.S Production workers: New hires a . . . . 315.5 279.3 45.7 Experienced workers . 87.0 53.3 28.8 Supervisors: New hires . . . . 160.0 260.0 166.7 Experienced workers . 109.2 80.0 60.4 Manufacturing engineers: New hires . . . . 864.0 466.7 155.0 Experienced workers . 156.7 100.0 72.3 a Newly hired empioyees, first 6 months on the job. SOURCE: John F. Krafcik, Training and the Auto Industry: International Companies, contractor report prepared for the Office of Technology Assessment, February 1990. on to managers at its other plants, though this has proved difficult, for it changes the job of every worker and manager at GM. 73 Even at NUMMI, the employment effects are not uniformly positive. There have been complaints about the pace of work and lack of seniority benefits in work assignments, 74 although no real strife has erupted. Also, it is not clear how much or how fast GM is able to transfer the knowledge the workers and managers of NUMMI have obtained from their experience with Toyota. Nevertheless, the ventures experiences have the potential of changing for the better the way some managers and workers think and work. So, too, do the other Japanese transplants. Other experiences are more problematic. In 1984, the Japanese firm NKK bought half of National Intergroup, Inc., making National Steel Corp. a Japanese affiliate. The Japanese chairman arrived in 1986, and the companys productivity has improved 16 percent since 1984-85, but it still remains one of the least profitable in the steel industry. Despite a $200 million annual capital improvement program, the company needs repairs for its blast furnaces, while unscheduled maintenance problems abound. The union there has one of the industrys most generous contracts, including a job security provision that restricts layoffs, but labor relations have been rocky, and the local unions have fought efficiency-boosting job flexibility. 75 Whatever lessons Japanese managers have to teach have been hard to pass on at National Steel. Finally, as noted above, Japanese firms may be less inclined to assign discretionary responsibility to American managers than other investors; certainly, they are less likely to hire American managers in America than U.S. firms are to hire Japanese managers in Japan. Can American workers and managers learn more from foreign direct investors than they already have? Probably, but there are limits. In some cases, the foreign parent is more a financial than a managerial presence, and many foreign affiliates are run the same as, or only slightly different from, American companies. In other cases, foreign companies may not have much to teach. In 1988, European affiliates had nearly 3 billion dollars worth of investment in motor vehicles and equipment in North America, but the quality and productivity records of the European auto manufacturers are worse than those of the Big Three American companies. 76 Unlike the Japanese case, there are few lessons to be learned from European auto production management. In terms of compensation and layoffs, foreign affiliates behave more or less like American companies. During the 1982 recession, U.S. manufacturing employment dropped 7 percent from the previous years employment, while sales dropped 5 percent. Foreign manufacturing affiliates employment dropped 4 percent, although sales increased 1.5 percent. Foreign affiliates manufacturing employment also dropped in 1985, as did U.S. manufacturing employment generally. While American firms have reputations abroad as fickle employers (it has sometimes been hard for them to recruit good employees in Japan because of their reputations as unstable employers during downturns), many foreign affiliates behave in similar ways when they do business in the United States. For example, in 1985, when the semiconductor industry worldwide went into a steep slump; at least two Japanese affiliates producing chips in U.S. plants (NEC and Toshiba Semiconductor) laid off workers in much the same way as their American counterparts in Silicon Valley .77 NEC official Koichi Shimbo told the San Jose Mercury News: When we are in the U. S., we do like the Americans. 78 In terms of compensation, foreign affiliates and American firms are very little different. While foreign affiliates pay higher compensation per employee overall than U.S. fins, this is due to the relatively heavy concentration of affiliates in highwage industries like banking. Within manufacturing, foreign affiliates and American firms pay nearly the same compensation to workers. 79 This is not surprising; if foreign firms failed to pay as much as
PAGE 109
104 Competing Economies: America, Europe, and the Pacific Rim Figure 3-n-Earnings and Reinvested Earnings: Foreign Direct Investors in the United States Billions of dollars 14 I A A A A -4 I 1 I 1 1 I I I I I 1 1977 78 79 80 81 82 83 84 85 86 87 88 89 ++ Total earnings ++Mfg. earnings + Reinvested earnings Mfg. reinvestment SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Survey of Current Business, various August and October issues covering the years 1977 through 1989, table 4. U.S. firms, they would have difficulty attracting workers. 80 PROFITS The way a multinational uses its profits also is an indicator of its commitment to foreign markets. In the 1960s, when American multinationals were investing heavily in Europe, concerns were raised about whether profits were reinvested or repatriated. Now, with FDI rising rapidly in the United States, we have come full circle, and Americans are wondering whether foreign companies use the profits made in the United States to benefit U.S. citizens, or instead send most of them back to the home country. According to Bureau of Economic Analysis (BEA) data, reinvestment depends heavily on earnings. The amount reinvested, particularly by manufacturing affiliates, bears a very close relationship to earnings, for all foreign direct investors in general, and for European and Japanese investors in particular (figures 3-11, 3-12, and 3-13). The difference between e arnings and reinvested earnings is distributed earnings, which are quite small for Japanese manufacturing affiliates and a bit larger for European affiliates. This fits with the well-known fact that Japanese manufacturing firms generally pay out very small dividends, compared with European or American manufacturers. If we look more broadly at all FDI (not just in manufacturing), distributed earnings are increasing, although the relationship with earnin gs is still strong. This is particularly striking for Japanese investment, whose distributed earnin gs increased modestly between 1982 and 1984, and then rapidly after 1984 (figure 3-12). The increase occurred in wholesale trade and a category called other, which includes retail trade, banking, finance, insurance, and real estate. In 1989,95 percent of Japanese affiliates distributed earnings came from these wholesale trade and the group labeled other, and in 1990, 96 percent. European affiliates distributed earnings originated differently; they came mainly from petroleum, other, and manufacturing, with very little from wholesaling. In 1988, manufacturing accounted for the largest reported share, 32 percent. European affiliates distributed earnings in 1988 and 1989 were reported in neither petroleum nor other separately, to avoid disclosure of data of individual companies. In 1988, the two sectors combined accounted for 58 percent of all distributed earnings, and in 1989, 68 percent.
PAGE 110
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 105 Figure 3-12Earnings and Reinvested Earnings: Japanese Direct Investors in the United States Billions of dollars 2.0 1.5 1.0 0.5 0. n -0.5 -1.0 1 I I 1 I I I I I I I 1977 78 79 80 81 82 83 84 85 86 87 88 89 + Earnings ~ Reinvested earnings + Mfg. earnings Mfg. reinvestmen t SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Survey of Current Business, various August and October issues covering the years 1977 through 1989, table 4. Figure 3-13Earnings and Reinvested Earnings: European Direct Investors in the United States Billions of dollars 10 8 6 4 2 0 2 1977 78 79 80 81 82 83 84 85 86 87 88 89 + Earnings +Reinvested earnings -++ Mfg. earnings + Mfg. reinvestment SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Survey of Current Business, various August and October issues covering the years 1977 through 1989, table 4.
PAGE 111
106 Competing Economies: America, Europe, and the Pacific Rim Figure 3-14Earnings and Reinvested Earnings: U.S. Direct Investment Abroad Billions of dollars 7 0 6 0 5 0 4 0 30, + A v 2 0 1 0 > 0 I 1980 81 82 83 84 85 86 87 88 89 + Earnings ++ Reinvested earnings + Mfg. earnings + Mfg. reinvestment SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, U.S. Direct Investment Abroad, Survey of Current Business, various August and October issues covering the years 1977 through 1989, table 20. One noteworthy feature of the data, going back to 1977, is that earnings of manufacturing affiliates are often small or negative. For Japanese affiliates they are hardly ever positive; since 1980, Japanese manufacturing affiliates earnings have been positive only once, in 1987, European manufacturing affiliates earnings dipped into the red only once, in 1982, a recession year. This observation raises several questions about how earnings are viewed. One possibility is that foreign investors in manufacturing, and Japanese investors in particular, are here mainly to gain market share, not profit, and can afford to sustain many years of financial losses. Eventually, of course, the firms must expect to profit from the increase in market share, but perhaps not yet; substantial Japanese direct investment in manufacturing is fairly recent. Another possibility is that foreign affiliates earnings are calculated with an eye to where the parent company would most like to pay corporate taxes and get tax breaks. Suggestions that affiliates in the United States are charged higher-thanmarket prices for both goods imported from the parent organization or for intangibles (e.g., R&D) come up from time to time, but so far are unresolved. It is possible that earnin gs of foreign affiliates are understated because parent firms prefer to deal with corporate taxes at home rather than in the United States. Whatever the resolution, Japanese affiliates in America are acting differently in making or reporting earnings and in reinvestment. American multinationals earnin gs and reinvested earnings in their overseas affiliates, overall and in manufacturing, have been positive throughout the 1980s (figure 3-14). This may reflect the fact that, compared with Japanese firms, American firms are under heavier pressure to make and distribute earnin gs, or it may have to do with differences in corporate tax rates and incentives here and abroad. INTERNATIONAL TRADE Foreign affiliates have a higher propensity to import than American firms, and the overall trade deficit associated with the operations of affiliates is substantial (figure 3-15). In 1988, the deficit in merchandise trade associated with U.S. affiliates of foreign investors (about $90 billion) was 75 percent of the total U.S. merchandise trade deficit ($120 billion). Between 1977 and 1988, U.S. affiliates of foreign firms increased their merchandise imports 11.3 percent annually, from less than $46 billion to nearly $150 billion, while their merchandise exports increased from $21 billion to $52 billion, 8.6 percent per year. The affiliates trade deficit accordingly increased from $21 billion in 1977, when the United States had an overall merchandise trade deficit of
PAGE 112
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation l 107 Figure 3-15-Merchandise Trade, FDIUS Affiliates Billions of dollars o 3 -2 0 -40 -60 -80 -100 w a I I I I I I I I I I 78 79 80 81 82 83 84 85 86 87 88 ~ Total Japa n ~ Europe SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates, 1977-80, 1985, table H-3, and subsequent series. $27 billion, to nearly $90 billion in 1988. In 1977, European affiliates accounted for 42 percent of the trade deficit associated with FDI, but their share dropped to 32 percent by 1988, while Japanese affiliates share increased from 28 percent of the total deficit to 57 percent (figure 3-15). 81 The nature of affiliates imports varies by country. In 1988, Japanese affiliates imported mostly through their wholesale trade establishments in the United States; 93 percent of their imports came to wholesalers and only about 7 percent to manufacturers. Forty-four percent of the merchandise imports of all Japanese affiliates were motor vehicles shipped to affiliated Japanese wholesalers ($33.4 billion of a total of $75.9 billion). Motor vehicle imports to wholesalers dwarfed the next largest category of Japanese affiliates imports, $12.8 billion in imports to electrical goods wholesalers. All this supports the point made earlier, that the preponderance of Japanese FDI operations in the United States is related to selling goods made in Japan, and this shows up in trade figures as much as in sales. While Japanese affiliates account for a substantial trade deficit in motor vehicles, there is a noteworthy countertrend. Honda, the first Japanese automaker to set up U.S. manufacturing operations, now exports cars from the United Statesin fact, it expects to export 70,000 cars from the United States this year. If it does, its exports will exceed auto exports by the Big Three U.S. automakers to all nations except Canada (which has a longstanding free trade agreement with the United States in motor vehicles) .82 The U.S. content of Hondas motor vehicles was low, only 25 percent, when the company began operations here in 1982, but Honda claims it will have 75 percent North American content in its U.S. and Canadian operations by 1992. 83 So while the balance of bilateral motor vehicle trade between the United States and Japan is still heavily tipped in favor of a Japanese surplus, the irony is that Japanese direct investment in the United States may end up contributing disproportionately to U.S. motor vehicle exports, too. European affiliates imports in 1988 also came mostly through wholesalers, but to a much smaller extent than Japans: 54 percent of European affiliates imports were done by wholesalers. Manufacturing affiliates accounted for 34 percent of all European affiliates merchandise imports. Like the Japanese, European affiliates largest single category of imports was motor vehicles imported by wholesalers, but autos were a much smaller proportion of their total merchandise imports, only 25 percent. The second largest category was imports by
PAGE 113
108 l Competing Economies: America, Europe, and the Pacific Rim manufacturers of electric and electronic equipment (6.6 percent of total merchandise imports). Altogether, foreign affiliates accounted for imports of motor vehicles and parts totaling $50.7 billion in 1988 (including imports by manufacturers and wholesalers), and a motor vehicle trade deficit of $46.1 billion. 84 This compares with a total U.S. trade deficit in motor vehicles of $49.2 billion in the same year. Manufacturing affiliates accounted for $21 billion in merchandise exports, and $29.3 billion in imports; wholesale trade affiliates accounted for a total of $35 billion in exports and $109.6 billion in imports. Clearly, the FDI deficit in auto imports by manufacturing affiliates is small compared to auto imports by wholesalers. Firms wishing to sell in Japan cannot do it so easily by establishing a wholesaling affiliate in that country, nor would many small U.S. firms have the resources to do so even if there were no institutional barriers against it. Many of the small American high-technology firms that have formed alliances with foreign partners cite as one of the benefits their increased access to the Japanese market (see the Nicolet example above). Sometimes the arrangements are more complicated: the Japanese tractor maker, Kubota, bought a stake in the American company Cummins Engine to take advantage of Cu mmins extensive European network, in preparation for European market integration in 1992. 85 Although FDI is associated with a large negative effect on the U.S. trade balance, it is misleading to think of FDI as the cause of our trade deficit. The fundamental causes are our puny national savings rate (greatly exacerbated by the Federal budget deficit), and the failure of U.S. manufacturing to keep up technologically with increasingly able competitors, principally Japanese. 86 According to economic theory, a nations current account trade balance (which includes trade in services, transfer payments from governments, and income from property abroad, as well as trade in goods) is determined by the national rates of savings and investment; over time, the current account trade deficit (or surplus) is equal to the difference between domestic investment and domestic saving. If foreign investors imports are persistently larger than their exports, this would tend to widen the U.S. current account trade deficit, but then the value of the dollar would presumably drop, making U.S. exports cheaper and returning the trade deficit to the level determined by the relation between domestic savings and domestic investment. In such circumstances, weakening of the dollar might be postponed if foreigners invested enough of their savings in the United States to sustain a widening current accounts trade deficit; this is what happened during the early 1980s. Later in the decade, the dollar declined and the merchandise trade deficit and current account deficit narrowed. However, this process is costly. A lower dollar raises the price of imported goods, and in the long run, reliance on a cheap dollar to right the trade balance tends to undermine the U.S. standard of living. 87 The most constructive way to get rid of the U.S. trade deficit is to produce goods that the world will buy because they are well-made and of good value. NATIONAL INTERESTS, BUSINESS INTERESTS What do nations want from firms, and what do firms do? Nations want things that make citizens better off: well-paid jobs, additions to knowledge and productivity, exports, investment. In some ways, foreign fins affiliates in the United States measure up well on many counts, compared with firms whose headquarters are in the United States, and less so on others. Foreign affiliates and U.S. firms are similar in their compensation of workers. Foreign affiliates do less R&D here per dollar of sales than do American firms, and they have a much greater propensity to import. In the latter regard, Japanese affiliates are noteworthy for their heavy importing, almost all of it from Japan. Some affiliates have made valuable contributions to workers skills and to managers competence, through training and object lessons; others operate in very much the same way as U.S. fins. They are about as reliable, in terms of job security, as American fins. All of this, of course, is on average. Japanese firms are particularly oriented to selling here; European fins investments apparently represent a more diverse set of goals. What all this means is that most of the differences in behavior between American and foreign firms are not very striking. Decisions about who ought to be allowed access to programs designed to improve competitiveness and living standards for Americans would therefore be more discrimin ating if they were made on the basis of individual fins behavior and performance, rather than strictly on nationality.
PAGE 114
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 109 This begs the question of political influence, deliberately. Whether foreign firms or interests have undue influence in national, state or local politics, or whether their influence is exerted in ways that will harm American competitiveness and living standards or aid it, is beyond the scope of this report. Political and national security concerns are relevant to the debate over access to publicly funded programs to enhance competitiveness, and they ought to be; their absence is not a dismissal. The Policy Environment: How America Treats FDI The above sections suggest that there are many dimensions of foreign investment to consider if the U.S. Government wants to adopt performance criteria for deciding how to treat affiliates of foreign firms. The discussion so far has focused on what measures might be appropriate if the decision were about whether affiliates were permitted to participate in government-funded programs to promote industrial competitiveness. So far, the Federal Government has not dealt with this issue systematically. It has come up a couple of times, but not specifically as a competitiveness issue. For example, a few European firms want to participate in Sematech, just as a few American firms want to participate (and as IBM has been allowed on a limited basis) in the European semiconductor manufacturing and development consortium, JESSI. 88 The decision in the United States has been to limit participation to American firms whose headquarters are in the United States and without a controlling foreign ownership position. Part of the rationale for the decision may have been national security. If that is so, then the position of the United States is either changing or is inconsistent with its stated principles and goals. The official position of the U.S. Government on direct investment, both U.S. direct investment abroad and FDI in the United States, is that firms investing offshore should be treated no differently from domestic fins. This so called national treatment standard is the mirror of the official U.S. position toward international trade in the GATT, but the legal principles and policies are not so well defined in the investment arena as they are in trade. 89 In a few sectors, this principle of neutrality is abridged by one of reciprocity, which stipulates that American firms must be treated abroad as foreign firms wish to be treated here. 90 The United States, like most nations, makes a number of exceptions to these standards. The most prominent is for national security. Exceptions to national treatment for national security reasons are made for two reasons: political sovereignty and military capability. Many nations have discrimin ated against foreign firms to keep them from gaining too great an influence over the nations economy or political decisionmaking. The standard for what constitutes too great an influence is soft, and often handled case by case. Nations also seek to assure that the capacity to produce military goods and services will be at the governments disposal when needed, and that there will be no unauthorized transfers of sensitive technologies or products. 91 These concerns are recognized by most nations as legitimate, and international agreements covering direct investment permit nations to make exceptions for national security purposes. 92 The ability of the government to make exceptions to the national treatment standard for national security purposes was recently strengthened, at least in theory, in the Omnibus Trade and Competitiveness Act of 1988, which added section 721, often called the Exon-Florio amendment, to the Defense Production Act. This provision allows the government to block foreign mergers, acquisitions, or takeovers of U.S. firms if there is a threat to national security .93 Implementation of the provision is done by the Committee on Foreign Investment in the United States (CFIUS), whose members come from various Federal departments and agencies. 94 So far, CFIUS has not ventured beyond a fairly narrow interpretation of national security. Critics argue that it should interpret national security more broadly to encompass strategic areas of the civilian economy, in the same way that Sematech received DoD funding on grounds that the ability to produce high-performance electronics products for national security depends on a competitive civilian industry. A crucial test of CFIUSs willingness to interpret national security more broadly was its recent decision to permit a Japanese company to acquire Semi-Gas Systems Inc., the leading U.S. producer of high-purity gas systems for semiconductor manufacture. The President, acting on CFIUS recommendation, decided that the purchase of Semi-Gas by Nippon Sanso would not threaten national security, and the Justice Department decided the sale would not violate antitrust laws. Semi-Gas and Nippon Sanso are first and second in world sales of
PAGE 115
110 Competing Economies: America, Europe, and the Pacific Rim semiconductor gas equipment, with Semi-Gas having 21.5 percent of the world market and Nippon Sanso 17 percent. After the acquisition, Nippon Sanso will control nearly 40 percent, far ahead of the next competitor, Air Products, which has 14 percent of world sales. 9s Although CFIUS has not yet broadened the definition of national security to include competitiveness of dual-use industries, the distinction between what is done for national security purposes and what is done to promote civilian industrial competitiveness is blurred. In some critical industries, such as electronics and telecommunications, the demarcation between the defense industrial base and overall U.S. industry is blurred; DoD must rely on technologies, people, and productive capacity that serve both civilian and military markets. If the Nations electronics sectors competitiveness declines, so too might DoDs ability to be able to fulfil the militarys production needs for either defense preparedness or for times of national emergency or war. Moreover, the Nations economic performance is at least equal in importance to military security, and policymakers are searching for ways in which the U.S. Government and industry can collaborate to strengthen Americas competitive position. 96 However, the U.S. Government, particularly the last two Administrations, has steadfastly maintained positions in favor of free markets and against national intervention to promote economic competitiveness. This has led, some argue, to a tendency to find national security rationales for programs designed to promote economic competitiveness, including exceptions to national treatment standards for foreign firms investing in the U.S. market. The United States probably cannot continue to invoke national security for all programs to promote civilian industrial competitiveness. The United States is behind on too many fronts, and in too many high-visibility industries, not to confront the issue of economic competitiveness, and the governments proper role relative to it, for its own sake. Anew government program with the unambiguous purpose of improving commercial technology is the Advanced Technology Program in the U.S. Department of Commerce. The law states that the programs mission is to improve the competitive position of the United States and its business and to help United States businesses create and develop generic technologies with commercial potential. 97 The Program may help joint R&D ventures with technical advice or may take part in the ventures, providing start-up funding or a minority share of the cost. Created in 1988, the Program got its frost funding ($10 million) in fiscal year 1990; the next year Congress upped the amount to $36 million and at the same time defined conditions under which foreign firms may participate in fiscal year 1991. The approach is to apply performance standards to both foreign affiliates wishing to participate and U.S.-owned companies. The performance standards stipulate that participating companies shall have investments in U.S. R&D and manufacturing (not limited to screwdriver assembly of imported components); a significant employment base; agreements to promote U.S. manufacture arising from any technologies developed in such ventures and to procure materials and components from the United States or Canada. The Secretary of Commerce is given the authority to find companies eligible for participation, using the performance standards as evidence that participation would be in the economic interest of the United States. In addition, reciprocity provisions apply to foreign participants. They may take part if the Secretary finds that their home country offers U.S.-owned companies comparable opportunities to participate in joint ventures, allows U.S. companies to invest on equal terms with other countries, and affords adequate protection of the intellectual property rights of U.S. companies. 98 For participation in government-sponsored programs such as the Advanced Technology Program, performance standards can be applied as a kind of screen, or they can be used on a case-by-case basis. While most observers seem to prefer performance requirements to discrimination based solely on ownership, 99 performance requirements are controversial, too. Graham and Krugman argue that performance requirements can introduce economic distortions, just as-trade protection can, that could reduce economic well-being and serve as a vehicle for political tampering. The use of performance standards to govern foreign affiliates participation in Federal programs, and other policy options, are discussed in chapter 2. While considerable information is available about the behavior and performance of foreign affiliates in the U.S. market, the kind of information that would allow an executive agency like CFIUS to discrimi-
PAGE 116
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 111 nate between investments that are likely to contribute to U.S. competitiveness and those that could endanger it is not always available, and some that is available is not used in todays permissive climate. The issue is very likely to heat up, however, in which case there will be increasing interest in how foreign affiliates contribute to the U.S. economy and technology. The data that exist tell us much, but they can also mask significant differences in contributions at the firm, industry, and country levels. Some of the most pressing questions, for example, about the flow of technology and value added when Japanese investors acquire control of high-technology American fins, cannot be fairly addressed with the data we have from government sources, and may require additional investigation. The behavior of foreign affiliates in general, and Japanese affiliates in particular, is similar enough to U.S. firms in many ways that it is not a simple thing to decide to exclude them on the basis of ownership alone. Neither is their behavior so similar to that of U.S. firms that national treatment standards are a matter of no consequence. 1 John W. Rutter, Direct Investment Update: Trends in International Direct Investment U.S. Department of Commerce, International Trade Adminstration staff report, September 1989. A U.S. affiliate of a foreign direct investor is defined as one in which an investor from another country owns or controls at least 10 percent of a domestic companys voting stock (or equivalent amount in an unincorporated enterprise). Definitions are given in U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: 1987 Benchmark Survey, Final Results (Washington DC: U.S. Government Printing Office, 1990). 2 OECD is the Organization for Economic Cooperation and Development. Its members include Austria, Belgium, Canada Denmark France, German Y Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden Switzerland, Turkey, the United Kingdom, and the United States. 3 This refers to the behavior of U.S. affiliates only; Japanese and many European countries have laws and traditions that make them much less prone to lay off workers at home. 4 For discussion of these conclusions, see U.S. Congress, Office of Technology Assessment, Paying the Bill: Manufacturing and Americas Trade Deficit, OTA-ITE-390 (Washington DC: U.S. Gov ernment Printing Office, June 1988) and Mahng Things Better: Competing in Manufacturing, OTA-ITE443 (Washington DC: U.S. Government Printing Office, February 1990). 5 The dollar may not fall if, as in the early 1980s, interest rates are high enough and confidence in the U.S. economy great enough that foreign investment is sufficient to offset the deficit in merchandise trade and support the dollar at a high leveltemporarily, at least. 6 Discussion of how the European Community treats foreign investors is in ch. 5. 7 Edward M. Graham and Paul R. Krugman, Foreign Direct Investment in the United States (Washington DC: Institute for International Economics, 1989), pp. 21-24. 8 Public Law 100418, 100th Cong., Title V, part II, sec. 5021, which amended the Defense Production Act of 1950, adding sec. 721. 9 Robert B. Reich Who Is Us? Harvard Business Review, January-February 1990. 10 Raymond J. Mataloni, Jr., U.S. Multinational Companies: Operations in 1988, Survey of Current Business, June 1990, pp. 31-44. 11 These measures of the characteristics of jobs in U.S. parent companies versus those in their foreign affiliates were put forth by Laura Tyson in her paper, Us, Them, and Competitiveness, The American Prospect, winter 1990-91. Tyson points out that even in developed countries, jobs in foreign affiliates are not the equal of those in the parent companies, although the discrepancies are not so great. Compensation per employee in developed country affiliates is 90 percent of that in parent companies, and assets per employee, 87 percent. The comparative data are from Mataloni, ibid. 12 U.S. Department of Commerce, Bureau of Economic Analysis, U.S. Direct Investment Abroad: 1982 Benchmark Survey Data (Alexandria, VA: National Technical Information Service, 1985), tables III.D.1, 111.H.1, and III.Q.1. The data on R&D spending are for majority-owned affiliates and parent companies of U.S. multinationals; other data are for non-bank affiliates, which include foreign companies in which a U.S. investor directly owns or controls at least 10 percent of the voting stock (or an equivalent portion in unincorporated enterprises). 13 OTA interview with Jean-Jacques Duby, Group Director, Science and Technology, IBM Europe, Paris, Oct. 8, 1990. 14 Rutter, op. cit. 15 Jeffrey H. Lowe, Gross Product of U.S. Affiiates of Foreign Companies, 1977-87, Survey of Current Business, June 1990, pp. 45-53; and Russell B. Scholl, The International Investment Position of the United States in 1990, Survey of Current Business, June 1991, pp. 29, 32. Historical series understate the value of U.S. direct investment abroad, and in comparison overstate the value of foreign direct investment in the United States, as the value of both is measured in book value. This valuation does not capture all the appreciation of investments, and the older an investment, the more likely it is to be undervalued in the official figures. Since for many decades the United States invested more abroad than foreign companies and governments invested here, comparisons of foreign direct investment with U.S. direct investment abroad can be somewhat misleading. However, there are significant inaccuracies in attempting to measure the true value of investments as well. Historical costs are used throughout this report. 16 The investment position of a foreign direct investor means its total equity in and net outstanding loans to, its U.S. affiliates. In contrast, the assets of an affiliate are equal to the equity of all owners, domestic and foreign, in the affiliate, less all liabilities of the affiliate. Sales is, as implied, the sales made by the affiliate in the United States. 17 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates of Foreign Companies, Preliminary 1988 Estimates (Washington, DC: U.S. Government Printing Office, 1990), table A-2. 18 Graham and Krugman, op. cit., p. 13. 19 U.S. Department of Commerce, Bureau of Economic Affairs, Foreign Direct Investment in the United States, Preliminary 1988 Estimates. op. cit., table A-1, and The U.S. National Income and Product Accounts, Revised Estimates, Survey of Current Business, July 1989, table 6.6B. 20 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Preliminary 1988 Estimates, op. cit., table E-8. 21 Wholesaling affiliates handle inventories of manufactured products and their sales and distribution to retailers. 22 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Preliminary 1988 Estimates, op. cit., tables G-6 and A-2. 23 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, 1987 Benchmark Survey,
PAGE 117
112 l Competing Economies: America, Europe, and the Pacific Rim Final Results (Washington, DC: U.S. Government Printing Office, August 1990), p. 148. 24 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, Preliminary 1988 Estimates (1990) op. cit., table E-8. 25 Most of these European affiliates were purchased, not built from scratch (e.g., Thomson Consumer Electronics consists of the old RCA consumer products group, which General Electric bought and then sold to Thomson SA. 26 This figure is a good illustration of the caution with which data on foreign direct investment should be treated. It is not so difficult for Americans with any knowledge of business to name a few European chemical companies, but most of us would have trouble thinking of a big Canadian company. Who are the big Canadian chemical companies investing in America? In this case, one is DuPont, the picture of an American company. A Canadian family, the Bronfman family, owns 23 percent of DuPonts stock, however, making it a foreign affiliate. Source: Michael Hodges, The Japanese Industrial Presence in America: Same Bed, Different Dreams, Millenium, winter 1989, p. 362. 27 For good reason. Honda was considered somewhat of an outsider in Japan, both by other companies and by the Ministry of International Trade and Industry (MITI); its share of the Japanese market was relatively small, while in the United States, Honda cars were becoming the most popular imports. Honda is also very dependent on exports, and was therefore more vulnerable to protectionist moves in other countries. Exports account for 70 percent of Hondas Japanese production. James P. Womack, Daniel T. Jones, and Daniel Roos, The Machine That Changed the World (New York NY: Rawson Associates, 1990), pp. 215-216. 28 James P. Womack and Daniel Roos, Case Study: The Automotive Industry, contract report for the Office of Technology Assessment, May 13, 1988. The 3 million figure also includes estimated production of 220,000 vehicles in a Suzuki-GM plant in Ontario. 29 John F. Krafcik and John Paul MacDuffie, Explaining High Performan ce Manufacturing: The International Automotive Assembly Plant Study, paper prepared for the International Motor Vehicles Program International Policy F O rum May 1989. 30 U.S. General Accounting Office, Foreign Investment: Growing Japanese Presence in the U.S. Auto Industry, GAO/NSIAD-88-111, March 1988, p. 33. This is not true of all U.S. parts makers; some are large, integrated companies like Rockwell and TRW, which have experienced engineering staffs. However, the American industry is more characterized than the Japanese or European industries by small firms with little or no ability to contribute to design and manufacturing engineering, 31 General Accounting Office, op. cit., p. 29. Another reason for sourcing major subassemblies like engines and transmissions from abroad is economies of scale, as noted in the GAO report. Because of the level of investment in assembly needed to support a facility for making engines, it maybe more cost effective to increase engine production in existing overseas facilities than to create new capacity to support new assembly transplants. 32 General Accounting Office, op. cit., p. 11. 33 All material about Hoechsts purchase of Celanese is from Robert R. Miller, The Impact of Merger and Acquisition Activity on Research and Development in U.S.-Based Companies, contract report to the Office of Technology Assessment November 1989. 34 See, for example, Stephen Kreider Yoder, Kubota Is Sued by Founders of U.S. Firm, Asian Wall Street Journal Weekly, July 16, 1990, p. 9; Christopher J. Chipello and Jacob M. Schlesinger, Tractor Maker Kubota Is Plowing Funds Into U.S. High-Tech Firms, Asian Wall Street Journal Weekly, July 23, 1990, p. 6. 35 Gina Kolata, Japanese Labs in U.S. Luring Americas Computer Experts, The New York Times, Nov. 11, 1990, p. 1. 36 Kenneth Flamm argues that a small group of Japanese suppliers of dynamic random access memory (DRAM) semiconductors had enough market power in the late 1980s to limit chip production raise prices in the world market and collect extraordinary profits, rather than aggressively compete with each other on price. (It should be noted that, according to some analysts, this behavior was, in effect, sanctioned by the U.S.-Japan Semiconductor Trade Agreement of 1986. Others believe that the trade agreement had little to do with the movements of DRAM prices.) At the same time, the integrated Japanese electronics companies used their chips (for which their cost was much lower than the world price) in their own computers, and began to compete very successfully for the first time in the U.S. personal computer market. See Kenneth Flamm, Semiconductors, in Gary Clyde Hufbauer (cd.) Europe 1992: An American Perspective (Washington DC: The Brookings Institution 1990), pp. 248-260. 37 The modem debate began in the late 1970s, when some proposed that the United States require a certain percentage of domestic content in cars assembled here. 38 Jean-Christophe de Gouville, European Carmakers GetReady To Battle the Japanese, Japan Econonic Journal, July 28, 1990, p. 16. 39 Electronic Engineering Times, Europeans May Miss Own Fab Boom July 16, 1990. 40 Other measures might include the patenting and licensing activity, and citations to technical articles. Data for these measures are not available for foreign firms investing in the United States. 41 It is possible that virtually all R&D done by foreign affiliates in the United States is done for, if not by, manufacturing affiliates. However, R&D intensity of U.S. firms is usually presented in terms of manufacturing R&D as a percent of manufacturing sales, so the data for foreign affiliates are presented here in the same way, for consistency. 42 R&D intensity may have been unusually low for Japanese manufacturing affiliates in 1987; all R&D spending, taken as a percent of sales of Japanese manufacturing affiliates, was unusually low that year-1.3 percent, compared to figures as high as 2 percent in other years. However, all R&D spending as a percent of manufacturing sales was consistently lower for Japanese affiliates than for European affiliates throughout the period 1977-88 (data for manufacturing R&D, by country, are available only for 1987). 43 Graham and Krugman, op. cit., pp. 58-62. 44 Data for investments in new plant and equipment are not available by country and by industry, so the more comprehensive figure for capital investments-property, plant and equipment-is used. No comparable figure for U.S. manufacturing firms is readily available. 45 Japanese investment in transportation equipment, almost all in greenfield motor vehicle plants, increased fivefold from 1985 to 1988, rising from $715 million (7 percent of all Japanese manufacturing assets in the United States) to $3.5 billion (12 percent) in the 3 years. Source: U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States, op. cit., 1985, table B-7 and 1988, table B-5. 46 Tsusho-sangyosho (MITI), Tsusho Hakusho (Commerce White Paper) Heisei 2 nen (1990), 7 gatsu, p. 192. 47 The overseas markets were the United States, the EC, the newly industrializing economies (NIEs), and ASEAN nations. The percentage of respondents citing increased local sales as the main reason for investing exceeded 80 percent in all except the NIEs, where the percentage was between 70 and 75 percent. 48 Tsusho-sangyosho, op. cit., p. 192. 49 This is inferred from the pattern of investment in the United States, rather than from Japanese literature. According to a recent MITI survey of international investors, collecting technology and market information ranked eighth (out of nine) on the list of reasons for investing in the United States, cited by only about 12 percent of respondents. The same motive ranked eighth out of 10 on the list of reasons for investing in the EC (cited by about 18 percent of respondents); ninth of 9 on the list for the ASEAN nations (cited by less
PAGE 118
Chapter 3--American Firms, Foreign Firms: Contributions to the Nation 113 than 10 percent of respondents); and did not figure in investments in NIEs. 50 Japanese Investments in Small U.S. High-tech Companies Grow, R&.D Magazine, September 1990, p. 18. 51 Nicolet Signs With Matsushita, Electronic Engineering Times, Sept. 24, 1990, p. 2. 52 For example, in a recent article the president of Meitec America, George Leslie, says, .Japanese firms are ideal candidates for strategic partnerships. In fact forging a strategic alliance with a Japanese firm not only can provide an additional source of funding but also can pave the way to Japanese market entry. See George Leslie, Getting a Read on Strategic Alliances, Electronic Engineering Times, Sept. 3, 1990, p. 35. 53 Ibid. 54 Letter from Senators Jeff Bingaman and Lloyd Bentsen to James F. Rill, Assistant Attorney General, Antitrust Division, U.S. Department of Justice, Sept. 27, 1990. See also Sheridan Tatsuno, U.S. Threatened by Rash of High-Tech Buyouts, New Technology Week, Monday, Aug. 6, 1990, p. 7. 55 For a short description of these arguments, see Tatsuno, ibid. 56 U.S. semiconductor manufacturers allege that seven types of semiconductor equipment became available to them 2 to 3 years after delivery to semiconductor firms in Japan. Bingaman and Bentsen, op. cit. In another example, the Manufacturers Alliance for Productivity and Innovation states in 1990 that intra-keiretsu purchasing patterns allowed Japanese investors sufficient market power to avoid competition from U.S. companies. Source: Phyllis A. Genther and Donald H. Dalton,~ Japanese Direct Investment in U.S. Manufacturing, International Trade Administration (Washington, DC: U.S. Department of Commerce, June 1990), p. 9. 57 Ibid., p. 17. 58 U.S. Electronics Strengthens Foothold in Japan, New Technology Week, Monday, Jan. 8, 1990. 59 Shadow of Suspicion, Far Eastern Economic Review, July 26, 1990. 60 Quoted study was done by Mordechai Kreinin of Michigan State University, and written up in The World Economy, a British Journal. 61 Urban C. Lehner and Alan Murray, Japanese Buying Binge Stokes U.S. Insecurity, Asian Wall Street Journal Weekly, June 25, 1990, p. 12. 62 Ibid. 63 Tsusho-sangyosho, op. cit., p. 189. 64 Lehner and Murray, op. cit. 65 Graham and Krugman, op. cit., p. 25. 66 Rutter, op. cit., p. 15. 67 Tsusho-sangyosho, op. cit., p. 188. 68 James Womack, Massachusetts Institute of Technology, personal communication with OTA staff. 69 Hodges, op. cit., p. 361. 70 Throughout the 1980s, Japanese exports of automobiles to the United States have been constrained by a formal VRA and, after its expiration in 1985, an informal pledge. These agreements made it very difficult for Japanese producers to increase exports, and encouraged them to invest in North American production facilities to increase sales. 71 Supposing, for example, that production from the greenfield plants of foreign direct investors substitutes entirely for imports, rather than for output of a domestically owned and located firm. There are always some domestic jobs (e.g., in transportation, warehousing, and sales) associated with imports, so that not all the new jobs involved in making, storing, transporting, and selling the goods from the greenfield plant are net additions to the jobs that would have been associated with imports. 72 Framingham is described as the worst American-owned auto plant. Womack et al., op. cit., p. 85. 73 Womack et al., op. cit., pp. 82-84. 74 See, for example, Mike Parker and Jane Slaughter, Management by Stress, Technology Review, October 1988, pp. 36-44. 75 Clare Ansberry, New Japanese Chief Seeks Turnarmmd at National Steel, Asian Wall Street Journal Weekly, Aug. 6, 1990, p. 23. 76 The high quality attained by the luxury auto manufacturers of Europe is not associated with high productivity and cost control. The mass production auto manufacturers of Europe have no better quality records and substantially worse productivity records than the American Big Three. John F. Krafcik and John Paul MacDuffie, Explaining High Performance Manufacturing: The International Automotive Assembly Plant Study, paper prepared for the International Motor Vehicles Program International Policy Forum May 1989. 77 David Sheridan, Worker Displacement in the California High-Tech Industry, contractor report to the Office of Technology Assessment May 1986. 78 Dedra Hauser, Overseas Employees Favored in Cutbacks, San Jose Mercury News, Oct. 29, 1985, pp. B9-B1O, cited in Sheridan, op. cit. 79 Graham and Krugman, op. cit., pp. 56-57. 80 They are sometimes able to pay a great deal more, however. That would be unlikely in the United States, which generally has higher wages than other countries, but in developing nations, employees of foreign affiliates are sometimes the best-paid workers around. 81 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates of Foreign Companies, various years. 82 John M. Kline, Trade Competitiveness and Corporate Nationality, Columbia Journal of World Business, fall 1989, p. 29. 83 Womack et al., op. cit., p. 217. Domestic content claims, according to these authors, are notoriously unreliable, but they judge Hondas claim is not far off the mark. 84 U.S. Department of Commerce, Bureau of Economic Analysis, Foreign Direct Investment in the United States: Operations of U.S. Affiliates of Foreign Companies, Preliminary 1988 Estimates (Washington DC: August 1990), tables G-3 and G-6. 85 Christopher J. Chipello, Tractor Maker Kubota Is Plowing Funds Into U.S. High-Tech Firms, Asian Wall Street Journal, July 23, 1990, p. 6. 86 Evidence and arguments for this conclusion are detailed in U.S. Congress, Office of Technology Assessment Paying the Bill: Manufacturing and Americas Trade Deficit, OTA-ITE-390 (Springtleld, VA: National Technical Information Service, June 1988). 87 A falling currency dimini shes standards of living by making imports more expensive. In addition, unless all our international debts are denomina ted in dollars, it also obliges us to work harder to pay our debts to foreign creditors; a falling or volatile currency will also make it harder to obtain dollar-denominated loans. 88 JESSI stands for the Joint European Submicron Silicon Initiative. Ch. 5 has a more extensive discussion of JESSI. 89 Kline, op. cit., p. 30. 90 Graham and Krugman point out that foreign or foreigncontrolled companies cannot acquire gas pipeline rights-of-way or certain mineral leases on Federal land if the investors home country denies such rights to American firms. Graham and Krugman, op. cit., p. 96. 91 Kline, op. cit., p. 27. 92 These exceptions do not necessarily limit defense production solely to domestic firms, however, as the large international trade in
PAGE 119
114 l Competing Economies: America, Europe, and the Pacific Rim military equipment attests. For example, the Department of Defenses buy American procurement regulations require not that the companies doing its production be American owned or controlled, but that the products involved must have at least a 50 percent U.S. content. The Department of Commerces trade fairs permit any firm with 51 percent U.S. content in its products assembled or produced abroad to participate, and the U.S. Export-Import Bank will make loans or loan guarantees to anyone whose products have at least a 50 percent U.S. content. Source: Kline, op. cit., pp. 26-27. 93 The Defense Production Act lapsed, temporarily at least, when in the closing hours of the 101st Congress, in October 1990, the Senate failed to adopt the conference report re-authorizing the act, and did not agree to concurrent resolutions passed by the House extending the act. 94 CFIUS members include representatives of the Departments of Commerce, Defense, Justice, and State; the Office of Management and Budget the U.S. Trade Representative, and the Council of Economic Advisers. 95 Bingaman and Bentsen, op. cit. Information in the letter was based on Dataquest, Inc. figures. 96 For a discussion of options that might be adopted to strengthen such government-industry collaboration, see U.S. Congress, Office of Technology Assessment, Making Things Better, op. cit. 97 Omnibus Trade and Competitiveness Act of 1988, Public Law 100-418, Sec. 5131. 98 U.S. Congress, House of Representatives, Conference Report 101-909 to Accompany HR. 5021, Making Appropriations for the Departments of Commerce, Justice and State, the Judiciary and Related Agencies for the Fiscal Year Ending Sept. 30, 1991, and for Other Purposes, Oct. 20, 1990. 99 See, for example, Kline, op. cit., p. 30.
PAGE 120
Chapter 4 U.S. Trade Policy
PAGE 121
CONTENTS Page INTRODUCTION AND SUMMARY . . 117 THE UNDERPINNINGS OF TRADE POLICY 118 The Rationale for Free or Liberal Trade . . 119 Refinements and New Trade Theory . . 119 Public Policy and Trade Theory . . . 121 THE U.S., TRADE POLICY APPARATUS . 124 FOREIGN MARKET BARRIERS . . . 125 Negotiating To Reduce Barriers . . . 127 Dispute Resolution Under GATT . . . 129 Section 301 . . . . . . . . 130 Japans Barriers l . . . . . . . 136 New Approaches . . . . . . . 137 SUBSIDIES AND DUMPING . . . . 138 Subsidies and Dumping: What They Are, Why They Matter . . . . . . 138 Countervailing Duty and Antidumping Cases 139 The Effectiveness of U.S. Countervailing Duty and Antidumping Law . . . . . 143 How Hard Are the Problems to Fix? . . 153 OTHER MEASURES AGAINST IMPORTS . 155 GOVERNMENT PROCUREMENT . . . 157 What the U.S. Public Sector Purchases . . 157 The U.S. Legal Regime.. . . . . . 159 The Effect of U.S. Procurement Law and Policy . . . . . . . . 160 Forcing Open Foreign Markets . . . . 162 EXPORT PROMOTION . . . . . 163 EXPORT CONTROLS . . . . . . 168 Reducing East-West Controls Dragon Competitiveness . . . . . . 170 Licensing Procedures . . . . . . 172 Boxes Box Page 4-A. GAIT and the Uruguary Round . . . 128 4-B. An Unfiled Dumping Case . . . . 146 4-C. East-West and North-South Controls; Cocom and Other Multilateral Export Control Forums . . . . . . . . 169 Figure s Figure Page 4-1. Import Penetration of the Japanese Semiconductor Market . . . . . 130 4-2. GATT Dispute Resolution Proceedings, 1947-86 . . . . . . . . 131 4-3. U.S. Trade Deficit With Japan and the World, 1983-1990 ..*..., . . . . .,.**.,. 136 4-4. New Countervailing Duty Cases Completed, January 1987 Through June 1990 . . . 140 4-5. New Antidumping Cases Completed, January 1987 Through June 1990 .. ., .. .+. . . 141 4-6. Decisions in U.S. Countervailing Duty Cases Completed January 1987 Through June 1990 . . . . . . . . 144 4-7.Decisions in U.S. Antidumping Cases Complete& January 1987 Through June 1990 . . . . . . . . 145 4-8. Government Purchases of All Goods and Durable Goods, 1990 . . . . . 158 4-9. Defense Purchases of Durable Goods, 1990, $83.6 Billion . . . . . . . 158 Tables Table Page 4-1. Some International Trade Functions of Some Key Agencies . . . . . . . 125 4-2. Summary of Some U.S. Laws Regarding International Trade . . . . . 126 4-3. Selected Section 301 Cases . . . . 132 4-4. New Countervailing Duty Cases Completed January l, 1987 Through June 30, 1990.... 141 4-5. New Antidumping Cases Completed Jan. l, 1987 Through June 30, 1990 . . 142 4-6. Outstanding Measures Against Subsidies as of June 30,1990 . . . . . 143 4-7. Outstanding Antidumping Measures as of June 30,1990 . . . . . 143 4-8. Section 201 Cases .. .. .. .. .. .., ... ..+..... 156 4-9. Purchases of Goods by the U.S. Public Sector in 1990 . . . . . . 157 4-10. Spending for Export Promotion in 1987..... 164
PAGE 122
Chapter 4 U.S. Trade Policy INTRODUCTION AND SUMMARY Many of the United States key trading partners, including the European Community (EC), Japan, Korea, and Taiwan, employ diverse tools to enhance manufacturing competitiveness. Tools like R&D support, infant industry protection, and favorable financing have been used to craft comprehensive trade and industry polices. On the whole, these policies have benefited Japan, Korea, and Taiwan, while posing a competitive challenge to the United States and the EC. The United States by and large has not imitated the proactive trade and industry policies of its trading partners. U.S. philosophy has been that manufacturing firms should make it on their own in the free market, with minimal help or interference from government. If this country were to modify its philosophy and seek ways in which government could enhance manufacturing competitiveness, two sets of policy tools could be used. The first set concerns the domestic economy. It includes improving education and training; using tax and fiscal policies to encourage long-term investment; costsharing commercially oriented R&D with industry; and promoting diffusion of best practice technology, especially to small and medium-sized manufacturers. Application could be uniform or selective--e.g., R&D support for certain technologies or industries. Many of these policies were discussed in Making Things Better 1 (also see ch. 2). The second group of policies, examined in this chapter, deal primarily with international trade. U.S. trade policy since World War II has normally sought no advantage for U.S. businesses beyond what they would obtain in a fully open world market. Rather, policy for the most part has tried to reduce, eliminate, or counteract foreign trade practices that distort the free market. The goal has been to ensure that U.S. companies can compete on an equal footing with foreign firms-or, as is sometimes said, to level the playing field. However, despite substantial progress in reducing quantitative restrictions on trade, this goal has not been fully met. Barriers against U.S. exports may persist for several years before the United States can get them removed. When foreign countries domestic policies confer advantages on their firms that result in dumped or subsidized U.S. imports, the United States can levy additional duties, called countervailing or antidumping duties, against the imports. These additional duties are intended to eliminate any foreign advantage; however, most often they do not. The failure of U.S. trade policy to meet its objectives is not so much a result of particular measures taken; the problem is that trade policy has been assigned a role that it cannot reasonably be expected to fulfill. To understand this point, it is necessary to look more closely at what other countries do. Some target their market barriers and domestic policies to promote specific industries, such as semiconductors and computers, that contribute disproportionately to a nations wealth and economic development, and on which U.S. manufacturing competitiveness depends. Some industries yield a high reward because they have increasing returns to scale and learnin g and spillover benefits to other industries. Developing such industries is often a race in which whoever gets ahead will likely stay ahead. A company with technical advantages or higher market share can reap greater economies of scale or learnin g, which will help it capture more market share and finance more R&D than its competitors, so that it can pull still further ahead. After World War II, the United States was the world leader in high-reward industries, and in a free and undistorted world market would likely have remained so. However, several other countries assisted their domestic firms, in part by protecting their home markets with tariffs, quotas, and other barriers. A protected home market can increase the domestic industrys market share and development while decreasing opportunities for foreign competitors. Home market protection can further enable the domestic industry to charge cartel prices, thus earning above-normal profits that can boost R&D programs. Profits in a protected home market can bankroll forays into export markets at low prices. Of course, home market protection can easily go astray, leading to an industry ill-suited for international competition; when managed properly in combination with other policies, however, it can aid a nations economic development.
PAGE 123
118 l Competing Economies: America, Europe, and the Pacific Rim In cases where home market protection has been successful, it has been as part of a more comprehensive strategy that included many domestic programs. This includes both society-wide programs-e. g., first-rate education, encouragement of household savings, and tax breaks for R&D and capital investmentand industry-specific programs, such as R&D projects, special tax breaks, preferential financing, and tolerance of cartel pricing in specific industries. By these means, some foreign governments have promoted selected industries to the point where they can earn large profits on their own, without need of government assistance (though assistance often continues). Other countries domestic programs and market protection have sometimes damaged U.S. industries. For the most part, U.S. trade policy plays out by noticing some of the advantages foreign firms enjoy, and then trying after-the-fact to eliminate or offset them, usually after substantial delay and often incompletely. Export policy focuses on foreign market barriers. Where foreign markets are closed to U.S. goods, the U.S. Government can try to get the barriers removed. However, not all barriers can be addressed, and many important ones have taken years to eliminate. Removal of specific barriers is an ineffective solution at best. It takes time and effort to identify barriers; there are so many that some will be missed. And once a barrier is identified, it takes time to negotiate its removal, if it can be removed at all. Import policy focuses on levying duties to counteract foreign subsidies and dumping. However, the laws approach of precisely compensating for observed subsidies and dumping has limited effectiveness. Identification and quantification of subsidies and dumping is a slow and laborious process. Once duties are in place, dumped or subsidized goods can come into the United States from a new country, requiring another investigation before duties can be assessed on the new goods. The formula to determine a duty that precisely neutralizes the foreign advantage does not fit the reality of modern industrial competition. Probably no formula could. Other aspects of U.S. trade policy are also less advantageous for U.S. companies than foreign counterparts policies. While many foreign governments procurement policies are attuned to fostering national industries, U.S. procurement policy is not. The Commerce Departments export promotion programs, while useful, are overshadowed in funding and effectiveness by similar programs in other countries. Export financing by the Export-Import Bank of the United States is sometimes less attractive than that offered by other countries export financing agencies. Finally, U.S. export controls for reasons of national security unduly hinder hightechnology exports; while many controls are necessary for national security, some are not. THE UNDERPINNINGS OF TRADE POLICY The conviction that free trade promotes social well-being is a cherished principle of economics and, not coincidentally, of postwar shapers of Americas trade policy. The arguments for free trade are powerful, and the prosperity that coincided with the GATT regime, which opened more nations to trade than at any other time in modern history, is usually regarded as convincing evidence of free trades benefits. Yet challenges to the free-trade regime continue and are gaining force. There is growing sentiment, especially among business managers, that domestic firms are unable to compete on an equal footing with foreign firms because of the support and protection foreign producers receive from their governments. Such sentiment is found not only in America but in Western Europe as well. According to one analysis, the quest of American industries for relief from import competition through special trade protection is driven partly by recent developments such as the overvalued dollar and high unemployment of the early 1980s. 2 In addition to temporary situations that make it difficult to export or compete with imports, companies in the United States and Europe are simply up against some very well-financed and technically sophisticated competitors from Japan and East Asia, many of which have government backing in various forms. The fear of loss of dominance--or even dissolution-is driving interest in various types of trade relief, ranging from outright protection to calls for different forms of managed trade. Is it time for a new guiding philosophy of trade? Few argue for widespread protection, but there is deep division between those who maintain that free trade invariably yields the greatest benefits and those who believe the time has come for some degree of managed trade. (Whether some degree of
PAGE 124
Chapter 4--U.S. Trade Policy l 119 managed trade is a realistic possibility is also in dispute. Critics see any management as the first step on the path to widespread protection.) Both free trade and managed trade have potential benefits and drawbacks. Even the option of a middle course, with modification of free trade that doesnt fit the nebulous term managed trade, would require a leap of faith, since evidence supporting any view of the philosophy of trade is equivocal. It should be stressed, as it often is not, that this is just as true of free trade as it is of other courses labeled industry policy, managed trade, and protectionism. The Rationale for Free or Liberal Trade Free trade is attractive for several reasons. One of the most powerful arguments in its favor is that a period of increasing prosperity has coincided with proponents would say it has been caused by international market opening. Many analyses attribute the twentyfold increase in trade in manufactured goods and the sevenfold increase in manufactured output to the accomplishments of the GATT. 3 The theoretical rationale for free trade is based on comparative advantage. Different products require differing amounts of labor, capital, or other resources to produce. Nations are endowed with these resources in different proportions. This combination of differences gives nations specific advantages in producing specific products. Nations like Canada and the United States, generously endowed with softwood forests, have advantages in producing construction lumber, plywood, and linerboard, for example. If nations specialize in producing what they make best, and there is free trade, everyone is better off than if every nation tries to produce the mix of goods it consumes by itself. This straightforward result only applies under certain conditions. These include perfect, or nearly perfect, competition, and no messy complications like barriers to entry (which make it difficult for new companies to enter a market), economies of scale or learnin g effects (which lower the costs of production as the volume of production or experience increase), and externalities (costs that producers do not take responsibility for, or benefits that are not available to them as profits). One of the earliest and simplest theories developed to explain why nations trade and the consequences thereof was developed by Eli Heckscher and Bertil Ohlin. The Heckscher-Ohlin theory conceived of a simplified world with two so-called factors of production (labor and capital), two products whose production requires these factors in different proportions, and two countries with differing abundance of labor and capital. The capital-rich country would have an advantage producing the capital-intensive good, and the country in which labor was plentiful would have an advantage in producing the labor-intensive good. Although in theory it is unlikely that either country would specialize in producing one good to the exclusion of the other, 4 the Heckscher-Ohlin model showed that if each exported what it was best suited to produce, the welfare of consumers in both countries would be greater than if there were no trade. In this view, tariffs, once viewed as the classic interference with free trade, are a net loss. While producers of the protected good in the home country may benefit from a tariff and the home country gets some revenue from it, the loss to consumers more than offsets these benefits. s Similarly, quotas can have negative results, as can the host of nontariff barriers that curtail free trade. This is the standard beginners guide to the economics of international trade. Like all models, it is built on a number of assumptions, all of which diverge from reality. The important question is how great a divergence exists between these assumptions and the real world. In some cases, the divergence is minor, and the result, free trade maximizing everyones welfare, is still valid. In other cases, the divergence is significant. Refinements and New Trade Theory Dividing factors of production into just two or three groups is oversimplified. After work in the 1950s by Wassily Leontief showed that the United States, thought to be a relatively capital-rich nation, was exporting goods that were more labor-intensive than those it was importing, a refinement was developed. That is, labor could be segregated by skill intensity, and the U.S. advantage was in producing goods and services especially intensive of skilled labor. Another oversimplification lay in using factor proportions as the only explanation of comparative advantage. Other factors can give a country an advantage in producing and exporting. Linder 6 postulated that the size and character of the home market was an important determinant of what a
PAGE 125
120 l Competing Economies: America, Europe, and the Pacific Rim country exported; after a product was developed for the domestic market, exports would begin when the domestic market was too saturated or too competitive to offer sufficient expansion or profit opportunities. Vernon and Wells explained trade flows through the product life cycle. The life cycle explanation argued that products would be developed and manufactured for the domestic market and later sold abroad. For the United States, products would be developed with labor-saving characteristics attractive to a relatively affluent customer. As other countries raised their standards of living, their markets would become ripe for American exports, at least until manufacturers in the foreign country learned how to make the product. Over time, as the technology matured and stabilized, countries with lower production costs (one example is lower labor costs) could master production, and the affluent country would lose its advantage and begin to import the product. Even with these refinements, trade theory was still far from reflecting real-world conditions. By the early 1980s, dissatisfaction with the theoretical basis of Americas trade policy was evident even within the economics profession. 7 Outside the profession, those concerned about the adequacy of the free-trade prescription and its theoretical basis became better informed and louder. The rise of Japan as an industrial power fueled the debate, for it was manifestly not a free trader, at least during the first three decades after the war, yet its industries were becoming more competitive and its standards of living higher. While economists may never reach a verdict on whether and how Japans trade policies and its economic prosperity are causally linked, Japan made everybody think more deeply about trade and trade theory. Developments in trade theory in the late 1970s and 1980s, often termed the new trade theory, have been summ ed up by saying .conventional trade theory views world trade as taking place entirely in goods like wheat; new trade theory sees it as being largely in goods like aircraft. In other words, new trade theory permits different outcomes and policy prescriptions for different kinds of goods. So what makes aircraft different from wheat, beyond the obvious physical and utilitarian characteristics? One of the most obvious differences is in the scale of production. While wheat farmin g is no longer just a matter of small family farms, there are thousands of wheat farmers, and none is large enough to exert much effect on the market. Aircraft, on the other hand, are produced by a handful of companies. There are two producers of large (100+ passenger seat) commercial aircraft in America and another in Europe, and three engine makers for large commercial jets. 9 There is a broad consensus that the world could not support additional entrants at the level of engine or aircraft assembler. There are more firms at the subassembly levels, but many expect those numbers to shrink. Increasing returns to scale accounts for this imposing structural difference between the two industries. What this means is that compared with wheat farming, making aircraf t requires a huge initial investment (typically $3 to $ 5 billion to design, develop, and certify a new model), but once it is made, producing one more airplane is relatively inexpensive ($50 to $150 million) .10 There are also increasing returns to scale, up to a point, in wheat farmin g. In all industries there is a range of increasing returns, and a point where any additional input (e.g., land, labor, capital, or knowledge, singly or in combination) begins to make the production enterprise less efficient. That point, where diminishing returns are gained as the scale of enterprise increases, is reached fairly quickly in wheat farmin g, so that the most economical scale of enterprise is small enough to allow many thousands of producers. Large commercial aircraft production is at the other end of the scale; increasing returns are still available to Boeing Commercial Aircraft Group, whose customer base is 60 percent of the world market and has been higher. While neither industry is typical, the patterns of world trade suggest that a greater amount of trade is in goods, like aircraft, that have increasing returns to scale over a wide range of production. If government assistance helps firms pull ahead of foreign competitors, scale economies can help keep them there. Another quality that sets aircraft apart from commodities like wheat is knowledge intensity. All products require some knowledge, at least for efficient high-quality production. But the knowledge embodied in a 747 is far greater than the knowledge embodied in a ton of wheat, and it is much more difficult to master the technology of designing and making an airplane than the techniques for growing wheat. The product cycle approach to explaining trade flows recognized the importance of technology, but it assumed a relatively fixed cycle in which technology is eventually
PAGE 126
Chapter 4--U.S. Trade Policy 121 diffused from the country of origin to producers in countries with other advantages, such as labor costs. For sophisticated and expensive technology such as in airplanes, diffusion could be slow, difficult, and expensive. Externalities complicate classic models of trade, too. Also called spillover effects, externalities are benefits or costs of a business or industry that are not captured or borne by the companies involved. Environmental pollution is a classic example of a negative externality; in most circumstances the generators of pollution have not had to bear the total damage or the entire burden of cleanup. One of the most frequently cited positive externalities is R&D. R&D done by one firm or industry often generates knowledge that other firms or industries can use without paying for it. Positive externalities translate into spillover effects that can benefit the society as a whole. Policies that foster growth of industries with positive externalities can create comparative advantage and produce higher growth in standards of living. New trade theory combines the ideas of increasing returns to scale and technological advantage, creating a virtuous circle. In industries with these increasing returns, a company with superior technology or greater market share can reap greater economies of scale or learning, which will help it capture additional market share and finance more R&D than its competitors, enabling it to pull still further ahead. In this way, a firm that starts out ahead can keep increasing its advantage, at least until other entrants that are well-funded and patient enough to weather long periods of loss and learnin g can make inroads as, for example, Airbus and the Japanese semiconductor producers have done. Public Policy and Trade Theory In industries with increasing returns to scale, significant positive externalities, and high knowledge intensity, there can be a sound economic case for departures from free trade, or protection. ll Economic models have postulated that it can benefit a nations balance of trade and standard of living to protect certain kinds of industries under certain circumstances, such as infant industry protection. 12 More cautious analysts suggest that policies to strengthen the research base (e.g., R&D tax credits) and encourage pooling of resources (e.g., policies to permit or encourage strategic alliances) would be useful, but that protection or subsidies would lead to a neighbor-beggaring world of retaliation. 13 This prescription may be more appropriate for technological leaders than for those facing dominant foreign competition. R&D incentives and encouragement for firms to pool resources may suffice to maintain competitiveness in a country with a well-developed technological base. However, without protection some domestic industries could face crushing competition from foreign firms whose technological superiority (perhaps attained through subsidies and protection by their own government) makes their production costs much lower. It is hard to prescribe policy for trade in the real world. Ever since World War II, the United States has maintained a strong interest and belief in free trade. As the principal architect and (until recently, perhaps) greatest proponent of the General Agreement on Tariffs and Trade (GATT), the United States regards itself as the worlds standard bearer for open markets and reasonable, unrestrained competition. While its market is still one of the worlds most open, the United States image as a staunch free trader is fading; as its industries become less competitive, the United States protects them more. The United States is only one culprit in what many regard as a watershedsome might even say a crisis-for GATT and the principle of free trade. Preeg cautions that [t]he GATT multilateral system is. .at the greatest risk of being overtaken by events. 14 What events? One is an increasing tendency for nations to negotiate quotas bilaterally or among trading blocs or customs unions. GATT has recorded over 200 quota arrangements that restrict industrialized countries imports in products such as textiles and apparel, steel, motor vehicles, semiconductors, machine tools, footwear, and consumer electronics. 15 These arrangements include the proliferation of voluntary restraint agreements (VRAs) that restrict trade between two nations. An example is the VRA between Japan and the United States in which Japan agreed to limit its exports of motor vehicles to the United States, from 1.76 million units in 1981 to 1.94 million units in 1985. 16 Another kind of restriction is represented by the Multifiber Arrangement (MFA), by which signatories agree to quotas on trade in textile and apparel products. Different countries and regions set their own quotas, which limit how fast imports in covered areas can grow. 17 Although done under GATT auspices and legal by 292-889 0 91 5 QL:3
PAGE 127
122 l Competing Economies: America, Europe, and the Pacific Rim GATT standards, the MFA represents an accommodation of GATT principles to reality rather than a change in GATT philosophy. Another possible threat to the GATT is regionalism. The GATT stresses nondiscrimination treating all trading partners the same and providing to all signatories most-favored-nation (MFN) status. 18 This principle may be eroding as more nations with geographic or other ties negotiate special trading arrangements. The recent U.S. Free Trade Agreements with Canada and Israel, and Europes Single Market Act are recent examples; more are possible, including a North American free trade agreement (FTA) and one between Australia and New Zealand. Not all these regional arrangements are inconsistent with the spirit of GATT: the U.S.-Canada FTA and the contemplated AustraliaNew Zealand FTA are viewed as building blocks for broader trade liberalization. 19 The European Community is a bit more of a puzzle. While many expect unification to open the market, the murky signs now available indicate that EC policies are more likely to open opportunities for investment in Europe than for exports to it. For some products, especially motor vehicles, the effect is to restrict trade with Japan. Finally, some analysts see industrial policies as challenges to flee trade. Nations use a variety of methods to support and nurture firms that are regarded as essential to national well-being. Nearly all use some means to assure domestic production of armaments and other goods vital to national security, and many promote the growth and development of economically strategic industries. The policies can range from outright protection for infant industries to policies whose protectionist intent is far more subtle and often unclear. For example, the laws and practices governing distribution of goods from wholesalers to retailers in Japan are often viewed as an effective form of protection against imports, as are public procurement of telecommunications equipment and services in Europe and Japan and U.S. national security requirements stipulating that some goods be purchased from domestic suppliers. Whether such policies are actually intended to protect domestic industries and companies is a topic for endless disagreement. But in fact, industrial policies often discourage imports, because nations often implement industrial policy to help their own industries catch up to the worlds leaders. Even policies designed to accelerate development in new technologies, such as high-definition television (HDTV), often have overtly protectionist overtones. The fact that the United States and the EC, both behind Japan in developing HDTV systems, are debating adoption of non-Japanese HDTV standards is an indication. 20 In short, no nation has an unblemished record of playing strictly by free trade rules. To the extent that the free-trade analyses of the gains from trade and penalties of protection are correct, increasing protectionism is a problem. There is a large class of goods-commodities or other unspecialized products -contributing about the average to value added, productivity, or knowledge, with relatively stable and widely available production technologies. This class probably includes most agricultural products and industrial goods such as toys, apparel and textiles, many chemicals, lumber and plywood, footwear, iron and steel, and a lot of industrial machinery. It is unlikely that significant sustained trade protection in these industries would benefit more than a narrow segment of society. This does not mean there is never a case for protection in such industries. Many nationsin fact, most industrialized nationsprotect their steel, apparel, textile, and lumber industries; protection of agriculture borders on legendary. Nations protect agriculture in part because of a deeply rooted conviction that it is unwise to depend on other nations for food, at least for the primary dietary staples. While most nations want to maintain friendly, or at least civil, relations with most others, none wants to be a hostage to the whims of another government in order to feed its citizens. This threat is remote, for little other than open aggression is likely to cause other nations to cut off food exports en masse (as most of the world has done recently to Iraq). But while a rational analysis might show that the vulnerability created by depending on imports for the bulk of ones food is not great, the penalty in the event of a worst-case incident is heavy; the popular consciousness often equates it with a loss of sovereignty. One reason to protect mature industries such as steel and textiles is to ease adjustment. The traditional economic model says little about mobility of labor and capital, assumin g that if an industry shrinks because its advantage declines then its workers and capital will migrate to industries with more advantage. Full employment is assumed, and the pain of the migration is mentioned only in
PAGE 128
Chapter 4--U.S. Trade Policy 123 passing, if at all. Adjustment rarely works this way, even when unemployment is low. Several things complicate the neat economic model. Both labor and capital may be less mobile than we would wish. Displaced workers, particularly the unskilled and semiskilled, typically undergo long periods of unemployment before finding new jobs, and most take a cut in pay and benefits when they do find new employment. According to the latest information from the Bureau of Labor Statistics, over half (55 percent) of workers displaced from Ml-time jobs had lower incomes when they were reemployed, most because their wages were lower in new full-time employment and some because their 21 Moreover, displacement new jobs were part-time. and reemployment often involves loss of important benefits; Podgursky and Swaim report that 29.4 percent of all reemployed blue-collar workers and 20.8 percent of reemployed white-collar workers had group health insurance in their old jobs but not in their new ones. 22 In the worst cases, dislocation has wider effects; downturns in industries have resulted in long-lasting economic disability for communities or regions. West Virginias economy never recovered from the downturn in the coal industry in the 1950s, despite attempts to mitigate its poverty and unemployment. Capital is not easily or painlessly mobile either. While it is not difficult to move money around, by shifting investments in a portfolio, for example, capital equipment may be specialized, and loss of competitiveness in an industry may result in scrapping of plant and equipment before their useful lives expire, or before they are fully depreciated. This kind of loss hardly compares to the trauma faced by workers and families confronting plant shutdowns or job losses, but it is a loss nonetheless. If the adjustment period is long enough, and everyone accepts the need for industry downsizing, much of the pain of adjustment can be mitigated or avoided. But if it is rapid, resulting in the loss of significant capacity in only a few years, there can be considerable upheaval for workers, families, businesses, and communities. Moreover, during contraction, it is often difficult to distinguish competitive subsectors from candidates for dissolution. For example, while we think of the apparel and textile industries as sectors in which America has no particular advantage, parts of both industries are still competitive. Industrial and household textiles do not face the same competitive pressure from imports as apparel textiles, and many of these companies can continue in business profitably. But many of the firms are integrated, producing both. If part of a companys business faces collapse, it becomes difficult for the other part to obtain the capital, workers, and other resources. While the beneficiaries of such protection are invariably many fewer than the number of people who pay, costs are limited if the protection is also limited in time and scope. Like the rationale for free trade, this often looks better on paper than in real life. In fact, governments seldom cut the lifeline that protection provides. Protection has continued for decades in textiles and apparel; there are few signs that the United States is prepared to end it, or that the industry could avoid further dislocations if all protection ended. Political considerations aside, there is widespread agreement that industries that do not contribute disproportionately to the national storehouse of knowledge and productivity and that do not have increasing returns to scale over a wide range should receive little if any protection. There is less agreement at the other end of the industrial spectrum, but a fair economic and a powerful political case can be made for including trade policy in the toolkit of measures to promote industrial development. Such protection has been used in countries that wished to jump-start their economic development. Japan and Korea, for example, have protected many developing industries and compiled a successful record in developing industrial competitiveness and technical competence in the rarefied atmosphere of domestic protection. 23 In the United States, however, consideration of infant-industry protection has been halfhearted at best, partly because of a strong conviction of the merits of free trade, and partly because high-reward industries developed and flourished here before anywhere else. The situation is different now. The United States retains great technological strengths, but U.S. manufacturing in general and many high-tech industries in particular are moving ahead less rapidly and surely than many of their best competitors, including Japanese and some European companies. In microelectronics, Japanese manufacturers dominate world markets and technology development in many products, starting with DRAM chips in the early 1980s. Japanese manufacturers
PAGE 129
124 Competing Economies: America, Europe, and the Pacific Rim have challenged the American lead in computers throughout the market, from laptop PCs to supercomputers, and few believe that they have reached their limit. After having pioneered scientific work in superconductivity (a technology that could contribute to many industrial products but is still mostly in the research stage), Americans and Europeans have watched Japanese companies take solid steps to incorporating superconducting materials in commercial products. And in high-resolution television, American companies have been mostly spectators in a game that involves European companies and governments struggling to catchup to the Japanese. Never before in modern history has the country been behind in developing high-technology industries, or more uncertain of its own ability to compete without greater government involvement. Those who see a place for infant-industry protection believe it should be limited. A recurring allegation is that countries continue infant industry protection long past the point of true vulnerability, which hurts not only foreign competitors but also domestic consumers and downstream industries. In principle, infant industry protection puts an industry on its feet, at which point it should fend for itself. That point is undefined, however, and we know from experience that life support systems are easier to start than to stop. Another problem, in the United States, at least, is that there is no political institution capable of or charged with identifying high-reward industries that need protection to start. Congress lacks the time and depth of knowledge to make such choices, and few trust existing public-sector institutions to choose wisely. If the United States does depart from its traditional policy to open trade in all circumstances, and pursues policies that proactively support U.S. industrial competitiveness, the country will likely need new government institutions and new relationships between government and the private sector (see ch. 2 for further discussion of this point). THE U.S. TRADE POLICY APPARATUS No one government agency has overall responsibility for U.S. trade policy. Table 4-1 lists the major players. International negotiations on trade issues are led by the Office of the United States Trade Representative (USTR), located within the Executive Office of the President. The USTRs main mission is to achieve greater access to foreign markets for U.S. goods, services, and investment. In setting negotiating priorities, the USTR consults with other agencies, Congress, and business. The USTR negotiates bilateral and multilateral treaties and investigates alleged foreign market barriers under Section 301 (and related sections) of the Trade Act of 1974, as amended. 24 The Commerce Department investigates allegations that imported goods are dumped or subsidized; conducts export promotion programs; and, with input from other agencies, interprets and enforces the export control laws. It studies competitive developments in U.S. and foreign industries. and advises the USTR and other agencies on the environment that U.S. industries need to stay competitive. The International Trade Commission (ITC) studies the health of U.S. industries and the effect of imports on them. The ITC evaluates the injury (or threat of injury) to U.S. industries that petition for protection from dumped or subsidized imports, or from imports under Section 201 (and related sections) of the Trade Act of 1974, as amended. 25 The Defense, Energy, State, and other departments advise the USTR, the Commerce Department, and other agencies about how trade policy could affect their interests. These departments assist the Commerce Department in interpreting and enforcing export controls, and the State Department leads negotiations regarding export controls. The Defense Department makes agreements with other countries regarding defense trade. The Export-Import Bank of the United States (Eximbank) helps companies get export financing. The Agriculture Department supports agricultural trade (including export promotion and export financing), administers price support programs, studies the business condition of U.S. agriculture, and acts as agricultures advocate within the governm ent. U.S. firms and industries can seek help from the government under various trade laws. Table 4-2 summarizes some of these laws, showing when they can apply, what agencies are involved, what relief can be obtained, and the usual costs to use the laws. The first law, Section 301, is concerned primarily with exports. Under Section 301, firms can complain about any act, policy, or practice by a foreign country that is unreasonable or dis criminatory and
PAGE 130
Chapter 4--U.S. Trade Policy 125 Table 4-l-Some International Trade Functions of Some Key Agencies Agency Function Office of the United States Trade Representative (USTR), in the Executive Office of the President Department of Commerce U.S. International Trade Commission (ITC) Department of Defense Department of State Department of Energy Export-Import Bank of the United States (Eximbank) Department of Agriculture Conducts most international negotiations, with assistance from other agencies (negotiations usually are aimed at increasing access to foreign markets). Investigates allegations of unfair foreign trade practices (most often foreign trade barriers) under Section 301. Investigates allegations of subsidies and dumping. Conducts export promotion programs. Has lead role in specifying and enforcing export controls. Studies and advises on competitiveness of particular industries. Determines whether U.S. industries are injured as required for relief under the laws regarding subsidies and dumping and under Section 201. Studies and advises on the competitiveness and health of particular U.S. industries, including the effect of imports. Assists Department of Commerce in specifying and enforcing export controls. Makes agreements with other countries regarding defense trade. Assists Department of Commerce in specifying and enforcing export controls; leads international negotiations regarding export controls. Assists Department of Commerce in specifying and enforcing export controls. Helps firms to obtain financing for exports. Conducts export promotion programs for agriculture. Leads many international negotiations regarding trade in agriculture. Collects and disseminates agricultural trade data. Helps business to obtain financing for agricultural exports. Administers price support programs for agriculture. SOURCE: Office of Technology Assessment, 1991. burdens or restricts United States commerce. 26 Usually the complaint concerns a foreign market barrier. Under Section 301, the U.S. Government can negotiate with the country concerned to remove the barrier. If the negotiations are unsuccessful, the United States can retaliate by imposing its own barriers against goods or services from that country. The other trade laws listed in table 4-2 concern imports. Two of these laws focus on subsidized and dumped imports. Subsidized imports are those for which the foreign producer or exporter receives a subsidy from its government. Dumped imports are those sold for less than fair value. U.S. firms can complain to the Commerce Department and the ITC about subsidized or dumped imports. If the Commerce Department finds subsidies or dumping, and if the ITC finds that the subsidized or dumped imports have caused or threatened material injury to the U.S. industry, additional duties (called countervailing or antidumping duties, respectively) normally will be levied on the imports. Finally, under Section 201 a U.S. industry can petition for relief from any increased imports that cause or threaten serious harm. The ITC determines whether the injury requirement is satisfied and, if so, recommends relief. However, the President ultimately decides what, if any, relief to grant. FOREIGN MARKET BARRIERS A major goal of U.S. trade policy is to reduce or eliminate obstacles to the sale of U.S. goods in foreign markets. However, despite U.S. efforts many barriers exist. While the United States itself limits some imports, the country is still one of the worlds most open markets, and on the whole U.S. exports
PAGE 131
Table 4-2-Summary of Some U.S. Laws Regarding International Trade Sections of Title 19 Agencies Action Nature of Approximate cost to Name of law U.S. Code responsible complained about relief granted petitioner in 1988 a Section 301 b 2411-2420 USTR investigates and de. . . . -. -_ -. termines any retaliation subject to presidential direction. Countervailing Duty Law 303, 1671 -1671h, Department of Commerce 675-1677k determines subsidies. ITC determines injury. Antidumping Law 1673-1677k Department of Commerce determines dumping. ITC determines injury. Section 201 b 2251-2254 ITC determines injury and recommends relief. President determines relief. Violations of a trade agreement, or other unreasonable foreign trade practices that restrain U.S. commerce (usually some form of foreign market barrier). Foreign subsidies on imported goods, injuring or threatening to injure a U.S. industry. Dumping of foreign goods in the U.S. market, injuring or threatening to injure a U.S. industry. Increased imports that cause or threaten serious injury to a U.S. industry. Negotiations to get foreign practice stopped, backed by threat of retaliation. $54,700-$305,400 Countervailing duties equal to the computed subsidy. Antidumping duties equal to the amount by which the goods are dumped. Protection in almost any form, and/or adjustment assistance. $138,100-$399,400 for initial investigation and ruling. $68,300-$106,700 for first court appeal (to Court of international Trade), if taken. $35,000-$50,000 for second court appeal (to Court of Appeals for the Federal Circuit), if taken. $43,900-$147,200 for each subsequent annual administrative review, if held. $151,000-$553,300 for initial investigation and ruling. $50,000-$83,000 for first court appeal (to Court of International Trade), if taken. $30,000-$37,500 for second court appeal (to Court of Appeals for the Federal Circuit), if taken. $42,000-$168,900 for each subsequent annual administrative review, if held. $202,000-$566,000 a~e ~mt ~~timat= are taken from a 1988 SUwey by the Genera] A~ufiing offi~ of ~ trade lawyers. (JrS. ~ngre~, General A~u~ing offi~, /~t~a~o~ Tfi: ~(./f$U/fofT~~e Re/?7deS by Srna//Business, NSIAD-69-69BR (Gaithersburg, MD: U.S. General Accounting Office, Dec. 1, 1968), pp. 6,6-11, 13,23. Each attorney was asked to give a lower and upper bound for the cost of each type of case. Each minimum value shown inthetable is the average of the minimum values given byeaoh attorney; similarty, each maximum value in the table is the average of the attorneys individual maximums. When an attorney in a particular instance gave an indeterminate lower or upper bound, that attorney was not included in the average. Because most indeterminate values tended to be large values for the upper bound, [the] average cost ranges tend to understate the actual oosts of pursuing the remedy. Ibid., p. 6. These cost estimates do not indude employee time devoted to the case. GAO noted (pp. 23-24) that its cost estimates were comparable to rough estimates by Professor John Jackson in Perspectives on the Jurisprudence of International Trade: Costs and Benefits of Legal Procedures in the United States, University of Michigan Law Review, April-May 1984, pp. 1570-1587. bof the Tr~e Act of 1974, pubii~ L93+18, as Amended. SOURCE: Office of Technology Assessment, 1991, and references cited.
PAGE 132
Chapter 4--U.S. Trade Policy l 127 face more obstacles than U.S. imports. As well as immediately worsening the U.S. balance of trade, foreign market protection can help foreign industries pull permanently ahead of their U.S. counterparts, often in high-reward sectors such as semiconductors and computers. Such barriers put the United States at risk of losing industries important for its wellbeing. Negotiating To Reduce Barriers Traditionally, the most important obstacles to international trade have been quotas (limits imposed by a country on how much or how many of a product may be imported into that country) and tariffs (taxes on imports levied by the importing country). In general, GATT has prohibited quotas for 40 years, 27 though there are some important exceptions. 28 Tariffs have been greatly reduced through a series of GATT negotiating rounds, with substantial leadership from the United States (see box 4-A). The United States has pursued lowering of tariffs by other means, such as the 1989 U.S.-Canada Free Trade Agreement, which, among other things, will gradually eliminate tariffs between these two countries. As quotas have been largely eliminated and tariffs greatly reduced, other barriers have assumed more importance. These include burdensome and slow customs procedures, limitations on the activities of foreign-owned fins, discriminatory regulations, government procurement practices that favor domestic firms, and the inclination not to buy foreign products. These barriers can be overt or hidden, and can result from government action, private action, or a combination of the two. Collectively, these obstacles are sometimes called nontariff barriers. 29 Nontariff barriers are often tightly bound up with domestic policies and are therefore harder to remove than tariffs or quotas. Despite its success in reducing tariffs and removing quotas, GATT has had more difficulty in reducing nontariff barriers. Accordingly, the United States has looked beyond GATT to eliminate these impediments. Each year the United States conducts many bilateral negotiations to eliminate particular nontariff barriers. 30 The USTR sets negotiating priorities primarily based on cases political importance, as communicated by industry, Congress, the press, and other executive departments. The USTRs negotiating priorities generally have not been based on an assessment of which industries have the most strategic importance for the nation. These bilateral negotiations have often succeeded in reducing or eliminating market barriers. However, often many years elapse without results. One example concerns how Taiwan assessed the value of imports for the purpose of levying percentage rate tariffs. Taiwan determin ed the value of imported goods by administrative rulings rather than by using the invoice price. By its nature, this approach tempted government officials to set values high. In 1979, Taiwan, in an agreement with the United States, obliged itself to switch over to valuing goods based on the invoice price by January 1, 1986. However, by August 1, 1986, Taiwan had not done so 31 Another example involves semiconductors. Despite several agreements by Japan to remove market barriers to semiconductor imports, the U.S. producers share of the Japanese market stayed at about 10 percent from 1973 to 1988. 32 This happened even though U.S. chips were competitive (especially during the first half of the period), as shown by the large market shares U.S. companies had in the rest of the world. 33 In September 1986, to settle a Section 301 case brought by the Semiconductor Industry Association, Japan promised to make certain efforts to increase market access. The Japanese Government professed that it strongly support[ed] expanded trade, stated that Japanese firms anticipated substantially increased purchases of foreign semiconductors, and agreed that the expected improvement in access should be gradual and steady over the agreements 5-year term. 34 During the negotiations, the United States argued that if Japans market were truly open, non-Japanese firms would capture at least 20 percent of it. In a confidential side letter to the U.S. negotiators, the Japanese Government stated that it understood, welcomed, and would make every effort to assist reaching the goal o f 20-percent import penetration by August 1991, when the agreement expired. 35 However, in the months after the agreement, despite a falling dollar, U.S. sales rose little. On March 31, 1987, the U.S. Government found Japans efforts to increase market access inadequate, 36 and on April 17 President Reagan levied punitive tariffs on certain personal computers and power hand tools from Japan. 37 In response, MITI encouraged Japanese firms to purchase more U.S. semicon-
PAGE 133
128 l Competing Economies: America, Europe, and the Pacific Rim Box 4-A-GATT and the Uruguay Round 1 GATT (the General Agreement on Thrills and Trade) is both an international agreement and an international organization. Its origin traces to the 1944 Bretton Woods Conference, which launched the World Bank (International Bank for Reconstruction and Development) and the International Monetary Fund. While the conferees did not consider trade issues as such, they went on record as recognizing the need for a third international organization to deal with trade. A charter for such an organization, called the International Trade Organization (IT0), was drafted and refined, under U.N. auspices, starting in 1946, culminating in the Havana Charter in 1948. However, the U.S. Congress would not ratify the Havana Charter; as the United States was then the worlds preeminent economic power, other countries declined to form an ITO without the United States. However, the President could ratify on his own, within a previous delegation of authority by Congress, a watered down agreement called the General Agreement on Tariffs and Trade (GATT. The difference was that GATT on its surface was just an agreement; it did not explicitly call for the creation of an international organization. For example, GATTs language does not provide for a general assembly or standing committees, but merely refers to the contracting parties acting in concert. Originally intended as a temporary measure until the United States endorsed an ITO, GAIT has provided the legal framework for trade among nations for over 40 years. In fact, the member countries did form an organization to carry out GATTs business. However, GATT has the flavor of a loose, hastily arranged outfit, reminiscent of the Articles of Confederation that preceded the U.S. Constitution. A key weakness is GATT's set of procedures for resolving disputes under its provisions, in which any country can, in effect, veto any ruling against it. GAITs primary mission has been reduction of trade barriers. With some exceptions, GATT immediately prohibited quotas. 2 From 1947 to 1979, GATT reduced tariffs in negotiating rounds through which countries would negotiate reciprocal lowering of tariffs. Each country, in return for concessions by others, would bind its tariffs for specified goods at certain rates. Each tariff concession granted to one country had to be granted to all member countries. Tariffs were greatly reduced by this process, leaving in 1979 an average tariff of only 4.7 percent for nonprimary products (products other than ores, timber, and the like) imported into industrial countries. 3 l~s ~x tiws h~vily from Jolm H. Jacb& The World Truding System: Luw and Policy of InternationaZRe2ations (Cardnhlge, MA: MIT Press, 1989), pp. 27-57. %AIT Articles XI-XIV, XIX-XXI. 3Johu H. Jaclcso~ op. cit., p. 53. ductors and to develop long-term relationships with one in 1976. The Japanese Government did not U.S. suppliers similar to those with Japanese suppliers. As a result, U.S. market share rose from 10.2 percent in 1988 to 11.0 percent in 1989 and 12.3 percent in 1990 (figure 4-1). 38 The total imports market share was larger, 11.9 percent in 1989 and 13.2 percent in 1990. 39 The first quarter of 1991 shows no further increase. 40 This was still far behind the progress envisioned by the U.S. negotiators in the 1986 agreement: steady growth to 20 percent by August 1991. A new agreement, in effect from August 1991, has set a target of 20 percent by the end of 1992. 41 Supercomputers provide yet another example. Since 1976, the Japanese Government heavily favored Japanese firms in its purchases of supercomputers, buying almost no U.S. machines even though they were generally superior. 42 Cray Research, Inc., the first firm to sell supercomputers, sold the first purchase supercomputers until 1983, when Japanese supercomputers were first available. Negotiations began in 1985; in August 1987 Japan agreed to make its procurement process more open to foreign firms, through, for example, improved announcement of 43 However, the Japanese upcoming procurements. Government continued to favor Japanese firms and bought very few U.S. machines. The government did this through a variety of means, such as a method of evaluating performance structured to favor Japanese machines, 44 In June 1989, the USTR lauched a Section 301 investigation, with the threat of retaliation if the Japanese Governments practice did not change. In June 1990, a second agreement was reached addressing performance measures and several other procurement issues. 45 This agreement may have come too late to be of much help to the U.S. supercomputer industry (see ch. 6).
PAGE 134
Chapter 4--U.S. Trade Policy 129 In recent years GATT has focused on nontariff barriers, which have become more important as quotas have been eliminated and tariffs lowered To some extent, GATT addressed nontariff barriers from the beginning. For example, GATTs national treatment provision requires that a countrys internal taxes and regulatory requirements not discrimin ate on the basis of whether a good is domestically produced or imported. 4 This provision prohibits explicit disc rimination against imports, but leaves open the crafting of laws and regulations that more subtly discourage imports. Optional supplementary agreements called GATT Codes further address nontariff barriers. A Code is effective only between two members who have both signed it. The United States and its major trading partners have signed several important Codes, including those on subsidies, dumping, and government procurement. While GATT and its Codes have eliminated some nontariff barriers, they have not succeded to the same extent as the effort to reduce quotas and tariffs. Nontariff barriers have been an important focus of the latest GATT negotiating round, the Uruguay Round, which began in 1986 and was scheduled for completion in 1990, but has been extended through 1992. Canada and the EC have proposed replacing GATT with an ITO, but that proposal will be considered only after the Uruguay Round is completed. The United States priorities in the Uruguay Round negotiations have included: 1. extension of GATT to require countries to afford certain minimum levels of protection for intellectual property (patents, copyrights, and the like); 2. extension of GATT to cover investment; 3. extension of GATT to cover services; 4. a stronger dispute settlement mechanism; 5. a stronger legal regime to minimize subsidies and dumping; 6. tighter limits on exceptions to GATT's requirements granted to developing countries; and 7. strengthening of GATT's coverage of trade in agricultural products. This last item, dealing with agriculture, is the top priority of many developing nations; progress on that item is a prerequisite for their consideration of other items. 4GATT Article III. There are many reasons why it can take a long panel of experts hears the case. If the panel finds a time to open a foreign market. First, a particular barrier must be identified and verified, and negotiations begun. Even these first steps are sigificant hurdles, given the prevalence of protection and the lean staffing of the USTRs office. Foreign countries can prolong negotiations, demand long transition periods to phase out barriers, and drag their feet on promised changes. Also, when one trade barrier is removed, another might take its place. Ultimately, success in opening foreign markets depends not only on the U.S. Governments diligence but also on its leverage with foreign countries. The United States has two main sources of leverage: dispute resolution under GATT 46 and the threat of retaliation under Section 301. Neither is very effective in promptly opening foreign markets. Dispute Resolution Under GATT Under GATT dispute resolution procedures, an aggrieved country must first try bilateral consultations; 47 if these are unsuccessful, an international violation and the GATT Council formally adopts that decision, the offending country must either stop the practice or offer compensation (typically a lessening of barriers on other products); if it does neither, the GATT Council can authorize the aggrieved country to retaliate by erecting protection of equivalent importance against the offending country. 48 This route has been well traveled. According to one tabulation, from 1947 to 1986, 233 cases came before GATT. 49 Most of these cases involved industrialized countries (figure 4-2). The United States brought 77 of these cases, with 37 directed at agriculture and 11 at manufactured goods .50 It was often difficult and time-consumin g to get and enforce GATT rulings, with the process typically taking a few years and some cases never being resolved. 51 In the Uruguay Round, the United States has already achieved some success in reducing the time involved. An interim agreement, effective May 1989, made it possible to hear a dispute and get a
PAGE 135
130 Competing Economies: America, Europe, and the Pacific Rim Figure 4-1import Penetration of the Japanese Semiconductor Market 15 10 5 1982 1983 1984 1985 1986 1987 1988 1989 199 0 = United States m Other countries SOURCE: Telefaxed data from the Semiconductor Industry Association, May 29, 1991. decision in about 10 months from the time consultations are requested. However, any country can still block the GATT Council from adopting an adverse decision as an official ruling; the defendant thus has veto power over the tribunal. And even if the GATT Council adopts the decision, a party found in violation can refuse to change or pay compensation, and can even veto any proposed retaliation by the injured country. While the pressure of international opinion can induce a country to accept GATT rulings, it still might stall for months or years. The United States has done its share of resisting GATT rulings, but nonetheless favors putting dispute settlement on a stronger institutional footing, so that all countries would be bound alike. However, it could well be many years before GATT dispute resolutions procedures provide for speedy, effective removal of market barriers. As long as these procedures are unreliable and while GATTs jurisdiction omits items of concern to the United States, such as services and intellectual property, there will be need for an alternative source of leverage, the threat of retaliation under Section 301. Section 301 Section 301 (and related sections) of the Trade Act of 1974, as amended, 52 can address any act, policy, or practice by a foreign country that is unreasonable or discrimin atory and burdens or restricts United States commerce. 53 In 1988, Congress clarified that Section 301 covers at least some government tolerance of private action, specifically the toleration by a foreign government of systematic anticompetitive activities by private firms or among private firms in the foreign country that have the effect of restricting access of United States goods to purchasing by such firms for reasons other than commercial considerations such as product quality, cost, and service. 54 Section 301 empowers the USTR to investigate allegations of such foreign practices. If they are found, the USTR will frost negotiate to stop them. If initial negotiations fail, Section 301 requires that GATTs dispute resolution process be invoked where applicable. 55 As a last resort, if all other measures fail, the USTR may retaliate. Retaliation might consist of increased tariffs or restrictions on specified goods imported from the offending country, whether or not the goods are related to the foreign practice. 56 From Section 301s enactment in 1974 through 1990, 83 Section 301 cases were initiated. Of these, 35 focused purely on raw and processed agricultural products; 37 in whole or in part on manufactured goods (excluding processed agricultural products and including cases concerning intellectual property); 10 on services; and 1 on investment. 57 Section 301 has been an important trade tool. Its use has prompted specific reforms by other countries; more generally, the threat of Section 301 retaliation for areas not covered by GATT, such as intellectual property and services, has made other nations more receptive to expanding GATTs coverage to those areas. 58 However, Section 301 generally has not achieved prompt removal of market barriers to manufactured goods. Table 4-3 lists all Section 301 cases that started in 1985 or later, concerned manufactured goods (excluding food, beverages, and tobacco), 59 and contained a finding by the USTR that there was a foreign market barrier that needed fixing. 60 Most of these cases have taken many years from the first notice of the problem to achievement of effective relief. Part of this delay comes from the time limits of the formal Section 301 case. The USTR has 1 year to investigate the alleged practice and negotiate for its elimination (up to 18 months if needed to accommodate a GATT dispute resolution procedure 61 ) and 7 months before deciding on and applying sanctions. 62 The USTR often takes the full time allotted, and
PAGE 136
Chapter 4--U.S. Trade Policy l 131 Figure 4-2-GAIT Dispute Resolution Proceedings, 1947-86 Country type: Developed Advanced developing Less developed Nonmarket economies Major filers: United States EC 1 1 1 1 1 1 1 I 1 1 200 150 100 50 0 50 100 150 20 0 Number of cases m Filed against Filed by SOURCE: John Jackson, The World Trading System: Law and Polbvof/ntemationa/ Economic Re/atkms Cambridae. MA: MIT Press, 1989), p. 99. foreign countries seldom make concessions unless under an imminent deadline with the threat of retaliation. Overall, however, the longest delays are in initiating an action. Most cases start with a petition from industry for relief. a3 It can be years before a U.S. firm or industry association is convinced that attempts to sell in a particular country are failing not because of cost or quality, but because of market barriers. Even then, a firm or industry association might balk at preparing a petition, given the expense (table 4-2), delay, and uncertain results. The USTR has initiated cases on its own, especially in recent years, 64 but those actions mostly involved longstanding problems that attracted considerable congressional attention. For example, under the socalled Super 301 provision, the Administration was directed to identify the highest priority offending countries and initiate Section 301 cases on their most important practices. 65 Cases 74 and 75 in table 4-3, dealing with the Japanese Governments procurement of satellites and supercomputers respectively, stemmed from this process. Delay can occur after a case is supposedly over. Case No. 52 in table 4-3 concerned Koreas intellectual property protection. Although Korea passed remedial legislation in 1987, many problems remained into 1989. These long delays arise because the United States is reluctant to retaliate or even start a Section 301 case. Retaliation, unless it ultimately induces the foreign country to change its practice, rarely solves the problems of the complaining U.S. industry. 66 Retaliatory restrictions on the import of goods from another country would in most cases violate GATT, and the open nature of Section 301 proceedings would instantly expose this violation to the world. a7 And while use of Section 301 does not violate GATT until retaliation occurs, starting an investigation does anger other countries, which perceive Section 301 as an assertion that U.S. law takes precedence over GATT. Such criticism might be somewhat hypocritical: if the target country maintains market barriers, overt or hidden, that could violate at least the spirit of GATT. However, this criticism is a political force to be reckoned with. Retaliation against a foreign trade barrier would itself be a trade barrier, and could provoke counterretaliation with still more market closures. For this reason sanctions, even as retaliation, are philosophically repugnant to the USTR and the Administration, for whom free trade is a paramount goal. On a
PAGE 137
Table 4-3-Selected Section 301 Cases Date of Effective date Case petition or and nature Subsequent Wait for No. Country Subject Prior history self-initiation of relief history relief 48 Japan Semiconductors: issues including access to Japanese market. l 49 Brazil Policies toward the informatics industry (computer and computerrelated products), includding import restrictions and inadequate intellectual property protection. Many sets of negotiations on market access since about 1972; U.S. firms market share stayed roughly constant at only 10 percent (far below what U.S. firms would have achieved in an open market) despite several ostensible liberalizing moves by the Japanese Government. Negotiations since 1984. June 14, 1985 July 31, 1986: Japan signed an agreement expressing the expectation of gradual, steady im provement in access; in a secret side letter, Japan acknowledged the U.S. goal that import penetration of Japanese market would rise over 5 years to 20 percent. (Apr. 17, 1987: United States retaliated for Japans failure to improve market access, by increasing duties on certain computers and power hand tools from Japan.) Serious Japanese Government effort to improve access commenced in 1989, leading to first substantial increase in import share of Japanese market, which was 13 percent in 1990 (improved from 10 percent, but much lower than expected from agreement). New agreement signed June 4, 1991, effective Aug. 1,1991, with target of 20-percent import penetration by the end of 1992. Sept. 16, 1985 Oct. 6, 1986: Brazil unsubstantial import dertook administrative rerestrictions continued forms designed to reinto at least Februduce import restrictions. ary 1991. 1987: Brazil passed remedial copyright legislation to protect computer software. Negotiations still ongoing About 18 years for some improvement in market share; goal of much greater market share not achieved after about 19 years. About 2 years to start removing import restrictions, with substantial restrictions lasting at Ieast 6-7 years; about 3 years to fix intellectual property protection. as of February 1991.
PAGE 138
Table 4-3-Selected Section 301 Cases-Continued Date of Effective date Case petition or and nature Subsequent Wait for No. Country Subject Prior history self-initiation of relief history relief l 52 Korea Inadequate protection of intellectual property rights, including patent protection and copyright protection for software. l 56 Taiwan Customs valuation system. 61 Brazil Lack of patent protection for pharmaceuticals. Extensive bilateral negotiations starting in March 1983 (patent) and in 1984 (copyright). In 1979 Taiwan agreed that by Jan. 1, 1986, it would value imports, for customs purposes, based on the invoice price, rather than determining the value by administrative rules. By Aug. 1, 1986, Taiwan still had not switched over. Negotiations since 1985. 68 Argentina Pharmaceuticals: inadeNegotiations since 1985. quate patent protection, and discriminatory registration requirements. N OV 4, 1985 July 1, 1987: Korea Many problems repassed remedial patent mained into 1989. and copyright legislation. Aug. 1, 1986 Oct. 1, 1986: Taiwan changed its customs valuation system to use invoice prices. June 11, 1987 (Oct. 20, 1988: United States retaliated by increasing duties on some imports from Brazil.) June 26, 1990: Brazils President announced that he would propose remedial legislation to Brazils legislature by Mar. 20, 1991. (July 2, 1990: United States removed its retaliatory duties.) Remedial legislation not passed as of spring, 1991. Aug. 10, 1988 Fall of 1989: Argentina No remedial legislaagreed to introduce retion enacted as of medial legislation by fall Mar. 5, 1991. of 1991. The petitioner then withdrew its petition. About 3-4 years to start improving intellectual property protection; many problems remained for about 5-6 years. About 7 years. At least about 6 years. At least about 6 years.
PAGE 139
Table 4-3-Selected Section 301 Cases-Continued Date of Effective date Case petition or and nature Subsequent Wait for No. Country Subject Prior history self-initiation of relief history relief 70 EC Export restrictions on cop per scrap, which allegedly depressed the price of scrap within the EC and raised it elsewhere, thereby giving an advantage to EC brass fabricators. Overall import licensing scheme, including prohib ited items, quotas, and lack of transparency. l 74 Japan Discriminatory government procurement of satellites. *73 Brazil l 75 Japan Discriminatory government procurement of supercomputers. Negotiations since 1984 or earlier. In 1983, Japan published its Long Range Vision on Space Development, in which Japan banned all public and private purchases of foreign satellites. The U.S. Government immediately initiated negotiations. Japan agreed to remove the ban on private purchases but held fast to its ban on public purchases. Discriminatory behavior since 1976, when U.S. machines were first available. Negotiations since 1985, including an agreement signed in 1987, which proved to be ineffective. ---NOV. 14, 1988 EC agreed to remove export restrictions for at least 1990. June 16, 1989 May 14, 1990: import regulation thoroughly changed, removing objectionable features. June 16, 1989 June 15, 1990: Japan signed an agreement for open procurement practices, except for R&D satellites. June 16, 1989 Mar. 22, 1990: Japan signed an agreement for open and fair procurement practices. As of May 1991, it was too early to tell if this agreement would stop Japans discrimination in procurement. 1 year, 2 months About 6 years About 7 years At least about 5 years since negotiations commenced; at least about 14 years since Japanese discrimination began.
PAGE 140
Table 4-3-Selected Section 301 Cases-Continued Date of Effective date Case petition or and nature Subsequent Wait for No. Country Subject Prior history self-initiation of relief history relief 79 Norway Discriminatory governJuly 11, 1989 Apr. 26, 1990: Norway 9 months ment procurement of took measures designed electronic toil collection to ensure fair considerequipment. ation of the petitioner in future procurements. Denotes cases initiated without a petition (self-initiated by USTR). NOTE: The prior and subsequent histories given for cases were found from readily available sources. Further research might reveal in some cases that the problem started earlier or lasted later. If so, the waiting times for relief in those cases would be longer than those stated. SOURCE: This table relies almost entirely on published USTR sources: Office of the United States Trade Representative, Section 301 Table of Cases, computer printout dated Jan. 17,1991, and Federal Register notices cited therein; Office of the United States Trade Representative, National Trade Estimate Reporf on Foreign Barriers, annual reports for 19851987, 1989-1991; Office of the United States Trade Representative, Report to Congress on Section 301 Developments Required by Section 309(a) (3)of the Trade Act of 1974, semiannual reports covering January 1989throughJune 1990; Office of the United States Trade Respresentative, United States, Japan Conclude SemiconductorAgreement, press release, No. 91-21, June 4,1891 (for Case No. 46); Donald Phillips, Assistant United States Trade Representative, testimony at hearings before the House Committee on Energy and Commerce, Subcommittee on Commerce, Consumer Protection and Competitiveness, May9, 1990, Serial No. 101-149, pp. 6-11 (for Case No. 48). Other sources were relied on for particular cases: discussion in this Chapter, and also Clyde V. Prestowitz, Jr., Tradjng P/aces; How WeA//owedJqoan To T*e The Lead(New York, NY: Basic Books, 1988), p. 62 (Case No. 46, prior history); USTR Official, personal communication, Feb. 27,1991 (Case No. 49, subsequent history into 1991); USTR Official, personal communication, Mar. 5,1991 (Case No. 61, subsequent history); USTR Official, personal communication, Mar. 5,1991 (Case No.68, events after initiation of investigation); chapter6 of this report (Case No. 75, showing that Japans discriminatory procurement ofsupercomputers started in 1976); Jonathan Streeter, Supermmputer Industry Analyst, International Trade Administration, U.S. Department of Commerce, personal communication, May 21,1991 (Case No. 75, subsequent history).
PAGE 141
136 l Competing Economies: America, Europe, and the Pacific Rim pragmatic level, sanctions could hurt downstream users of the affected products. For example, when Japan violated the 1986 semiconductor agreement and USTR drew up a proposed list of Japanese products on which to increase tariffs in retaliation, many U.S. firms that purchased items on the proposed list testified and lobbied against the action. 68 Sanctions can also hurt U.S. firms with operations or joint venture partners concerned. Japans Barriers With the limited effectiveness of in the country GATT dispute resolution procedures and Section 301, foreign market barriers often persist for years. This is especially true with Japan. In 1990, the United States had a $42 billion bilateral trade deficit with Japan, accounting for about 40 percent of the United States total trade deficit (figure 4-3). This deficit has many causes, including often superior Japanese products, U.S. dissavings and high Japanese savings, and inadequate efforts by some U.S. firms to sell to Japan. However, one important cause is a rich network of market barriers, including official government action and regulation, unofficial advice from government to industry, government toleration or encouragement of anticompetitive activity, and a business culture in which Japanese companies maintain long-term relationships and prefer to deal with other Japanese businesses-.@ Figure 4-3-U.S. Trade Deficit With Japan and the World, 1983-1990 Billions of 1990 dollars $200 I I mltl ----l $100 $50 $0 1983 1984 1985 1986 1987 1988 1989 1990 m Deficit with Japan m Deficit with the rest of the world NOTE: Amounts were converted to 1990 dollars using GNP deflators from the Department of Commerce, Bureau of Economic Analysis. SOURCE: International Monetary Fund, DirwtionofTra&Statisti@, 1990, pp. 402-403, and U.S. Department of Commerce, Bureau of Economic Analysis, Survey of Current Business, vol. 71, No. 5, May 1991, pp. S-16, S-17. Some have likened Japans trade obstacles to an onion: when one layer is peeled away, another is waiting below. This happened with semiconductors and supercomputers, as discussed above. It occurred from 1980 to 1983, when U.S. producers of aluminum bats tried to have their bats certified for use in Japan by the government-run childrens baseball league. First certification was flatly denied, even though that went against an agreement signed by Japan in 1979 as part of GATTs Tokyo Round. Then certification was offered but conditioned on meeting dis criminatory standards requiring an alloy seldom used in the United States and abase plug not used in U.S.-made bats. Then the standards were changed, but U.S. factories, because they were outside Japan, could not be qualified as meeting the standards, so that each lot of bats would have to be opened and inspected individually. Finally the inspection issue was resolved, but Japanese distributors refused to carry U.S.-made bats, effectively shutting U.S. producers out of the market. Each step required high-level U.S. Government involvement. 70 As another example, in 1989 a well-known American manufacturer approached the USTR with a fully documented problem. Two Japanese trade journals would not run its ads because of pressure from the firms Japanese competitors. Glen Fukushima of the USTR staff raised this case informally with a contact in MITI, and the problem was quickly resolved. Mr. Fukushima commented that, While the outcome of this case is positive, it is troubling to realize that it required the intervention of USTR and MITI, potential congressional involvement, and 1 year of frustrated effort for a major, well-established, and well-endowed American company even to place an ad in two Japanese trade journals-several stages removed from actually making a sale to Japanese users. 71 The difficult task of opening Japanese markets is made worse by the scant U.S. Government resources devoted to the problem. The equivalent of only about 15 full time staff at the USTRs office work on trade with Japan. 72 While USTR is assisted by other agencies, U.S. negotiating teams have often been far smaller and less well prepared than Japanese teams .73 The U.S. agencies have faster staff turnover than their Japanese counterparts, especially at the most senior (political) levels. This has caused the United States to repeat mistakes and has made interagency coordination more difficult. 74
PAGE 142
Chapter 4--U.S. Trade Policy l 137 New Approaches Difficulties in removing Japanese trade protection have led the U.S. Government to consider alternative approaches. In 1986, the United States tried a new approach, putting forward its expectation that the foreign share of Japans semiconductor market would rise to 20 percent over 5 years. Rather than trying to identify every market barrier, a seemingly hopeless task, the United States in effect asked Japan to identify and remove barriers on its own as needed to reach the target import level. While the 20-percent goal has not been achieved, the 1986 U.S.-Japan semiconductor agreement, together with retaliation for Japans breach of it, did improve U.S. market share after more than a decade of failed attempts. This results-oriented approach was endorsed in February 1989 by the top-level industry advisory group to the USTR, 75 but apparently has not been used in other cases. Another approach has been to try to remove many trade barriers at once in a given sector. This approach was used in five sectorstelecommunications, electronics, forest products, medical equipment and pharmaceuticals, and autos and auto parts in the Market-Oriented Sector-Selective (MOSS) talks, begun in 1984. Overall, the talks focused high-level attention on certain market obstacles, reduced them, and opened communication channels to resolve follow-up disputes. The talks seem to have helped U.S. firms increase their exports to Japan, though many companies reported continuing problems. The most substantial increases in exports occurred in telecommunications and in medical equipment and pharmaceuticals, for reasons including easily identifiable barriers that the Japanese Government could readily fix; strong U.S. presence already in Japan poised to take advantage of market liberalization; some industry support within Japan for market liberalization; and negotiators ability and teamwork. Subsequent monitoring by the U.S. Government was crucial to translating the agreements into tangible trade gains. 76 For example, because of the MOSS talks Japan made it easier to certify U.S.-made cellular telephones for use in Japan, but follow-up U.S. Government intervention was necessary to get permission for Motorola to serve the Tokyo cellular market. 77 Yet another approach considers the whole economy. The Structural Impediments Initiative (SII), launched by President Bush and Prime Minister Uno in July 1989, sought to identify and solve structural problems in both countries that stand as impediments to trade and to balance of payments adjustment. 78 In a June 1990 joint report, each side resolved to modify economic practices responsible for much of the bilateral trade imbalance. Japan pledged to decrease its trade surplus, for example, by increasing spending on public works projects, liberalizing restrictions on large retail stores, and enforcing its antimonopoly laws. The United States stated that to decrease its trade deficit it would, among other things, encourage more consumer savings; reduce the budget deficit; support commercially relevant R&D; reform various areas of Federal or State regulation, such as antitrust, product liability, and export controls; improve export promotion efforts; and improve education and work force training. 79 The countries agreed to seven follow-up meetings in the next 3 years, at the level of deputy or assistant cabinet officer and vice or deputy minister, to review progress, discuss problems, and produce annual joint reports. 80 The U.S. commitments stopped far short of the fundamental and wide-reaching changes necessary to reverse the competitive decline. In May 1991, the Administration noted that Japan had made welcome progress in a number of areas during the SII agreements frost year, though additional progress in all areas is necessary in order to contribute further to the goals of opening markets, reducing trade and current account imbalances, and improving the quality of life in Japan. 81 Most observers outside the U.S. Government are less optimistic. Some commentators believe that the SII talks will not substantially open Japanese markets, and that the only approach that will work is to demand specific market shares, as was done for semiconductors. 82 The U.S. Government has also used a systematic approach to deal with the ECs evolving trade and industry policies. An interagency task force, led by USTR, was formed in early 1988 to study EC-92 and help avoid a Fortress Europe with strong trade barriers to outsiders. The task force has 12 working groups on specific issues, such as technical standards, quantitative restrictions, investment, and rules of origin. 83 It has established U.S. Government positions, developed relationships with EC officials, and negotiated to influence EC legislation before it is agreed to at the EC level. The task force has influenced EC policies in areas including standards-
PAGE 143
138 l Competing Economies: America, Europe, and the Pacific Rim setting procedures, reciprocity provisions, and rules of origin. SUBSIDIES AND DUMPING Subsidies and Dumping: What They Are, Why They Matter Governments can by domestic policies give their national industries advantages in international competition. Occasionally governments have made outright payments to firms; more commonly, benefits take a less direct form, such as R&D support, tax breaks, loans on preferential terms, and provision of raw materials at below-market prices. Recognizing that such benefits can adversely affect industries in other countries, GATT sometimes permits those countries as a countermeasure to levy extra import duties called countervailing duties. Countervailing duties are allowed only if the benefits constitute a subsidy under the law. GATT does not directly define subsidies; U.S. law defines a subsidy as any bounty or grant paid upon the manufacture or production or export of any article or merchandisc. ~ Subsidies are of two types: export and domestic. Export subsidies apply only when goods are exported. For example, in 1985 New Zealand paid its producers of carbon steel wire rod 10.5 percent of the value of the exported product. 85 Other examples include preferential loans to finance exports and a reduction in corporate income tax conditioned on export performance. GATT largely prohibits export subsidies. 86 Domestic subsidies are those paid whether or not goods are exported. Benefits might be in proportion to the firms total production--e.g., the provision of a raw material at a below-market price. Or benefits might be given to the company with no clear relation to a particular product; this could occur, for example, with R&D support or preferential loans for capital investment. Under U.S. law, domestic subsidies must be limited to a specific enterprise or industry, or group of enterprises or industries. 87 Broader benefits enjoyed by companies in other countries, such as abetter educated work force or easier access to long-term financing for all manufacturers, do not count as subsidies. As acknowledged by the GATT Subsidies Code, domestic subsidies are widely used as important instruments for the promotion of social and economic objectives. 88 Accordingly, GATT does not prohibit domestic subsidies. However, the GAT T Subsidies Code considers that domestic subsidies, by giving one nations businesses an advantage in international competition, may cause or threaten to cause injury to a domestic industry in another nation. 89 In such circumstances, GATT permits the second nation to levy a countervailing duty. It is imposed in addition to any other duties normally collected and must be no more than the value of the subsidy. For example, if a foreign nation provided coal at below-market prices to steel companies, saving $10 per ton of steel produced, then the maximurn countervailing duty allowed would be $10 per ton. It is not always possible to levy countervailing duties to fully capture and neutralize the benefits foreign companies receive from their governments. GATT permits an alternative countermeasure: antidumping duties, which can be imposed only when a foreign frm is dumping. Dumping is the selling of goods in an export market at less than a benchmark fair value. The benchmark is the price in the home market in the ordinary course of trade." 90 If no such home market price exists because, for example, there are no home market sales or the home market sales are not made in arms length deals 91 then the benchmark can be either the price in the ordinary course of trade in a third country, or a price constructed from the cost of production. plus a reasonable addition for selling cost and profit."92 Although GATT appears to relegate costs to a subordinate role, as an option only when home market price cannot be used, the United States and other major users of antidumping duties interpret these provisions as giving costs a primary role. These countries provide that sustained home market sales at a price below the cost of production (excluding profit) will not be deemed to be in the ordinary course of trade and should be disregarded in determining g the home market price. Therefore, any home market price actually used will normally be at least the cost of production excluding profit; sales below this level will be considered dumping. If all of the home market sales for a sustained period are below the cost of production excluding profit, then there will be no home market price in the ordinary course of trade, so that the benchmark used will be a constructed price consisting of costs including profit. 93
PAGE 144
Chapter 4--U.S. Trade Policy l 139 GATT states that dumping is to be condemned if it causes or threatens material injury to an established industry, or materially retards the establishment of a domestic industry, in the importing country. % GATT does not prohibit dumping in these circumstances. Rather, as with domestic subsidies, it allows the importing country to levy an antidumping duty, which must not be more than the amount necessary to bring the price up to the benchmark value. This amount is called the dumping margin. Antidumping duties, like countervailing duties, are in addition to any other duties normally collected. Sustained dumping by foreign producers is often a sign that those producers enjoy an advantage in international competition due to either government policies or societal features such as industry structure and habitual business practices, which themselves may result in part from government policies. This is true for both price-price dumping (selling in an export market for less than at home or in another export market) and price-cost dumping (selling in an export market for less than costs). With price-price dumping, the foreign producer can use its sales in the high-priced market (usually the home market) to subsidize its sales in the low-priced export market. The price difference is often due to protection in the high-priced market. 95 While one way to end the foreign fins advantage might be to remove the foreign market barriers, in practice that can be difficult. 96 Sometimes the foreign producers home government tolerates or even encourages cartel pricing; this elevates the home market price beyond what it would be with protection alone and increases the profits from home market sales that are available to subsidize low-priced exports. Price-cost dumping indicates that the foreign supplier has a special advantage. Sustained sales below cost are normally possible only if the sales are somehow subsidized. In some cases, countervailing duty law could be used instead of antidumping law. Countervailing duty law could address subsidies provided by the government, and even those provided by related firms (e.g., in a Korean chaebol or Japanese keiretsu).sT But a fros cross-subsidy between product lines would not come under countervailing duty law, so that only antidumping law could provide relief. Sustained sales below cost could be part of a long-term plan to gain experience and market share in a high-reward industry, which could lead to high future profits that might more than pay back the loan with interest. 98 In this case, the sales need not be subsidized, they need only be financed. For example, in the mid-to-late 1980s, capital for industrial growth was available to Japanese firms on better terms and with less pressure for short-term profits than in the United States. This included funding from the government, banks, and related firms. 99 Japanese dumping of DRAMs in the mid1980s can be viewed in this light; U.S. firms did not have the financial backing to undergo the massive losses that the larger Japanese firms swallowed, so most U.S. DRAM producers exited the field. Japanese firms emerged as the dominant producers. When used in high-reward industries, both priceprice and price-cost dumping can enable producers to achieve economies of scale and learning. Both types of dumping can help utilize excess capacity in industries experiencing slack demand. For example, during the late 1970s and the early 1980s, there was a worldwide excess steelmaking capacity. Steel firms in the EC, Japan, and many other countries had substantial market protection at home while they dumped in the more open U.S. market. l00 Since their facilities would otherwise stand idle, foreign firms could cut their losses by selling in the U.S. market for more than their variable costs, even if that was less than they charged at home, and less than their fully allocated costs. 101 In the United States, it has often been said that subsidies and dumping are unfair. This label irritates foreign firms and countries accused of these practices, who often do not consider that they have done anything unfair. Rather than debating whether these practices are unfair, it is more helpful to recognize that government action, societal structure, or a combination of the two can give foreign firms a special advantage against which U.S. companies often cannot compete unaided. Countervailing and antidumping duties should be seen not as a response to so-called unfairness, but rather as an interface or buffer mechanism to ameliorate difficulties caused by interdependence among different economic systems. "102 Countervailing Duty and Antidumping Cases The United States and other countries have laws to provide for investigations of subsidies and dumping and assessment of countervailing and antidump-
PAGE 145
140 Competing Economies: America, Europe, and the Pacific Rim ing duties. 103 Figures 4-4 and 4-5 show the countries most active in these proceedings, based on new investigations completed from January 1987 through June 1990. During this period the United States was the leading user of countervailing duty proceedings; Latin America, the Pacific Rim, and the EC were leading targets. The United States, Australia, Canada, and the EC most often used antidumping proceedings, while the Pacific Rim (especially Japan and Korea), the EC, nonmarket economies, Central and South America, and the United States were the primary targets. Antidumping cases were far more common than countervailing duty cases. Tables 4-4 and 4-5 detail who brought cases against whom and how many resulted in some final government action. 104 Tables 4-6 and 4-7 list how many final actions by each country were outstanding as of June 30, 1990. 105 From these figures and tables it appears that in recent years the United States has been by far the biggest user of both countervailing and antidumping duties. This does not necessarily mean, however, that the United States is more aggressive in their use. The United States has a large, relatively open market that attracts imports. Countries with more market barriers have less need to impose countervailing and antidumping duties. The United States principal trading partners use subsidies more often, 106 so the United States has more occasion to levy countervailing duties. As for antidumping cases, the United States has not always been out in front. In the 1980s, Australia brought 30 percent of all reported cases; the United States brought 28 percent, Canada 20 percent, and the EC 19 percent. 107 GATT allows countervailing and antidumping duties only if the effect of the dumping or subsidization is such as to cause or threaten material injury to an established domestic industry, or is such as to retard materially the establishment of a domestic industry. 108 This condition is called the injury test or requirement. Under U.S. law, the injury or threat of injury must be caused by the dumped or subsidized imports rather than by the dumping or subsidization itself. 109 U.S. law imposes the injury test in every antidumping case; for countervailing duty cases, however, the injury requirement applies only to countries that have signed the GATT Subsidies Code (this includes the United States major trading partners), have assumed obligations with respect to the United States which are substantially equivalent to obligations under the GATT Subsidies Code, or are otherwise entitled by international agreement to application of the injury test. 110 The Department of Commerce determines whether goods are subsidized or dumped and by what Figure 4-4-New Countervailing Duty Cases Completed, January 1987 Through June 1990 United States Australia New Zealand Canada Other Western Europe South Korea Other EC Other Pacific Rim Latin America 1 1 7 2 2 7 12 14 19 50 40 30 20 10 0 10 20 30 40 50 = Initiated against m Initiated by SOURCE: Semiannual filings by signatories of the GAIT Subsidies Code to the Committee of Signatories, provided by the U.S. Department of Commerce, International Trade Administration.
PAGE 146
Chapter 4--U.S. Trade Policy l 141 Figure 4-5-New Antidumping Cases Completed, January 1987 Through June 1990 United States Australia Canada Mexico New Zealand Other Western Europe Other South Korea Central & South America EC Japan Other Pacific Rim Nonmarket economies 1-.. 3 3 1 137 2 83 13 57 7 13 n 3 6 7 4 3 8 52 50 55 200 150 100 50 0 50 100 150 20 0 = Initiated against m Initiated by SOURCE: Semiannual filings by signatories of the GAIT Antidumping Code to the Committee on Anti-Dumping Practices, provided by the U.S. Department of Commerce, International Trade Administration. Table 4-4-New Countervailing Duty Cases Completed January 1,1987 Through June 30, 1990 Cases filed against: Other Other United New Western South Pacific Latin States Australia Zealand Canada Europe Korea Other EC Rim America Total Cases filed by: United States Final action . . . . No final action . . . . Australia Final action . . . . No final action . . . . New Zealand Final action . . . . No final action . . . . Canada Final action . . . . No final action . . . . Total Final action . . . . No final action . . . . 1 0 1 0 1 0 1 0 1 1 2 0 0 0 3 1 1 3 4 3 1 1 3 2 0 1 1 1 0 2 3 1 1 3 6 4 1 1 4 6 8 9 5 3 0 0 1 3 1 3 8 10 6 9 30 18 2 6 1 3 3 4 36 29 SOURCE: Semiannual reports submitted by signatories of the GATT Subsidies Code to the Commitee of Signatories, provided by the U.S. Department of Commerce, International Trade Administration. Additional explanation is given in the accompanying text. amount, and the ITC determines whether the injury grounds for believing that subsidies or dumping test is satisfied. While the government can start an have occurred and that the injury requirement is investigation on its own, that very rarely happens; satisfied. The petition must identify any known usually an investigation begins only when industry foreign manufacturers and exporters, as well as U.S. petitions the government and sets out sufficient importers, wholesalers, and retailers of the foreign
PAGE 147
142 l Competing Economies: America, Europe, and the Pacific Rim ma cum co. OJo ma No I-I% C9 No 70 0)0 mm l-o l-ml -0 C4ml Oln 0. T-O lo lo l-ml 0. No WC-9 0. Com . . s
PAGE 148
Chapter 4--U.S. Trade Policy 143 Table 4-6--Outstanding Measures Against Subsidies as of June 30, 1990 Negotiated quantity restraint or Country Duties other agreement Total United States . . 73 13 86 Canada . . . NA a NA a 9 Australia . . . NA a NA a 1 a NA Not available. Canada and Australia did not describe the nature of their outstanding orders. NOTE: All other signatories reported no outstanding orders. SOURCE: Semiannual reports submitted by signatories of the GAIT Subsidies Code to the Committee of Signatories, provided by the U.S. Department of Commerce, International Trade Administration. Table 4-7-Outstanding Antidumping Measures as of June 30, 1990 Price Country Duties undertakings Total United States . . . 192 6 198 EC a . . . . . . 82 53 135 Canada . . . . . 69 12 81 Australia . . . . . 12 0 12 New Zealand . . . N A b N A b 11 a Data for the EC is as of Feb. 1, 1990. The EC filed no Semiannual report for the first half of 1990;that would normally imply that no additional action had been taken. b NA = Not applicable. New Zealand did not describe the nature of its outstanding orders. NOTE: All other signatories reported no outstanding orders. SOURCE: Semiannual reports submitted by the signatories of the GAIT Antidumping Code to the Committee on Anti-Dumping Practices. goods. Those respondents are invited to participate in the investigation, along with the government of the foreign country involved. Figures 4-6 and 4-7 show the progress of antidumping and countervailing duty cases through the Commerce Department and the ITC, along with the number of cases that have progressed through each stage for all cases completed between January 1987 and June 1990. The figures show that cases with the injury requirement progress through three stages. At each stage, the case can end with no relief; only if all three stages are passed is relief granted. First, the ITC makes a preliminary determination of whether there is reason to believe that the injury requirement could be satisfied. Second, the Commerce Department determin es whether the imports at issue are subsidized or dumped, and if so by how much. Third, the ITC makes its final determination of whether the injury test is satisfied. If subsidies or dumping are found, and if any applicable injury test is passed, the government issues an order, enforced by the Customs Service, assessing countervailing or antidumping duties in an amount equal to the subsidy value or dumping margin. In dumping cases, each subsequent shipment is assessed an antidumping duty based on how much that particular shipment is priced below the benchmark fair value. Thus, foreign exporters can avoid paying any duty by raising prices up to the fair value, which they often do. In both subsidy and dumping cases, annual reviews are conducted on request to determine whether the amount of subsidy or the fair value, respectively, has changed. If so, the duties already paid for that year are adjusted; the exporter then pays any additional amount due or is refunded any overpayment. The U.S. Government sometimes accepts an agreement by the foreign government or exporters to eliminate the subsidy or limit the quantity of exports to the United States, in lieu of a countervailing duty. 111 Or, instead of levying an antidumping duty, the United States might accept an agreement by the foreign exporters to raise prices. 112 The Effectiveness of U.S. Countervailing Duty and Antidumping Law Countervailing duty and antidumping laws have helped somewhat to protect U.S. industries from advantaged foreign competition. However, the effect of these laws is limited. The reasons come in two clusters. First, the imposition of countervailing and antidumping duties is slow and uncertain. Second, even if they were promptly applied, countervailing and antidumping duties would often be inadequate to neutralize the advantage of foreign companies. Delay and Uncertainty in Imposing Duties The government normally will start an investigation only when an affected U.S. business files a petition. This takes considerable time and effort. The firm must suspect that subsidies or dumping are at work; it could take months or even years to piece together why certain sales were lost and conclude that there is more than just ordinary stiff competition. For example, a company might suspect dumping only after it is forced to make a series of painful price cuts in response to foreign competition. The company must be aware that a legal remedy exists; some fins, especially smaller ones, do not know that. Plus there is a certain inertia to overcome. The Torrington Co. waited until March 1988 to file petitions regarding dumping of ball bearings from
PAGE 149
144 l Competing Economies: America, Europe, and the Pacific Rim Figure 4-6-Decisions in U.S. Countervailing Duty Cases Completed, January 1987 Through June 1990 1 Total cases filed 3 9 I I I 23 case s 16 cases not subject to injury test ITC: is injury or threat no Case ends: reasonably possible? no cases no relief I 1 1 I n I yes 23 case s Department of Commerce: no Case ends: are there subsidies? 9 cases no relief I I I J I I I yes I 20 cases 10 cases not subject to injury test yes ITC: does injury or threat 0 Case ends: of injury exist? 5 cases ro relief 2 partial 2 yes 13 cases 2 partial 2 I Receive relief from subsidies 23 cases and 2 partial NOTES: 1. This table does not include 7 cases that were terminated before a final determination was made and 5 cases for which insufficient data was provided. 2. In these cases, injury was found for some of the products at issue but not for others. SOURCE: Unpublished data, U.S. International Trade Commission. several countries, even though dumping and injury were apparent by early 1987. 113 When asked about the delay, a Torrington official stated: Most businessmen worry about the day-to-day order book. To start a trade action, you must sit back. 114 To file a petition, a firm must present sufficient evidence of both injury and subsidies or dumping. Regarding subsidies, the petitioner must identify particular subsidy programs before the government will investigate them, but it can be difficult for U.S. firms to learn about these programs. With dumping, the petitioner must first get evidence of the prices at which the imported goods are sold in the United States. These prices would often be known only by the foreign producer and the domestic purchaser; those parties would not want to help the petitioner because any antidumping duties would be assessed against their goods. The petitioner then would have to document the foreign manufacturers home market price or costs of production. This too can be difficult, especially if the petitioner has no presence in that country. Evidence concerning injury can be elusive, because at issue is the state of the whole domestic industry, not just the fro(s) filing the petition. The other companies in the domestic industry might not wish to give petitioners the information they need; for example, they might be unwilling to share sensitive business information, or they might be foreign-owned, with loyalties to the respondents Camp. 11s Even if a firm comes up with the necessary information, other factors could dissuade it from preparing a petition and fighting the case. The frost is expense, which typically ranges in the hundreds of thousands of dollars for the initial case (table 4-2), not counting employee time. Additional expense is incurred for court appeals, which are frequently taken, and for subsequent annual reviews, which are frequently conducted. Some recent cases have been quite expensive. For example, in 1989 Bethlehem Steel filed both an antidumping and a countervailing duty petition against steel rails from Canada, which resulted in substantial duties. 116 To win the case at the Commerce Department and ITC, and to defend the victory on appeal, took over $1 million. 117 When cases are filed against more than one country, the expense increases. The ball bearings cases filed by the Torrington Co. in 1988 included nine countries,
PAGE 150
Chapter 4--U.S. Trade Policy l 145 Figure 4-7Decisions in U.S. Antidumping Cases Completed, January 1987 Through June 1990 i Total cases filed 109 1 109 cases ITC: is injury or threat reasonably possible? M n + Yes 92 cases Department of Commerce: is there dumping? m 1 yes 85 cases 1 partial 2 1 partial 2 ITC does injury of threat of injury exist? = 1 yes 50 cases 16 partial 3 16 partial 3 Receive relief from dumping 50 cases and 16 partial NOTES: 1. This table does not include 10 cases that were terminated (9 cases) or suspended (1 case) before a final determination was made. 2. In this case, dumping was found for some of the products at issue but not for others. 3. In these cases, injury was found for some of the products at issue but not for others. SOURCE: Unpublished data, international Trade Commission. at an estimated cost of $4 to $5 million. 118 The high costs can discourage firms, especially smaller ones, from filing petitions. 119 In estimating the expense, a firm must consider the many reasons why it could lose the case and that even a successful case might not provide effective relief. The expense, uncertainty, and difficulties of gathering information dissuaded U.S. manufacturers of rock crushing machines from filing a case in the late 198% (box 4-b). In response to the high legal and other fees as reported by the General Accounting Office (table 4-2), Congress in 1988 stepped up the role of the ITCs Trade Remedy Assistance Office (TRAO). 120 TRAO can help small businesses represent themselves. TRAO gives behind-the-scenes technical and legal advice, including explaining the law and procedures, identifying relevant precedents, consulting on strategy, and reviewing draft petitions and briefs prepared by the firm. Since the 1988 legislation, TRAO has helped several companies file and pursue dumping cases. While companies have been pleased with the assistance, there are some inherent limitations to how much TRAO can help. Petitioners representing themselves have no access to confilential business information submitted by other parties; the ITC makes this information available only to outside counsel and consultants. This information is central to many cases, and lack of access puts a party at a tremendous disadvantage in formulating strategy and framin g legal arguments. Further, even with TRAOs help, it requires immense effort for a firm to represent itself. Businesses can be discouraged from filing petitions because they fear retaliation-by the accused firm or a related one that might be the prospective petitioners supplier or customer for some products, or by the foreign government. For example, it is widely believed that the U.S. commercial airframe manufacturers, Boeing and McDonnell Douglas, for many years did not file a countervailing duty case against large subsidies received by the members of the Airbus consortium from the governments of Britain, Germany, and France, for fear that those governments would direct their national airlines to buy fewer U.S. airplanes. As another example, in the spring of 1991, a survey by the U.S. and Japanese Governments found that auto parts were priced 108
PAGE 151
146 l Competing Economies: America, Europe, and the Pacific Rim Box 4-BAn Unfiled Dumping Case l The U.S. rock crushing machinery industry is composed of about nine companies with total annual sales of about $175 million (including new units and spare parts). These machines are used in many public works and construction projects. The steady decline in infrastructure spending over the 1980s greatly reduced the demand for crushing equipment; 1989 sales were half those of 1979. The situation was made worse by sales lost to low-priced imports from Japan, Korea, and Taiwan that appeared in the early 1980s. After 1985, as the dollar was devalued the U.S. firms expected some relief, yet the imports dollar price did not rise. Manufacturers lost sales in both 1987 and 1988 of some 20 to 30 machines out of 200 or 300 total sales in the United States; the west coast firms were hardest hit. Though the U.S. producers might well have been able to win a dumping case against the imports, they never filed one. The industrys attempts, and the problems it faced, show how hard it can be to use antidumping law, especially for small businesses. One west coast producer, referred to here as Firm A, organized an ad hoc committee of manufacturers and acted as the spokesman to the government. The commerce Departments International Trade Administration (ITA) gave Firm A the legal definition of dumping. Since they were unable to determine the Japanese firms home market prices, the U.S. manufacturers would have to show dumping on the basis that the Japanese machines in the U.S. market were priced below the cost of production. This required an analysis of the costs of production facing the Japanese companies. The major costs of making rock crushing machines are materials and labor. Because the overwhelming material input is steel, which as a commodity would normally cost about the same anywhere in the developed world, Firm A believed that the material costs of U.S. and Japanese machines might be equal. Using thi S assumption, and knowing that the Japanese producers also had shipping and duty costs, Firm A was fairly sure that Japanese labor and other costs could not be low enough for the price of the imported machines to be above cost. However, the foregoing analysis was not sufficiently rigorous to get ITA to initiate an antidumping case. The U.S. manufacturers understood this, but did not know how to proceed. ITA had sent some questionnaires whose answers would supply the necessary information, but those forms were imposing and Firm A thought they would be very difficult to complete. An analysis along the lines given below might have been sufficient to induce ITA to initiate a case. The analysis below, summarized in table B-1 2 is for one representative Japanese model and its counterpart produced by Firm A. The analysis would have to have been performed for all import models at issue. List Price: List prices were readily available from promotional literature. List prices represent the maximum possible retail price, and most machines are sold below list. Unconfirmed reports of Japanese machines being sold at substantial discounts, some as high as 40 percent, were not uncommon. For the calculation, the list price was used initially, with discounts considered afterward. Dealer Markup: It was assumed that retailers of imported crushing machinery receive the same percentage markup as retailers of U.S. made machinery. U.S. machinery producers believed that the markups were similar, though they could not prove it. Duties and Shipping Costs: Table B-lEstimated Cost Breakdown of Rock Crushing Machinery, Jaw Crusher Type Japanese U.S. firm firm List price . . . . . . . 1,600 1,000 Less dealer mark-up . . . . -320 -200 Wholesale price (to producer) . . 1,280 800 Less duties . . . . . . O -24 Less shipping costs . . . . 0 -54 Factory price . . . . . . 1,280 722 Costs: Material . . . . . 377 377 Direct labor . . . . 194 136 Variable overhead . . . 349 209 Fixed overhead . . . . 147 74 Total costs . . . . . . 1,067 796 Profit (negative number for loss) . 213 -74 (Factory price minus total costs) Dumping margin . . . . . 9.3% (Loss divided by wholesale price) NOTE: The figures given do not represent dollars. They are given in an undisclosed unit of currency, to protect proprietary business information. The table contains slight rounding errors. SOURCE: Office of Technology Assessment, 1991. Data on shipping and duty charges were readily available and required no estimation. ITMS bX is b~ed on interviews with firm exeeutivea, the trade lawyer consulted by the f-, and a former Department of Commerw ofticial. 2~e fi~es in @ble B-1 do not represent dollars. The eurreney unit used is not disclosed, in order to prOteCt prOpfiem hs*s information.
PAGE 152
Chapter 4--U.S. Trade Policy l 147 Factory Price: Factory price is the amount the producer receives, not counting any duties and shipping costs. To prove dumping based on below-cost sales, factory price must be less than the costs of production. Material Costs: The major material input was carbon steel. According to U.S. manufacturers, all other materials costs were insignificant. Carbon steel is a commodity good traded worldwide. Steel prices faced by Japanese competitors might then be assumed to be equal to those faced by U.S. firms. As the comparable machines weigh the same amount, it was assumed that the same amount of steel is used. Thus, material costs were assumed to be equal. Direct Labor Costs: Japanese labor costs had to be estimated. The Department of Labor does make international comparisons of wage rates, but not for specific industries. The rock crusher workers in the United States were paid more than the average industrial wage, but because no data is available on rock crusher workers in Japan, as a conservative assumption the average industrial wage in Japan was used. The labor cost estimate also assumed that Japanese labor was 10 percent more productive; that is, if it took 100 hours to build a U.S. machine, it took 90 hours to build a Japanese machine. Variable Overhead and Indirect Labor Costs: The estimates of Japanese variable overhead costs (e.g., energy consumption, worker benefits, and workmans compensation) were more difficult to make than estimates of wages. Without access to hard data, variable overhead costs incurred per hour of direct labor in Japan were conservatively assumed to be two-thirds as much as the variable overhead costs incurred per hour of direct labor in the United States. (The figure in the table also reflects the assumption given above that Japanese labor was 10 percent more productive.) Fixed Overhead: Data on fixed overhead costs in Japan are not available. For the U.S. machine, fixed overhead was about 16 percent of the total of the other costs (material, direct labor, variable overhead). Without direct data, fixed overhead for the Japanese machine might conservatively be assumed to be about 10 percent of the total of the other costs. As table B-1 shows, under these assumptions Japanese firms were dumping. Even with sales at list price, the dumping margin (the cost minus price, divided by price) was 9.3 percent. Moreover, both domestic and foreign rock crushers were rarely sold at list price. One reported Japanese sale was at 40 percent off list. If the wholesale price was similarly reduced by 40 percent, then, using the same cost assumptions, that sale would have a dumping margin of 89 percent. To win a dumping case, the U.S. manufacturers would also have to prove injury. Under the law, a sizable number of sales lost to dumped goods can constitute injury if those sales were lost on the basis of price. In both 1987 and 1988, an estimated 20 to 30 imported crushers were sold in the United States, all in the west coast region, rimarily with Firm A and one other firm. All evidence available to Firm A from its customers indicated competing p that price was, in fact, the overriding factor. Comments such as, we liked the way you [the domestic producers] do business but, their price was just too low were typical. These lost sales probably constituted roughly 10 percent of the U.S. market for new machinery. Lost sales of this magnitude, if they could be proved, would stand a good chance of satisfying the injury requirement. However, the U.S. producers did not perform the above analysis, for they did not know how. Instead, they consulted an experienced trade lawyer. The lawyer explained that, while a dumping case had a fair chance of success, it would cost about $1 million. While this was a substantial amount to an industry with an annual profit only in the range of $5 to $10 million, especially since the chance of winning the case was far from certain, for the industry as a whole it probably would have been manageable. However, because the imports competed only in the west coast market, many of the companies whose sales were concentrated elsewhere did not feel injured enough to justify a large expense. The two firms bearing the brunt of the injury did not feel able to bear the bulk of the cases cost. In principle, the firms could have filed a petition on their own, and if the Commerce Department accepted the petition then the Commerce Department and the International Trade Commission would investigate dumping and injury respectively, whether or not the U.S. producers hired a lawyer. However, this arrangement would have required considerable employee time to develop the petition. Moreover, the lack of an attorney would pose two serious practical problems. First, a petition would need to contain companies confidential business information Continued on next page
PAGE 153
148 Competing Economies: America, Europe, and the Pacific Rim Box 4-BAn Unfiled Dumping Case l Continued regarding sales, prices, customers, profits, and so on. They could not share this information with each other. An outside attorney could take all pertinent information from the firms in confidence, combine it as needed in the petition, and designate certain information in the petition as confidential. Without an outside attorney, the firms were stuck As the case developed, the other side (foreign manufacturers and importers) would submit confidential business information of its own as key elements of its case. While an outside attorney for the U.S. producers could have access to such information, the companies themselves could not. Thus, if they argued the case without an outside attorney, the U.S. firms would have no way of refuting key arguments of the other side, since they could not know what those arguments were. 3 The Department of Commerce could initiate an investigation on its own on behalf of the U.S. industry. However, the Department of Commerce has done so only once in hundreds of cases. While the Department may have considered doing so in this case, it ultimately decided not to. The producers were forced to ride out the dumping without remedy. In this case, riding it out seems to have worked. The continued low value of the dollar and increased labor costs in Japan and other Asian countries have combined to push the price of imports up. Additionally, the continued decline in U.S. demand may have dissuaded foreign firms from buying a large market share through dumping. Domestic producers say the Japanese are still selling in the United States, but not with the same sell at any price fervor. This case highlights some problems faced by small firms in trying to attain relief from dumping. The problems included difficulty in getting needed data on foreign firms costs and prices, difficulty in collecting industry-wide data, difficulty in justifying the payment of legal fees (especially since success was uncertain), and a general lack of familiarity with the law. g~e~ Woulclbe given noncontidentid summaries of the confidential information, but those summaries would usually omit important details. percent higher in Japan than in the United States. will the duties collected be based on verified The Commerce Departments Auto Parts Advisory Committee, a group of 35 industry representatives, advised that the Commerce Department initiate a dumping case on its own. The Committee explained that industry would be reluctant to file antidumping petitions, partly because of the need to preserve business relationships with Japanese firms. 121 Filing petitions can also anger U.S. customers who stand to be deprived of cheap imported goods. After a petition is filed, a case takes time to yield results. Provisional duties are not imposed until the ITC first finds reason to believe the injury testis met and the Department of Commerce makes a preliminary finding of dumping or subsidies. This usually takes over 5 months for dumping cases and almost 3 months for subsidy cases. 122 The duties ordinarily apply only to goods that clear customs after the preliminary determinations are made and published, though in rare cases the Commerce Department will seek to collect duties retroactively for up to 90 days. 123 The Commerce Departments preliminary determin ation is often based on unverified statements by the foreign manufacturers. Only after a final determin ation, usually another 21/2 months, information. l24 Even these relatively modest delays could be important for products with short life cycles. Moreover, these delays must be added to the time it takes for a business to suspect subsidies or dumping and to prepare a petition. Additional delay can occur if further proceedings, and perhaps court appeals, are necessary to get duties assessed. One example is that after a countervailing duty order is in place, subsidy levels can be increased; it will take at least a year or two before the Commerce Department determines the new subsidy level in an annual review. In while increased duties will be assessed retroactively, the underresponse to the subsidy in the interim could allow the foreign producer to capture market share. An extreme example of delay pending review is the case of portable electric typewriters from Japan. In May 1980, Smith Corona Corp. obtained an antidumping order against these products. 126 Within 2 years after this order, Japanese firms started exporting portable typewriters that included memory. In January 1987, the Commerce Department ruled that the typewriters with memory did not come under the scope of the 1980 antidumping order. 127
PAGE 154
Chapter 4--U.S. Trade Policy l 149 Three court appeals later, in September 1990, the U.S. Court of Appeals for the Federal Circuit ruled that these typewriters did come under the 1980 order, and ordered the collection of antidumping duties. 128 However, antidumping duties could no longer be collected on typewriters imported in the intervening 8 or 9 years. 129 Delays of several years can also occur because of ineffective enforcement; sometimes the duties are never collected. 130 Effective relief is sometimes held up because one case after another must be filed. Each countervailing duty or antidumping case must name a specific product or group of products already entering the United States from a specific country of origin. However, after a case is won and countervailing or antidumping duties stem the flow of goods from the country named, the flow of dumped or subsidized goods can start up again from other countries. The Torrington Co. claims that this happened in the late 1980s with ball bearings, when several multinational firms switched production to facilities in other countries. 131 If foreign producers can easily switch production capacity among different final products, a producer might export one product until a countervailing duty or antidumping order issues, and then switch to another product. This occurred with steel in the 1980s. 132 After hundreds of cases were filed starting in 1982, in September 1984 President Reagan announced that countervailing duty and antidumping cases had been slow, cumbersome and incomplete in remedying [the U.S. steel industrys] injuries. Antidumping and countervailing duty orders are simple to circumvent for goods whose final stage assembly is easily moved across national borders. Such a move can put the goods out of the reach of a previous order; 134 thus, a new pattern of subsidized or dumped imports could go unimpeded until a new case was filed and a new order obtained. In 1988, Congress added so-called anticircumvention provisions so that the initial order can sometimes cover the new import patterns. However, these provisions apply only in very limited circumstances and will likely have little effect. 135 It might not be possible to strengthen these provisions while staying consistent with GATT as it now stands. 136 In the Uruguay Round, the United States and the EC have proposed that GATT should explicitly permit certain anticircumvention measures, to deal not only with cases of relocated final assembly but also with relocated parts production, as occurred with ball bearings. Another possible impediment to relief, in the case of subsidies, is the specificity test. As discussed, a government program is counted as a subsidy only if it in fact selectively benefits a specific firm or industry, or group of firms or industries; programs granting equal benefit to a broad range of firms and industries are not counted. Foreign countries can phrase their laws so as to paint benefits as widely available, even when in practice they are not. Former U.S. Government officials representing foreign firms in trade cases could coach foreign governments on how to do this. When foreign laws appear to grant benefits to a wide range of fins, often as a practical matter the difficult task of proving that the benefits are indeed selective falls on petitioners. In other ways as well, the Commerce Department at times does not recognize certain subsidies. For example, the Commerce Department does not count R&D subsidies if the results are made public, but making the results public does not negate the benefits to the firm of being among the first to know and the most familiar with the new discoveries. As another example, if a foreign government disguises a long-term loan as a series of short-term loans, the Commerce Department often will count it as shortterm; this can decrease the computed subsidy if, as is most often the case, the market interest rate for long-term loans exceeds the market rate for shortterm loans. Another important limitation on the effectiveness of countervailing duty and antidumping law is the injury test. U.S. law, implementing GATTs requirement, provides that antidumping and countervailing duties cannot be levied unless by reason of imports of [dumped or subsidized] merchandise or by reason of sales (or the likelihood of sales of that merchandise) either an industry in the United States is materially injured or threatened with material injury, or the establishment of an industry in the United States is materially retarded. Congress has further defined material injury as harm which is not inconsequential, immaterial, or unimportant, and has specified certain factors to be considered, such as the volume and trend of imports, whether the imports depress or suppress prices, and sales, profits, and capacity utilization in the affected industry. 137 It is up to the ITC, which determin es injury in U.S. cases, to evaluate and weigh the factors. The ITC has considerable discretion and the courts seldom overturn its decisions.
PAGE 155
150 Competing Economies: America, Europe, and the Pacific Rim As shown in figure 4-7, the injury test is the primary reason why petitioners lose antidumping cases. Well under one-tenth of the cases (6 cases, plus 1 partial case, out of 109) are lost because no dumping is found. In contrast, one-third of the cases (37 out of 109) are completely lost because no injury is found, and about half of the cases (53 out of 109) are lost completely or in part because no injury is found. Figure 4-6 shows that, while a factor, the injury test in countervailing duty cases is less important. Of 39 cases, 23 were subject to the injury test. Of those, 5 whole cases and 2 partial cases were lost because of the injury test. The injury test adds expense. The legal fees and other expenses needed to prove injury are often greater than those needed to prove subsidies or dumping. Sometimes they are quite large. For example, in 1988 and 1989 The Torrington Co. spent a quarter of a million dollars on economic consultants merely to criticize and comment on an economic model proposed by Commission staff to analyze the injury. Torrington believed its opposition spent an equivalent sum. 138 While much of this expense is unavoidable due to the complex, technical subject matter, the ITCs approach to injury determinations does add to it. The ITC determines each case on its own facts, without articulating or following rules of general applicability. One such rule could be that a drop in profit of a certain percentage of sales, if linked to the imports, normally establishes the requisite injury. Without safe harbor rules such as this, petitioners are compelled to develop the facts and argue the law for all possible factors that the ITC might considerprofits, sales, capacity utilization, and so on. The ITCs case-bycase approach also makes case outcomes less predictable, which can discourage petitioners from filing complaints. The injury test poses a particular problem for firms seeking relief in good economic times, at the beginning of a product cycle, or before much damage has been done. As the ITC has interpreted the test in some recent cases, industries that appear fairly healthy in an absolute sense and whose health has not recently declined will likely fail the injury test, even if they would have done far better if not for the subsidized or dumped imports. This happened in a case decided in January 1989 concerning digital readout (DRO) systems from Japan. The Commerce Department had found large dumping margins on DRO systems, ranging from 39 to 55 percent. 139 DRO systems consist of two components: a transducer, which measures the position of equipment such as metalworking machine tools during operation, and a console, which displays the resultt in digital form. In determining injury, the International Trade Commission considered console production and transducer production to be two separate industries, and made a separate determin ation for each industry. Four Commissioners, a majority of the sixmember International Trade Commission, ruled that the required injury was not present for either industry. The Commissioners never considered whether the dumped imports made the industries less successful than they otherwise would have been. l40 Two Commissioners dissented. They pointed to strong economic evidence, supplemented by an admission by one of the major Japanese manufacturers, that the Japanese firms had maintained their 141 They also noted market share only by dumping. strong evidence that the dumped imports had substantially lowered the price for DRO systems in the U.S. market. 142 These two Commissioners believed that injury should be considered primarily in terms of the effect of the dumping, rather than the general health of the industry. The majority approach can make it hard for U.S. firms to win dumping cases early in a product cycle (when markets are growing) or before significant damage has been done. This can put companies in a dilemma: file a case early and risk that not enough injury will be found; or file later, when the damage is substantial and hard to reverse 143 The DRO systems case shows another hurdle in the application of the injury test. In assessing the health of the domestic industry, the Commission considered export sales together with domestic sales. Under this approach, even if subsidies or dumping severely hurts the domestic industry in its home market (the only market that U.S. countervailing or antidumping duties can address), no relief will be obtained if the U.S. industrys export sales of consoles are strong enough to make the overall industry healthy. In the DRO case, export sales were apparently much more profitable than domestic sales, which appeared to be made at a 1 OSS l44 The injury test permits the respondents to claim that the domestic industrys troubles came not because it was undersold but for other reasons. When credible alternative reasons are offered, often as a practical matter the burden of proof shifts to the
PAGE 156
Chapter 4--U.S. Trade Policy 151 petitioners to show that the low-priced imports were responsible for at least some of the material injury. This issue arose in a case brought in 1990 by National Label Co., a firm in Lafayette Hill, Pennsylvania. 145 National Label made pressure-sensitive polyvinyl chloride covers for alkaline and some other batteries. These covers used a new technology that expands battery life. l% National Label had been making the battery covers under a 1986 patent and trade secrets license from Zweckform, a German firm, which gave National Label the exclusive right to exploit the technology in North America. Relations between the two firms soured, however, and in December 1988 Zweckform told National Label it was terminating the license, and (according to National Label) began approaching potential U.S. customers. National Label asserted that Zweckform did not have the right to terminate its exclusive license and sued Zweckform in Federal court to stop it from selling in the U.S. market. 147 In January 1990, National Label also filed an antidumping petition, alleging that Zweckform was dumping battery covers in the U.S. market with the intent to drive National Label out of the business. National Label maintained that it had been injured in at least two ways: first, that it had lowered its prices to keep from losing sales to its principal customer, Eveready; and second, that it had lost a sale to a prospective new customer, Mutec, a potentially powerful new entrant into the battery market backed by Matsushita and Kodak. Regarding its Eveready account, National Label submitted sworn statements by its president that in March 1989, Zweckform representatives visited Evereadys office in Chicago and an Eveready official present at that meeting told him later that Zweckform had solicited to sell Eveready battery covers for the U.S. market. In April 1989, according to the statement, Zweckform gave him a choice between selling his battery cover business to Zweckform or being driven out of the business through deliberate below-cost pricing by Zweckform in the U.S. market. 148 How ever, Zweckform disputed the facts of the case, claimin g that the March 1989 meeting in Chicago was to discuss the European market only, and that Zweckform had never approached Eveready regarding the U.S. market. 149 Apparently Eveready told the Commission that Zweckforms version of the facts was correct. The Commission agreed with Zweckform and found that Zweckform had not offered to sell to Eveready, and therefore could not be held responsible for National Label lowering its prices to Eveready. 150 Regarding the prospective Mutec account, Zweckform said that it won the sale to Mutec because National Labels samples did not meet Mutecs specifications, rather than because National Label charged more. Mutec gave testimony agreeing with Zweckform. National Label disputed this point, arguing that Mutecs rejection of the samples was a sham. National Label argued that Mutecs parent Matsushita had found National Labels samples acceptable; Eveready and other firms found National Labels cover acceptable; the rejection of National Labels samples came after Mutec had already chosen Zweckform as a supplier, with the antidumping case started or anticipated, and it was then in Mutecs interest to claim that National Labels samples were unsatisfactory; and problems found in samples can often be fixed, but here there was no chance to do so because thanks to Zweckforms low price, Mutec had decided to purchase from Zweckform before it ever tested National Labels samples. 151 The commission agreed with Zweckform and found that Zweckforms low price was not what had caused National Label to lose the sale to Mutec 152 OTA does not know whose version in this case is correct. The ITCs decision was based on confidential information not available to OTA. But this case does illustrate some difficulties facing a petitioner. In both cases of alleged injury, key testimony as to the real cause of National Labels injury came from Zweckforms customer (Mutec) or potential customer (Eveready)--parties predisposed to take Zweckforms side. The details of the tests performed by Mutec on National Labels samples were considered Mutecs confidential business information. While National Labels outside counsel could see this information, National Labels president, who was a technical expert, could not. 153 Nationa l Labels president believed that this handicapped his ability to challenge the rejection. l54 Offsetting Foreign Advantages Even if antidumping and countervailing duties were applied reliably and immediately to subsidized and dumped imports, this often would not offset the advantages of foreign firms. Some practices and effects slip through the cracks; they cannot be addressed as either subsidies or dumping. Some-
PAGE 157
152 l Competing Economies: America, Europe, and the Pacific Rim times the foreign practice can be stopped by negotiation (perhaps using Section 301 or GATT dispute resolution procedures), though such negotiations, like negotiations to remove market barriers, are likely to be slow and difficult. Countervailing and antidumping duties affect only sales in the U.S. market. Subsidies can give foreign firms an advantage in their home and third country markets, and dumping can mean an advantage for foreign firms in third country markets. Such subsidies have been addressed in several Section 301 cases (most involved agricultural products), and third country dumping of memory chips by Japan was addressed in the simultaneous settlement of a Section 301 case and three antidumping cases. Realizing that its firms needed relief from third country dumping, the United States negotiated a special co mmitment by Japan not to dump in the United States or in third countries. 155 Another practice that can fall through the cracks is sales in the U.S. market at very low profit margins-high enough to sustain foreign manufacturers, but low enough to squeeze U.S. manufacturers. If the same price were charged at home, such sales would not constitute dumping. Lower profit margins might be more acceptable in another country in which conditions for raising capital are different. Or the foreign firm might be crosssubsidizing its low-priced exports to the United States with profits from other product lines, as Japanese supercomputer manufacturers have begun doing. 156 That cross-subsidization could not be countervailed. Perhaps the most profound Iimitation is that subsidies are quantified in a manner that systematically understates their true value. To calculate the countervailing duty, the Commerce Department frost puts a monetary value, called the subsidy value, on the benefit conferred. For example, the subsidy value of a low interest loan would typically be the interest saved by the company compared to the interest that it would have had to pay on a commercial loan. Next, the Commerce Department allocates the subsidy value over time. For example, the subsidy value of a low-interest loan could be allocated over the life of the loan. Then, for each period used for calculating countervailing duties, the countervailing duty rate for imports is that part of the subsidy value allocated to that period, divided by the value of the goods produced (or sold) during the same period with the help of the subsidy. For several reasons, a countervailing duty computed in this manner often does not compensate for the advantage received by a foreign producer. Sometimes the subsidy is essential in order for the foreign producer to export in the first place. Often government financing makes it possible to get crucial bank loans, without which the business might never get off the ground. In the 1960s and 1970s, the Japanese Government subsidized its computer industry by means of low-interest loans from the Japan Development Bank (JDB) to the Japan Electric Computer Co. (JECC), a joint venture of the Japanese computer manufacturers. JECCs function was to finance computer sales; JECC bought computers for cash from the manufacturers and leased them to users. The loans from JDB made it possible for JECC to acquire additional commercial loans. Without JECCs help, most of the Japanese manufacturers probably would not have been able to borrow money to finance sales of computers. These manufacturers would then have been restricted to customers who could pay cash, and as a result Japans computer industry would have developed more slowly. 157 But if U.S. manufacturers had filed a countervailing duty case, what would the countervailing duty have been? As calculated by one scholar, the subsidy implicit in the JDB loans from 1961 to 1969 was 2 billion yen, while the Japanese computer fins sales during those years were about 230 billion yen. 158 Therefore the subsidy, and the countervailing duty, would have been computed as less than 1 percent of the value of the goods sold. A similar mismatch between the countervailing duty computed and the advantage conferred on the foreign firm can occur with government support of R&D. An example is MITIs VLSI project in the late 1970s, which concerned techniques for making denser semiconductor chips. This project was important in helping the Japanese semiconductor and computer industries catch up with their U.S. counterparts, 159 and was important to the development of Japans supercomputer industry. 160 But if U.S. firms were to petition for relief in 1991, what countervailing duties might be imposed? The VLSI project is relevant to producing a wide range of electronics products; those of the participating firms will probably total well over $40 billion in 1991. 161 In
PAGE 158
Chapter 4--U.S. Trade Policy 153 comparison, the $150 million contributed to the VLSI project by MITI, of which little if any would be allocated to 1991, would yield a minuscule countervailing duty, far less than 1 percent of sales. Why does the countervailing duty calculation fail to capture the full worth of the R&D? Perhaps the most important reason is that the approach to calculating countervailing duties assumes incorrectly that benefits are static. Under the law, subsidies are used once, and then their effect is gone; the only issue is how to divide up the benefit over time and products. In fact, under the law an R&D project in the late 1970s, such as the VLSI project, would be considered irrelevant to sales in the 1990s; the projects payback period would be considered over and its effects dissipated. In reality, however, subsidies can create new opportunities and help a business seize an advantage years later; the effect can grow, rather than diminish, with time. In high reward industries this is particularly likely. 162 The effects of Japans VLSI project probably increase with each passing year. The same can be true, for example, with low interest loans. Under the law, preferential financing to Japanese computer companies in the 1960s is probably irrelevant to sales of supercomputers and other computers in the 1990s; but in fact, the effect of those loans is strongly felt in 1991 (see ch. 6). Under the law, it is not possible to capture the benefit that such government action confers on industries. Legalities aside, this task is very difficult. The foreign industrial policies discussed in this report are replete with actions intended to benefit national industries. To track the effects of all actions that helped a particular industry is a hopeless task. How Hard Are the Problems to Fix? The ineffective response of countervailing duty and antidumping law is not easy to fix. First, the law is inherently reactive. Duties are imposed in response to specific observed foreign practices; therefore, some delay occurs before the duties are imposed. In this way the approach of U.S. trade law differs fundamentally from one in which the U.S. Government would proactively seek to create advantages for its national industries, as other governments do. However, there are ways to improve the effectiveness of countervailing and antidumping duty law. For example, the injury test could be attuned to be more receptive to companies still doing fairly well, before serious damage is done; the Commerce Department could be directed to suspend closing out of customs entries as soon as a case is started, and, once imposed, to collect duties retroactively as far as possible; and the Commerce Department could be more aggressive in finding and evaluating subsidies. These changes might be challenged as inconsistent with GAITs requirements. For example, an attempt to change the evaluation of subsidies to more completely capture the subsidys full effect could run afoul of GATTs requirement that countervailing duties not exceed the estimated bounty or subsidy determined to have been granted. 163 Precisely what GATT allows is not always clear; ambiguities can be resolved only by a GATT ruling on a dispute between two countries, or by amendments to clarify the rules. Some countries have maintained that current U.S. practices violate GATT e.g., the use of a minimum of 10 percent general administrative overhead and 8 percent profit when constructing a products fair value based on foreign manufacturers costs, 164 and the practice of assuming until proven otherwise that petitions are supported by the majority of the domestic industry. This latter practice has already been found to violate the GAT Antidumping Code. 165 The United States has been negotiating in the Uruguay Round to allow countervailing and antidumping cases to be more effective. Along with the EC, it has pressed to provide clear authority for certain expanded anticircumvention measures, The United States has also pressed to prohibit domestic subsidies in certain cases, just like export subsidies are now prohibited. However, the United States faces an uphill battle in GATT, with few allies and with many nations pushing to weaken countervailing duty and antidumptig law. Several practical concerns counsel caution in trying to enhance the laws effectiveness. Imposing countervailing and antidumping duties entails substantial administrative costs, starting with an expensive investigation. Further, the computations used to determin e subsidies and dumping rely on imprecise data that may contain a fair amount of estimation. At times, duties will be levied when there really were no subsidies or dumping. The stronger the laws, the more often such mistakes will occur, throwing up misguided barriers to imports. 292-889 0 91 6 QL:3
PAGE 159
154 Competing Economies: America, Europe, and the Pacific Rim This problem is magnified because countervailing duty and antidumping petitions can be filed to harass ones competitors. While cases cost petitioners, they also cost the respondents. In fact, the burden on respondent firms can be greater, because the Commerce Department and the ITC, in developing the facts of the case, often scrutinize these fins business records. Just the threat of countervailing or antidumping duties creates uncertainty that can impede imports. And if, perhaps because of the proceedings inherent imprecision, very small countervailing or antidumping duties are imposed, there is the threat that in subsequent annual reviews these duties will be retroactively increased. In the case of antidumping law, the imprecision is not just in evaluating data, but also in how dumping is defined. The problem is that what is defined as dumping does not always justify a response. While dumping is often a sign that the foreign firm has a special advantage, this is not always the case. Dumping is sometimes a short-term ploy to get new customers; U.S. firms can often meet this type of competition on their own. As calculated, dumping could be found when a foreign firm prices identically at home and in the United States, if the prices in both markets fluctuate together. Or dumping could be found simply because of exchange rate fluctuations. 166 But if the definition of dumping excluded these transient effects, the law would also miss, at least for a while, situations that do require a government response. Another problem with dumping calculations concerns how to compare U.S. and home market selling prices. A fair comparison requires that differences in selling costs in the two markets be taken into account to net back to ex-factory prices for sale to each market; but when the foreign exporter sells through a U.S. subsidiary it seems impossible to write rules that will always fairly compare the two selling costs. It is not always possible to separate cases that warrant antidumping duties from those that do not. Even when foreign firms enjoy a special advantage, countervailing or antidumping duties might not be helpful. The U.S. industry might be well able to take care of itself, while important downstream industries could benefit from the cheap imports. For certain industries, the U.S. economy might be best served by the downsizing or eliminating the industry in question, with resources put to better use elsewhere. As useful as countervailing and antidumping duties are in some cases-and foreign firms and countries often target high-reward industries where the U.S. industry appears vulnerable-in other cases, they are not. Yet another consideration is that foreign countries can use these duties against U.S. exports. These countries include traditional users of countervailing and antidumping law such as the EC and Australia and newer users such as Mexico. Imprecision in computing dumping and subsidies, and the broad reach of the dumping label to include some conduct not warranting a response, can comeback to bite U.S. exporters. In some cases the United States has subsidized its industries, usually in an incidental rather than planned way. For example, governmentsupported R&D for military aircraft has helped the U.S. commercial aircraft industry (see ch. 8). 167 Another example is the governments support of half a billion dollars over 5 years for Sematech, an R&D consortium concerned with semiconductor manufacturing. If the United States decides to increase support for its industries, they could become more vulnerable to trade remedies in other countries. Where subsidy and dumping laws are ineffective, Section 301 might apply. In 1988, Congress specified that Section 301 applies to foreign export targeting, defined as any government plan or scheme consisting of a combination of coordinated actions bestowed on a specific enterprise, industry, or group thereof, with the effect of enhancing export competitiveness. 168 Section 301 would thus seem to cover a variety of foreign policy tools inadequately covered by U.S. antidumping and countervailing duty law, such as R&D programs and preferential financing. As of August 1991, no Section 301 case had been brought based on this provision, and such use of Section 301 would pose problems. There would be the same problems of delay, angering foreign countries, and soon as when Section 301 is invoked to eliminate foreign market barriers. Other countries are particularly averse to being told how to run their domestic economies. And Section 301 cannot reach back in time: a policy discontinued years or even decades ago, such as market protection or R&D support, can be the cause of an export surge today.
PAGE 160
Chapter 4--U.S. Trade Policy 155 OTHER MEASURES AGAINST IMPORTS Restrictions on particular imports can play a role in strategies for both mature industries facing difficult adjustments and for sunrise industries that promise fast growth. The appropriateness of protection depends on a host of factors and does not hinge on the presence of any individual foreign practice such as subsidies or dumping. To justify such protection countries sometimes invoke GATTs Article XIX, the so-called escape clause The escape clause addresses situations in l which, because of a combination of unforeseen developments and the countrys GATT obligations, a product is imported in such increased quantities as to cause or threaten serious injury to a nations domestic producers. Under these circumstances, a country may temporarily suspend performance of its GATT obligations to the extent and for such time as may be necessary to prevent or remedy such injury. The nation may raise tariffs, or impose quotas or nontariff barriers, but must compensate the countries whose exports have been blocked. If agreement can be reached, the compensation can consist of concessions (e.g., reduced tariffs) on other items; if no agreement is reached, an affected country may impose substantially equivalent barriers on imports from the nation using the escape clause. 171 Typically, the affected country would raise tariffs on products of its choosing. U.S. law implements the escape clause in Section 201 and the following sections of the Trade Act of 1974, as amended. 172 While GAPP requires only that the increased imports cause or threaten serious injury, U.S. law specifies that the increased imports be at least as great as any other cause of injury. 173 The U S statute, reflecting GATT, requires that an existing industry be injured; the injury test cannot be met by showing that the imports are preventing the establishment of an industry. 174 In these ways and perhaps others, the injury test under Section 201 is more stringent than the injury test in subsidy and dumping cases. To assess injury, U.S. law specifies factors such as idling of facilities, profit levels, and unemployment levels. 175 The ITC dete rmines injury under Section 201. A petition from industry or organized labor normally triggers an ITC investigation, 176 and the ITC typically takes 4 months to decide whether the injury requirement is satisfied. 177 If the injury requirement is not satisfied, the case ends with no governmental action; if the requirement is satisfied, then the ITC decides whether to recommend protection. The protection can be tariffs and/or quotas, lasting up to 8 years. 178 The ITC can recommend that the President negotiate with other countries to address the underlying cause of the increased imports or ameliorate their effect (as by restraining other countries exports), or take any other action authorized under law. 179 The ITC usually submits its recommendations to the President 6 months after the petition was filed. 180 The President then has 2 months to decide what trade relief, if any, to grant. 181 The President is supposed to weigh the social benefits and costs, and consider efforts being made by the domestic industry to adjust to the import competition and the position of the domestic industry in the United States economy. 182 The President has great flexibility in selecting the form of relief; he may choose one or more measures, including tariffs, quotas, auctioning off quotas to foreign producers, assistance to the domestic industry, negotiation of agreements with foreign countries, and any other action which may be taken by the President under authority of law. 183 Relief can last up to 8 years. 184 While relief is in effect, the ITC reports to the President every 2 years on industrys progress in adjusting to import competition. The President can then reduce, modify, or termin ate the relief based on changed economic circumstances or his determination that the industry has not made adequate efforts. 185 Table 4-8 summ arizes the use of Section 201 during 1975-1990. Of 62 cases, the ITC found injury in 33, and in 30 recommended protection. The President granted protection in 14 of the cases. Overall, a petitioners chance of obtaining some protection was about 23 percent. Further, the President frequently reduced the level of protection recommended. 186 Use of Section 201 tapered off considerably after the 1970s. The change probably occurred in part because the Trade Act of 1979 made it easier to win dumping and countervailing duty cases, so that those cases were often a more attractive route to relief. 187
PAGE 161
156 Competing Economies: America, Europe, and the Pacific Rim Table 4-8-Section 201 Cases Cases filed ITC President during years ITC found recommended granted Years shown injury protection protection 1975-90 . 62 33 C 3 0 d 1 4 e 1980-90 . 20 7 f 7 g 5 h 1986-90 . 2 0 0 0 a lncludes all caes in which the majority of the Commissioners voting on the remedy to be used recommended some form of protection, plus the one case (filed in 1977) where half of these Commissioners so voted. b Protection consisted of one or more of the following: increased tariffs (used 7 times), quotas (used 3 times), tariff rate quotas (increased tariffs charged only on imports above quotas, used once), negotiated agreements to limit foreign countries exports (used 4 times), and price supports (used once). c lncludes 6 cases in which injury was found only with respect to Some of the products at issue; includes 4 cases in which the ITCs vote on injury was evenly divided. d lncludes 6 cases in which injury was found only with respect to some of the products at issue; includes 2 cases in which the ITCs vote on injury was evenly divided. e All were cases in which a dear majority found injury and the ITC recommended protection. Includes 4 cases in which injury was found only with respect to some of the products at issue. f lncludes 2 cases in which injury was found only with respect to some of the products at issue. g Includes 2 C a S e S in which injury was found only with respect to some of the products at issue. h lncludes 2 cases in which injury was found only with respect to some of the products at issue. SOURCE: U.S. General Accounting Office, International Trade: Activity Under Section 201 of the Trade Act of 197.4 NSIAD-87-96FS (Gaithersburg, MD: U.S. General Accounting Office, Mar. 30, 1987), pp. 6-9; U.S. International Trade Commission Annual Reports, 1987-1990. In addition, the ITC found the injury test not satisfied in a highly visible case filed in 1980 concerning automobiles. To many, this seemed an ideal case for relief if ever there was one: an unexpected rise in oil prices suddenly made large U.S. cam an unattractive purchase compared to smaller Japanese cars. The ITCs negative decision in that case conveyed the message that the injury test under Section 201 would be very hard to meet. Although Congress in 1988 changed the statute in a way that likely would have changed the result in the autos case, 188 there has not been a resurgence of interest in Section 201; in fact, only one case has been filed since. Section 201 can be a useful trade tool. It operates within the GATT framework and therefore tends to be accepted by other countries. It has two advantages over countervailing duty and antidumping proceedings: it does not require the proof of subsidies or dumping; and it does not irritate foreign firms or countries by calling them unfair. Section 201 has significant limitations that restrict its usefulness, especially for high-reward industries. It applies only to an already existing domestic industry that is producing an article that competes with the imported goods. 189 It thus cannot be used to promote the formation of a new industry. 190 The industry must be seriously injured or threatened in an absolute sense; it is not enough that the industry would have done much better without the increased imports. 191 The Commission will find that a threat of serious injury exists only if serious injury, although not yet existing, is clearly imminent if import trends continued unabated. 192 These requirements make it hard to get relief at the beginning of a product cycle, when markets are growing rapidly. Section 201 has not often been invoked on behalf of high-reward industries. One prominent highreward sector, electronics, is represented by only two casestelevision receivers, filed in 1976, and CB radio transceivers, filed in 1977. The television case concerned a U.S. industry in decline, and the CB radio case concerned a potential growth industry that, according to the ITC, was in deep trouble because of the Federal Communications Commissions decision to expand the citizens band from 23 to 40 channels. 193 Section 201 has not been invoked for supercomputers, mainframe computers, workstations, personal computers, or semiconductors, where Japanese companies are a strong competitive threat to a U.S. industry. Even for mature industries, the injury requirement can make it difficult to get relief before much damage has been done. For example, in July 1984 the ITC found no threat of serious injury to the nonrubber footwear industry even though imports had increased in volume from51 percent of domestic consumption in 1979 to 65 percent in 1983. 194 The ITC reasoned in part that imports were predominantly at the low end and domestic production was predominantly at the high end, so that rather than displacing domestic nonrubber footwear production, imports have served to complement it. 195 In value terms, the imports had increased only from 36 percent of domestic consumption in 1979 to 44 percent of domestic consumption in 1983. 196 As the ITC spoke, however, the situation was changing. The industry filed another petition for relief, and in July 1985, one year after its first ruling, the ITC found a threat of serious injury. Import volume had risen to71 percent of domestic consumption in 1984, and 77 percent in the first quarter of 1985. The
PAGE 162
Chapter 4--U.S. Trade Policy l 157 imports had moved upscale; their value was 54 percent of domestic consumption in 1984 and 62 percent in the first quarter of 1985. 197 Even if the legal hurdles are passed, relief still depends on the President. Recent Administrations have been averse to granting import relief in Section 201 cases. This results from a free trade orientation and the obligation under GATT to pay compensation. GOVERNMENT PROCUREMENT In market economies, governments seldom own industries, and must procure goods and services from private enterprises in order to function. Government procurement contracts cover everything from pencil sharpeners to cleaning services to supercomputers. If procurement is open and competitive, governments take bids and choose the suppliers that fulfill their needs at the cheapest prices. Often this does not happen. Some foreign governments use public procurement to develop or maintain key industries and technologies. For example, the Japanese Government has procured Japanese supercomputers instead of faster U.S. machines to help develop its domestic industry (see ch. 6). The United States seldom uses procurement as a tool to promote industries or technologies. Its basic policy is to conserve tax dollars and be fair to all U.S. suppliers by buying at arms length. Accordingly, procurement law has elaborate procedures to ensure that the process is transparent, meaning that firms have adequate notice and opportunity to bid and that the selection is made according to criteria announced when bids are solicited. U.S. law allows disappointed bidders to protest; if the required procedures were not followed, a resolicitation could be required. Another policy uses government purchases to protect domestic jobs and promote various social goals. This policy is reflected in laws giving preferences to goods manufactured in the United Statesincluding Buy American laws and preferences for businesses that are small, minorityowned, or in depressed areas. National security concerns can limit purchases to U.S. goods. These laws are not meant to help commercial competitiveness, although they occasionally do--e.g., the requirement that DoD buy only domestically made supercomputers. But often these preferences do not apply, so that suppliers of foreign goods can compete on an equal footing. Many foreign procurement markets are not so open. Standard practice in many other countries is to grant notice of procurements to, and fairly consider bids from, certain domestic firms. This imbalance of access was perpetuated by international agreements. These agreements include bilateral memoranda of understanding (MOUs) dealing with defense procurements and the GATT Procurement Code. The United States is negotiating to improve access to foreign procurements, but the negotiations are difficult and even modest improvements are not assured. The imbalance of access and the fact that foreign governments often use procurement as a strategic tool to enhance competitiveness while the United States seldom does mean that U.S. and foreign procurement policies together have a negative effect on U.S. competitiveness. What the U.S. Public Sector Purchases In 1990, the U.S. public sector purchased about 9 percent of the goods produced in the United States (see table 4-9). Half of these purchases were for defense needs. 198 The defense share of public purchases of durable goods was greater: 68 percent (see figure 4-8). Durable goods include computers, semiconductors, and many other important high technology products. The effect of defense procurement on competitiveness is not clear. A large portion of defense purchases have no commercial use: military ships, missiles, military communications and electronic Table 4-9-Purchases of Goods by the U.S. Public sector in 1990 (in billions of dollars) Percent of GNP for good8 Durable Nondurable Total ($2,144bn) Federal defense a . 83.6 13.1 96.7 4.5 Federal nondefense 5.7 4.6 10.3 0.5 State and local . 33.0 53.9 86.9 4.1 Total ............. 122. 3 71.6 193.9 9.O b a This spending is almost all by the Department of Defense. A small amount, probably under 1 percent, is by the Department of Energy and NASA for defense-related work. b Figures do not add up because of rounding. SOURCE: Department of Commerce, Bureau of Economic Analysis, Government Division, computer printouts, Feb. 28, 1991, and Economic Report of the President, 1991, table B-6.
PAGE 163
158 Competing Economies: America, Europe, and the Pacific Rim Figure 4-8-Government Purchases of All Goods and Durable Goods, 1990 All goods Durable goods $193.9 billion $122.3 billion Federal defense billion % Federal nondefense $10.3 billion 5% State and local $86.9 billion 45% Federal nondefense $5.7 billion 5% ederal defense $83.6 billion 68% State and local $33 billion 27% SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, unpublished data. equipment, and combat vehicles (see figure 4-9). The largest portion (31 percent in 1990) is for aircraft. Although defense purchases of military aircraft have long generated spillover benefits to the civilian aircraft industry, these benefits are dwindling (see ch. 8, and vol. 2). Defense procurement includes some items with commercial use, such as general purpose computers and automobiles, but their procurement represents only a tiny fraction of the U.S. market for the products. For example, defense purchases of general purpose computers in 1990 were worth $1.56 billion, 199 less than 1 percent of the estimated total U.S. market sales of $176 billion. 200 Defense purchases of noncombat motor vehicles in 1990 were worth $984 million, 201 less than 1 percent of the estimated $139 billion in shipments of motor vehicles and car bodies made in the United States. 202 These relatively small quantities limit the power of defense procurement to affect commercial industries. On the other hand, defense needs for certain niche products have sometimes provided crucial early demand. Since its first sale in 1976, the U.S. supercomputer industry has benefited greatly from defense-related purchases by the Departments of Defense and Energy. For many years they were the major customers, and if not for early defense purchases and software generated in national laboratories, Cray Research Co., since 1976 the worlds leading supplier of supercomputers, might not be in the supercomputer business today. Figure 4-9-Defense Purchases of Durable Goods, 1990, $83.6 Billion Mi Spa Oth L Aircraft 31% General purose d 2 computers % Stock funds 7% Weapons 1% Y \ Military C&E 7% ~ Non-combat vehicles 1% Ships 12% Combat vehicles 3% SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis, unpublished data. Federal nondefense procurements, while more oriented toward commercial products, are relatively small. In contrast to many other developed countries, the United States runs no national health service, no national railway, no national airline, no national telecommunications system, and no national university system. Therefore its public sector does not provide as large a market for medical equipment, rolling stock, aircraft, telecommunications equipment, computers, etc. For example, U.S. Government purchases of telecommunications equipment accounted for 17 percent of the domestic market in 1989. 203 In contrast, up to 1988 roughly 90 percent of the equipment sold in the European Community was reportedly delivered to government entities. 204 State and local governments together are more important than the Federal Government as public
PAGE 164
Chapter 4--U.S. Trade Policy l 159 purchasers of nondefense goods in the United States. As shown in table 4-9, they purchased goods worth $87 billion ($33 billion for durables) in 1990. This was 4.1 percent of the U.S. GNP for goods. Indeed, State and local governments are significant purchasers of certain products. In 1982 (the latest year for which figures are available), about 24 percent of State and local government purchases of durable goods were for motor vehicles. The same proportion spent on vehicles in 1990 equates to about $7.9 billion in current dollars. 205 This is about 6 percent of the $139 billion in 1990 shipments of motor vehicles and car bodies made in the United States. The U.S. Legal Regime Buy American and Other National Preferences The Buy American Act was passed in 1933 as an unabashed protective measure, providing a barrier against goods of foreign origin in the area of Federal Government purchasing. 206 This was one of many laws introduced during the depression years to use procurement to promote employment and other socioeconomic goals. 207 The law has been amended several times and interpreted by regulations. The Buy American regime applies to goods to be used within the United States. In general, domestic goods will be favored unless they are priced more than 6 percent higher than foreign goods; 208 if the U.S. bidder is a small business or in a labor surplus area, the preference increases to 12 percent. 209 If a product is manufactured in the United States, and if the cost of domestic components going into the product is over half the total component costs, then it is domestic; otherwise, it is foreign. 210 This preference does not always apply. Two major exceptions arise under the GATT Procurement Code and bilateral defense MOUs, discussed below. Other exceptions stem from the GATT Civil Aircraft Code and the U. S-Canada and U.S.-Israel Free Trade Agreements. 211 The Department of Defense (DoD) procures under a slightly different set of rules. Products from countries that do not have special agreements with the United States can be subject to a 50 percent price preference. DoD must disadvantage foreign bids either by adding 50 percent to the value exclusive of duties, or by adding 6 percent inclusive of duties, whichever is greater. The 6 percent preference increases to 12 percent in competition with domestic bids from small businesses, or in areas of labor surplus. After this adjustment, the lowest bid wins. 212 Many Little Buy American acts grant additional preferences to particular products such as machine tools and ball bearings. Some of these preferences are absolute, forbidding the purchase of foreign products no matter what they cost. Aside from Buy American, there are some absolute preferences for products of U.S. businesses that are small, minority-owned, or located in disadvantaged areas; in fiscal year 1990, 19.9 billion dollars worth of contracts for goods and services were made under these preferences, out of total Federal contract actions worth $191 billion. 213 Defense Procurement and Memoranda of Understanding The United States has negotiated defense MOUs with most major trading partners, though not Japan, Korea, or Taiwan. About half of defense procurement is open to MOU partners on an equal basis. 214 Under the MOUs, the United States waives the general Buy American preference, 215 through not some of the specific Buy American preferences. 216 The MOUs do not affect small business, minority business, and labor surplus area preferences; they do not apply when a U.S. or Canadian source is required to preserve the defense industrial base; and they do not apply to classified military procurements that require a U.S. source for security reasons. The GATT Procurement Code and the Trade Agreements Act The GATT Procurement Code 217 and the Trade Agreements Act of 1979 eliminate the general and some specific Buy American preferences. 218 The Code requires signatories to give each others goods national treatment, treating the foreign goods the same as domestic goods. The Code currently applies only to goods; it does not cover purchases by State and local governments, or military weapons or other goods necessary to maintain national security. 219 Further, the Code covers only purchases worth at least 130,000 Special Drawing Rights (SDRs), equal to about $180,000 in January 1981. 220 Finally, the Code applies only to agencies specifically listed in a country-by-country Annex; some countries have specified other exemptions. For many countries, including the EC and the United States, agencies dealing with telecommunications, transport, energy, and water are not covered. Also, the United States
PAGE 165
160 l Competing Economies: America, Europe, and the Pacific Rim has exempted from Code coverage purchases reserved for small or minority businesses. The GATT Procurement Code is implemented in U.S. law by the Trade Agreements Act of 1979. 221 The Code Annex lists most agencies of the U.S. Government. Some exceptions are the Department of Energy, the Department of Transportation, the Bureau of Reclamations at the Department of the Interior, the Army Corps of Engineers within DoD, and, for some products, DoD. The Act grants national treatment not only to Code co-signatories, but also to countries with whom the United States has equivalent bilateral treaties, and to many developing countries on a nonreciprocal basis. 222 State and Local Procurement Like the Federal Government, all the States have adopted the rule of competitive procurement. According to a 1987 survey, 27 States have no Buy American laws on their books; 223 the other States have some Buy American laws, though sometimes they are not implemented. For example, in 1968 a California appellate court found Californias Buy American law to violate the U.S. Constitution; 224 the law has since been inoperative. Kansas law permits the State government to reject foreign bids, but this is not done. 225 States may offer preferences to small business, minority businesses, and in-State suppliers. 226 Some local governments have preferences of their own. The Effect of U.S. Procurement Law and Policy U.S. procurement law and policy in international negotiations have been of limited value in advancing U.S. competitiveness. This is because the United States does not use procurement strategically, and procurements are on balance more open than foreign procurement markets. Buy American Preferences Buy American preferences can raise the prices the government pays for goods. It is not clear whether Buy Americans benefits outweigh this and other possible costs. The 6 percent Buy American preference is intended to promote employment, but it applies with no consideration of whether additional jobs are needed. 227 Specific Buy American preferences, usually the result of heavy lobbying from industry associations or regional interests, have supported many industries: food, clothing, coal, paper, valves and piping, machine tools, aircraft ejection seats, supercomputers, etc. This targeting is not based on competitiveness strategy. It carries the danger that the protected industry will grow complacent and uncompetitive. The Buy American laws are complex, highly technical, and difficult for bidders and agencies alike to understand and apply. 228 They add complexity, cost, and delay to an already burdened Federal procurement process. 229 Occasionally, defense procurement aids competitiveness in dual-use industries. For example, the 1987 Defense Appropriations Act stipulated that none of the funds in this Act maybe used to purchase any supercomputer which is not manufactured in the United States. This restriction has probably helped the U.S. supercomputer industry, not only by assuring revenues but also by promoting the development of software for domestic machines. However, on the whole special Buy American preferences might do less good than harm (higher prices paid, inefficiencies perpetuated, and administrative burden). While reliable data are not available, 231 the effect of the general Buy American preference is likely to be modest, since defense MOUs and the GATT Procurement Code override the preference when they apply and the civilian preference is only 6 (sometimes 12) percent for nondefense goods. Buy American preferences are a lightning rod for foreign criticism as a symbol of U.S. protectionism. 232 Perhaps Buy Americans best use is as a bargaining chip to open up foreign markets. Negotiations to expand the scope of the GATT Procurement Code have tended in that direction. Unequal Access The defense MOUs, first negotiated in the late 1970s, have perpetuated the imbalance in access to government procurements. These MOUs probably cover about half of DoD procurement. Since defense accounts for about 90 percent of government procurements of goods (table 4-9), and 80 percent of Federal procurement generally (including goods and services), 233 these MOUs open a substantial part of U.S. Government procurement to foreign competition. 234 These MOUs did not open foreign government procurement to U.S. competition in the same way. To begin with, defense spending is less important in other countries. While defense procurement of
PAGE 166
Chapter 4--U.S. Trade Policy 161 goods and services is 80 percent of government procurement in the United States, it is only an estimated 25 to 50 percent of the procurement of its allies. 235 In addition, the defense MOUs were intended as defense cooperation agreements, and were not designed to open foreign procurement markets to U.S. goods. The primary goal was to secure the goodwill and cooperation of military allies. It was expected that the MOUs would reduce the U.S. surplus in defense trade, which they did. 236 While 17 of the 19 MOUs in principle were supposed to grant reciprocal national treatment, 237 they required national treatment only to the extent consistent with the other .238 this huge loopcountrys laws and regulations, hole allowed buy-national preferences to remain. These MOUs did not require foreign defense procurement to be transparent; in large measure it is not. Foreign procurements are often not well publicized. Also, the foreign governments selection criteria are often less open and predictable, with U.S. bids not guaranteed full consideration. This uncertainty sometimes deters U.S. firms from preparing bids. And foreign governments have often required U.S. prime contractors to use local subcontractors, which can hurt U.S. small businesses that cannot easily export on their own. 239 Reliable data on how much procurement these MOUs cover in the United States and elsewhere, and how much trade results in both directions, is not available. 240 However, the lesser importance of defense procurement in foreign countries and the lack of true reciprocity suggest that the MOUs have cost U.S. firms home market share without opening export markets enough to compensate. This imbalance prompted a debate within DoD in the late 1980s about whether the MOUs are to the United States 24l it also aroused congressional conadvantage, cern. 242 In 1989, the Administration began renegotiating the MOUs to achieve better access by U.S. firms to foreign defense markets. As of April 1991, MOUs had been renegotiated with France, Italy, and The Netherlands. The new terms require some procedural fairness but still do not require waivers of buy-national preferences. 243 The GATT Procurement Code might have reduced the imbalance in access to government procurements. In addition to requiring national treatment, the Code requires covered purchases to be transparent. Upcoming procurements are to be announced in advance, giving firms at least 30 days to bid, and evaluation is limited to criteria specified in advance. The Codes provisions, originally written in 1979 and amended in 1986, were modeled after U.S. practice. Recent data on procurement and trade under the Code are not publicly available because the Administration has classified it. 244 Detailed data are available only for 1981, the first year during which the Code was applied. The Administration had predicted that the Code would open $20 to $25 billion annually in foreign procurements to U.S. competition on an equal basis. However, for 1981, the General Accounting Office found that the Code covered a total of only $4.2 billion of purchases by all foreign signatories. In contrast, $18 billion of U.S. procurement was covered under the Code. 245 Thus, roughly 80 percent of the trade opportunities mandated by the Code were in the U.S. market, and only 20 percent in foreign markets. This roughly 80/20 ratio persisted through 1987. 246 This ratio contrasts sharply with the ratio of the United States total procurement to foreign signatories combined total procurement, which is at most about 50/50. 247 While it may seem that the United States gave up more than it got with the Procurement Code, the matter is not that simple. Many of the purchases covered by the Code would have been open to foreign competition anyway. Close to 90 percent of U.S. purchases covered by the Code have been for defense, 248 and a substantial portion of those have been covered under MOUs. Where MOUs do not apply, if not for the Procurement Code the Buy American regime would most often govern procurement, but even under Buy American foreign goods are sometimes purchased. Thus, it is possible that the new opportunities for U.S. exports under the Code were the same or even greater than the new opportunities for foreign countries exports to the U.S. market; there is no way to tell with the available data. Another way of assessing the effect of the Code is to look at actual purchases made under it. While the Code does not guarantee any sales, which depend on competitive bids, the level of sales is one indication of how effectively the Code has opened a given market. In 1981, $3.3 billion of foreign goods were purchased by the U.S. Government under the Code, but of that over $3 billion was for fuel and related products that were sold by longstanding arrange-
PAGE 167
162 Competing Economies: America, Europe, and the Pacific Rim ments predating the Code. 249 Only about $270 million of nonfuel-related foreign goods were purchased, or about 1.5 percent of covered purchases. Much of that might have occurred anyway under defense MOUs. For trade in the other direction, while recognizing that data were unreliable, GAO estimated that only about $210 million of U.S. goods were sold in 1981 to foreign governments under the Code (none of it fuel-related), or about 5 percent of the covered purchases. 250 In sum, the trade volumes were low, and it is not clear whether the Code contributed more to U.S. exports or imports. Startup problems might have depressed the trade in 1981. 251 Amendments effective in 1988 might have increased trade volumes in both directions. One persistent problem is that a large part of civilian procurement in the EC and other countries is made by agencies not covered under the Procurement Code, including those responsible for telecommunications, transport, energy, and water. In these sectors, which probably account for 25 to 50 percent of the EC procurement market, 252 the EC member governments favored domestic over U.S. suppliers. For example, in the last 30 years, there have been virtually no sales to the European Community of American electric utility equipment such as boilers and turbine generators, 253 even though General Electric and Westinghouse were the top two world suppliers as a percent of world sales between 1955 and 1984. 254 The only sales have been to countries with no production facilities of their own. EC members have favored national producers of telecommunications equipment. For example, in 1987, France, Germany, Italy, the United Kingdom, and Belgium on average satisfied 73 percent of their demand for digital switching equipment from domestic producers, despite substantially higher costs. 255 In the United States these sectors are largely open. Many U.S. utilities and telecommunications companies are privately owned and procure openly and competitively as a matter of course. Governmentowned utilities have also purchased substantial shares of equipment from foreign sources. For example, from 1977 to 1981 the largest Federal purchasers of heavy electrical equipment collectively spent about 26 percent of their money abroad. 256 The United States has been negotiating to bring government entities buying equipment in these sectors within the GATT Procurement Code. 257 The EC contends that if publicly held utilities in these sectors are to be covered by the Code, then privately held utilities should in many cases be subject to similar procurement requirements. The ECs position is that if: 1) there are substantial barriers to entry in the industry; and, 2) there is substantial government influence over the utility (e.g., as regulation of the utilitys operation), then there is insufficient assurance that the utility would, on its own, procure as a private firm does, on the basis of quality, cost, service, etc. (Both British Telecom and the Bell operating Companies in the United States, for example, would meet these two conditions.) The EC internally is following its own precepts: effective January 1, 1993, it will regulate procurement by privately owned utilities to prevent favoring domestic goods over the goods of other EC members. 258 The United States opposes this change, arguing that such utilities would, on their own, procure on the basis of cost, quality, etc., because that would maximize their profit, and that regulating such utilities would saddle them with a needless administrative burden. As of June 1991, the resolution of this issue was unclear. The EC also wishes to include procurements by State and local governments in the new Code. However, there is some question as to the extent of the U.S. Governments legal power over State procurement. So far, the Office of the United States Trade Representative (USTR) has made inquiries as to whether States might be willing to make a voluntary commitment to follow the Code. These commitments could then be incorporated into Federal law. 259 Also under negotiation is the coverage of services and rules to ensure that a fair hearing is given to unsuccessful bidders who complain that the required procedures were not followed. Forcing Open Foreign Markets Whether or not the GATT Procurement Code applies in a given case, the United States considers closed foreign procurement markets, like other trade barriers, unjustified. Under the Buy American Act of 1988, also known as Title VII of the Omnibus Trade and Competitiveness Act of 1988, 260 the USTR annually lists countries that discriminate in their procurement against U.S. products. The list is to include not only violations of the Procurement Code, but also any significant and persistent patterns of
PAGE 168
Chapter 4--U.S. Trade Policy l 163 discrimin ation, causing identifiable harm to U.S. businesses, by countries whose products or services are acquired in significant amounts by the U.S. Government. 261 When a country is listed, the USTR is to negotiate to have the practice changed; if the country does not stop, the President can retaliate by discrimin ating against that country in U.S. Government procurement. 262 In its first report, in April 1990, the USTR described substantial procurement problems in several countries with markets of particular significance, and noted special concern over access to the heavy electric equipment and telecommunications sectors (currently excluded from the Procurement Code) in the EC, Germany, France, and Italy. However the USTR determined that no countries met the criteria of the statute. 263 It is difficult to understand how the examples the USTR noted for special concern fail to meet criteria for listing under Title VII. The discrimin ation appears to be significant and persistent, and cause considerable harm to U.S. fins; the noted countries products and services are purchased by the U.S. Government in significant amounts. In its second report, in April 1991, the USTR listed just one country, Norway, for violating its Code obligations in its procurement of an electronic toll collection system. 264 The USTR decided again, however, not to list the EC, Germany, France, and Italy for their non-Code-covered discrimination in the procurement of heavy electrical equipment and telecommunications equipment, even though this discrimin ation remain[ed] of particular concern. Why has the Administration not used Title VII more? Partly because threatening countries with sanctions during the Procurement Code negotiations, especially for actions not inconsistent with the current Code, could antagonize them and set back the negotiations. In April 1990, the Administration believed that keeping Congress informed and promising to evaluate the progress of the negotiations in the next annual review would facilitate the negotiations in progress while keeping in reserve a prod for recalcitrant countries. 265 In April 1991, the Administration reported that progress had been made in the past year in negotiations to end the cited discrimination by expanding the Codes scope and that it would continue to monitor developments. EXPORT PROMOTION Exporting is difficult. To export, a firm must: evaluate foreign markets; learn how business is done abroad; identify, contact, evaluate, and select potential customers, agents, distributors, and/or partners; learn about foreign customers special needs and perhaps modify the product; prepare special labeling for each market and possibly special packaging as well; conduct an advertising campaign abroad (unless that is done by a local agent); and arrange for shipping, insurance, and customs clearance. exporter must also follow any regulatory requirements of the target country, and is sometimes disadvantaged by foreign tariffs or nontariff barriers. Frequently, special export financing must be obtained. And these efforts require the exporter to overcome separations due to geography, language, and culture. The U.S. Government assists exporters in these efforts. The Department of Agriculture provides a full range of services for agricultural exports. But exporters of manufactured goods must go to more than one agency to get a full range of services. The Export-Import Bank of the United States (Eximbank) assists with export financing, and many other services are provided by the Commerce Departments International Trade Administration (ITA). The activities of these and other agencies are coordinated by the Trade Promotion Coordinating Committee, chaired by the Secretary of Commerce. 266 Within ITA, the U.S. and Foreign Commercial Service (USFCs) is primarily responsible for gathering information on export markets and making it available to business. In summ er 1991, USFCS had about 200 commercial officers and 490 foreign national employees providing export counseling in U.S. embassies abroad and a total staff of about 300 in field offices in the United States. 267 USFCS emphasizes industry sectors with substantial export potential. Otherwise, USFCS does not target any particular sector for strategic or other reasons. Also within ITA, Trade Development (TD) analyzes trends by industry sector and works with industry associations to promote exports, and International Economic Policy (IEP) monitors foreign trade prac-
PAGE 169
164 l Competing Economies: America, Europe, and the Pacific Rim in fiscal year 1991. 272 But a policy of recovering costs horn firms using USFCS services has had some perverse effects. In the 1980s, USFCS published Basic Guide to Exporting, a step-by-step introductory manual on how to export. While the guide cost only $8.50 per copy, the need to collect this amount made it harder for businesses to get the book; USFCS field offices could not sell it because the revenue belonged to the Government Printing Office. The need to recover publication costs similarly crippled ITAs Associate Office Program, under which local business organizations help promote ITA products, services, and activities. 273 USFCS recent problems have not stopped with money matters. In January 1989, the General Accounting Office reported other weaknesses, includting a mismanaged trade show program and an inefficient information management system. 274 Starting in July 1989, USFCS, with a new Director General, undertook an extensive long-term strategic review of its operations. Partly through surveys, interviews, and discussion groups, USFCS sought to determine anew what business most important needs were, how USFCS could be most helpful, and what tasks were better left to other organizations. 275 Instead of trying to be all things to all people, USFCS decided to focus on the needs of the infrequent exporter-one that exported more than once per year but less than once per week. USFCS concluded that firms exporting less needed basic information that could be disseminated most efficiently by organizations such as State governments, tices and proposes strategies for opening foreign markets. 268 The United States spends proportionately much less on export promotion than do many other countries. An unreleased 1988 report by the Department of Commerce 269 compared export promotion programs in eight nations: Belgium, Canada, France, Germany, Italy, Sweden, the United Kingdom, and the United States. The report shows that in 1987 all of the other nations spent between 2 and 25 times as great a fraction of GNP on export promotion as the United States (table 4-10). The United States ranked sixth in export promotion spending in proportion to total exports. As well as spending relatively less overall, the United States allocates funds lopsidedly for agricultural rather than industrial exports compared to other countries. The United States only spends one-eleventh as much on industrial exports as on agricultural exports, per dollar of each type of export. In contrast, France spends 29 times as much on industrial exports, the United Kingdom 5 times, Belgium 1.6 times, and Italy one-ftith as much. The USFCS budget was particularly tight in the late 1980s, when the budget for foreign operations was not adjusted to compensate for the dollars declining purchasing power abroad. 270 Foreign posts had no money for market research contracts and often had inadequate libraries. Sometimes foreign service officers did not even have funds to return phone calls from firms in the United States. 271 The USFCS budget has recently improved somewhat, from $75 million in fiscal year 1987 to $108 million Table 4-10-Spending for Export Promotion in 1987 United United Belgium Canada France Germany Italy Sweden Kingdom States Central government spending . . . . . Total spending a . . . . On agricultural exports . On industrial exports . . Totai spending per $1,000 of ali exports . . . . Of agricultural exports . . Of industrial exports . . Totai spending per $1,000 GN b . . . . $1,000 Centrai Government spending . . . . Capita . . . . . $61.5m $102.lm $5.lm $97.Om $209.3m $219.3m $30.7m $188.6m $10.Om $72.4m $2.2m $70.2m $190.9m $194.lm $2.7m $191.4m $261.6m $294.Om $173.Om $121.Om $45.8m $62.9m $4.5m $58.4m $484.3m $546.8m $43.7m $503.1 m $330.lm $340.7m $2.5m $338.2m $0.71 $0.46 $0.74 $6.00 NA NA $2.18 $0.09 $2.61 $0.35 NA NA $2.00 $9.30 $1.78 $1.65 NA NA $1.45 $0.29 $1.53 $1.16 $5.95 $0.54 $0.40 $1.48 $0.47 $0.11 $0.29 $0.46 $0.43 $0.06 $1.07 $6.35 $6.02 $21.44 $1.95 $6.19 $0.68 $1.67 $0.64 $3.74 $1.33 $8.72 $0.98 $3.42 $0.29 $1.20 m-million a Total spending comprises not just central government spending, but spending from local governments, quasi-government agencies, and cooperating nongovernment agencies. b Gross Domestic product (GDP) is used instead of Gross National Product (GNP) for Canada and ltaly. SOURCE: Unreleased Department of Commerce Report, Export Prornotion Activities of Major Competitor Nations, May 1988.
PAGE 170
Chapter 4--U.S. Trade Policy l 165 industry associations, and chambers of commerce. USFCS considered that very experienced exporters could, for the most part, fend for themselves, although they needed help in removing foreign market barriers. USFCS found that industrys top two priority needs were for timely, product-specific information, and for assistance in locating and dealing with overseas agents, distributors, and end users. Accordingly, it has switched the focus of its information gathering from general information about sectors to specific information about particular products, and has worked harder to help industry with foreign contacts. An example of the latter is the Matchmaker program, by which U.S. exporters travel abroad and interview about 10 interested local firms per day. Another example, a joint effort among USFCS, TD, and IEP, is the 5-year Japan Corporate Program, for which the Commerce Department selected 18 firms (out of about 130 applicants) that showed exceptional promise for being able to export to Japan. The companies, selected in late 1990, committed to: visit Japan four times per year (at least twice per year by the firms President or CEO); print labeling and promotional material in Japanese; participate in at least one trade show each year; modify products as needed for the Japanese market; and arrange for after-sale service and maintenance. Commerce Department in return is providing marketing data and promotional support, which included the participation of Commerce Secretary Mosbacher in kicking off the program in Tokyo in April 1991. USFCS has increased both its counseling regarding options and sources for financing exports, and assistance in complying with export control regulations. USFCS is also trying to improve information flow among its headquarters, domestic field officers, and overseas posts, though this effort depends largely on an automated information system whose target completion date is September 30, 1993. The Commerce Department does not provide export financing. Often credit terms are an important part of competitiveness in export sales, especially to developing countries. All major economic powers have some governmental export credit agency (ECA) that helps to finance exports. 276 In the United States, this function is carried out primarily by the Eximbank, an independent agency of the U.S. Government, and (for agricultural products) the Commodity Credit Corporation, a corporation owned by the U.S. Government and run by the Department of Agriculture. 277 Eximbank can fill in after private sector financing leaves off, so as to provide, when possible, export financing comparable to that available for exports from other countries. While U.S. commercial banks are often reluctant to provide export financing services because they are unfamiliar with the risks in the destination country, Eximbanks specialized knowledge of these risks enables it to provide such services. The need for Exirnbank is greater now that U.S. commercial banks, burned by bad foreign debts in the 1980s, are more cautious about foreign loans. Even when banks are willing to provide services, their interest rates and other fees may be higher than the fees for government-backed export financing abroad. Eximbanks activities have cost money; at the end of fiscal year 1990, Eximbank had an accumulated deficit of $6.5 billion, financed by borrowing from the government. 278 If an exporter grants a foreign customer time to pay, Eximbank may provide insurance against the risk of nonpayment. Similarly, if a U.S. or foreign bank lends money to the customer, Eximbank can provide the bank with a guarantee against default on the loan. Insurance and guarantees help guard against both normal commercial risk of nonpayment (bankruptcy of buyer, change in market demand for buyers product, etc.) and political risks (war, expropriation, asset freezes, etc.). Eximbank guarantees can be crucial to obtainin g private sector loans, and can help firms get these loans on better terms. Eximbank can lend money itself, though it normally does so only to match particular loans offered by foreign ECA S 279 These loans may be direct to the customer, or indirect by rediscounting bank loans (lending to a bank money that the bank in turn lends at higher interest to the customer). Eximbank also makes short-term working capital loans to exporters. Eximbanks total outstanding commitments, including loans, guarantees, and insurance, cannot exceed $40 billion, though the first $25 billion of insurance and guarantees are counted only at a quarter of their value so that a total of $58.75 billion is possible. As of September 30, 1990, actual out-
PAGE 171
166 l Competing Economies: America, Europe, and the Pacific Rim standing commitments were $13.3 billion in loans and $16.7 billion in insurance and guarantees, for a total of $30 billion. There are also ceilings set each year on new commitments. For fiscal year 1990, the ceiling was $612 million for loans and $10.2 billion for guarantees and insurance. The actual commitments made were $612 million in loans (the full amount authorized) and $8.2 billion in guarantees and insurance. These amounts are modest compared to the total U.S. exports in calendar year 1990 of about $390 billion in goods and $133 billion in services. Agricultural exports receive proportionately much more government financing. The Commodity Credit Corporation made commitments of $911 million in loans and $4.5 billion in guarantees in fiscal year 1990, 280 compared with about $40 billion in agricultural exports in fiscal year 1990. 281 Congress apparently envisioned that Eximbanks financing be focused on strategically selected industries. In 1983, Congress amended the Export-Import Bank Act of 1945 to require that Eximbank provide in an annual report a detailed description of all actions which have been taken or will be taken to: l l l maintain the competitive position of key linkage industries [this refers primarily to industries that benefit downstream industries]; support industries which are engaged in the export of high value added products; support industries which axe engaged in the development of new capital goods technology; and preserve and create high skilled jobs in the United States economy. 282 However, it is not clear to what extent Eximbanks activities have been strategically focused. 283 Further, Eximbanks effectiveness is limited in two ways. First, Eximbank grants creditonaneeds basis, while European and Japanese ECAs grant credit on an entitlement basis. European and Japanese governments make broad, long-term determinations about which exports to assist and by what means, and then provide sufficient resources and administrative freedom to assist exports in all credit worthy cases within the guidelines. 284 However, in the United States, increasing budget discipline and a strong free trade, antigovernment interventionist philosophy have combined to require that applicants demonstrate on a case-by-case basis that they need the financing to close the deal. 285 The case-by-case approach increases paperwork, which along with the uncertainty regarding approval can discourage would-be applicants. Eximbanks effectiveness has also been limited in matching the very favorable terms of some foreign loans. The United States and its major trading partners have subscribed to an Arrangement on Guidelines for Officially Supported Export Credits. The Arrangement limits the attractiveness of ECA financing by specifying interest rate floors, minimum up-front cash payments, and maximum payback periods. 286 The Arrangement has largely prevented U.S. firms from being disadvantaged by superior financing by foreign ECAs. However, the Arrangement has a big loophole that lets countries combine export loans or guarantees with grants of aid for development. The combination of export financing with aid is called mixed credits or tied aid (since the aid is tied to the purchases from the donor country). It is difficult to assess the impact of tied aid on U.S. manufacturing competitiveness. An Eximbank study asked firms in five capital goods industries (telecommunications equipment, computers, electric power generating machinery, rail transportation products, and earth moving equipment) to estimate sales to developing countries that were lost because of foreign tied aid. The firms estimated that for the years 1985 through 1988, an average of between $400 and $800 million per year of sales were lost in all sectors combined. These lost sales were at least 10 percent of the combined markets of all developing countries in 8 out of 12 subsectors, and were at least 30 percent in 4 of these subsectors. However, the lost sales represented at least 10 percent of the U.S. industrys total output in only 4 subsectors 287 and in no cases represented at least 30 percent. These figures do not tell the whole story. Lost initial sales can mean lost follow-on sales, especially when compatibility is a concern. The survey responses indicated that foreign tied aid had overall decreased exports by at least 10 percent in 7 out of 12 subsectors. 288 This result is open to question. To balance its case studies, Eximbank performed a macroeconomic analysis of the impact of tied aid. The analysis failed to confirm this significant negative impact. 289 The United States ties a large part of foreign aid to purchases. However, since 1974, U.S. aid has
PAGE 172
Chapter 4--U.S. Trade Policy l 167 focused on agriculture, nutrition, health, and education, rather than the large capital projects favored by some major donor countries. From 1984 through 1987, U.S. bilateral development assistance was $15.2 billion, amounting to 34.8 percent of the aid given by the seven largest donors; but the figure drops to $1.1 billion, or 5.3 percent of the sevencountry total, when only extractive industries, power, transport, telecommunications, and other capital goods are counted. 290 Thus, tied aid programs in other donor countries offer more opportunity to manufacturing industries. This is no accident. Most other aid donors use tied aid to promote particular industries, while the United States, except for agriculture, does not. 291 Japan in particular uses foreign aid to promote selected industries. Japan grants about as much nonmilitary foreign aid as the United States. 292 However, 44 percent of Japanese aid goes to economic infrastructure, compared to only 4 percent of U.S. aid. 293 Most of Japans aid (about 60 percent in 1987) is in the form of subsidized loans, so a given amount of aid (subsidy) results in a greater amount of purchases; most U.S. aid (probably around 75 percent in 1987) is provided as grants. 294 Japan focuses its aid on Asian countries, which represent one-third of Japans trade with developing countries and whose relatively advanced economies have great potential to buy capital goods and services. The United States, on the other hand, gives the bulk of its aid to Central America and the Middle East, which represent less than a tenth of U.S. trade with developing countries.295 Japanese Government and business practices can magnify the competitiveness effect of aid, even if the aid in principle is not tied. Unlike the United States, Japan awards aid primarily based on proposals received. The requirements for proposals are so complex that often foreign governments need help from Japanese fins, which steer proposals in the direction of providing follow-on work for themselves or affiliated companies. Japan sometimes ties only the design and pl anning phase of a project. The use of a Japanese company for that phase makes use of Japanese sources for follow-on construction quite likely. 296 To minimize the adverse competitive impact from other countries use of tied aid, the United States has attempted to close or reduce the loophole that permits aid to be mixed with export credits. Effective July 1988, the Arrangement was amended, at U.S. urging, so that mixed credits must have at least a 35 percent grant component (50 percent in the case of aid to the least developed countries). The intent was to make mixed credits so expensive that their use would decline; however, this did not occur in 1989 and 1990. 297 The United States has continued attempts to limit the use of tied aid. It is negotiating for further amendments to the Arrangement, notably one to prohibit mixed credits when a project is commercially viable without a grant element. It has also gotten an informal agreement to avoid using mixed credits in Eastern Europe; as of July 1991, this agreement was working. 298 To strengthen its negotiating position, the United States has matched or even overmatched some foreign mixed credit offers, to show foreign countries that they have nothing to gain by offering mixed credits. These offers have used Eximbanks war chest of grant money, which can be combined with loan funds. In fiscal year 1990, $53.8 million in war chest grant funds were used, 299 supplemented by $30 million in grant funds from the Agency for International Development (AID), resulting in $228 million in concessionary loans. For fiscal year 1991, Eximbank had used war chest grant funds of $58.1 million by July, yielding concessionary loans of $130.8 million, though Eximbank reported that it might by the end of fiscal year 1991 use the entire $150 million in grants available, for a total of about $500 million in loans. 300 While sales have been won in this manner, the war chest sums are much smaller than those available for tied aid programs in several other countries, notably Japan, France, and Germany, whose annual loans with tied aid are measured in billions of dollars. 301 The U.S. Trade and Development Program (TDP) also offers tied aid. TDP started within AID but was spun off as an independent agency in 1981. This program funds feasibility studies and other project pl anning services by U.S. firms in developing and middle-income countries. It resembles Japans tied aid funding for pl anning and design. TDP provides project plannin g funding only for projects that are priorities of the host country and present a good opportunity for sales of U.S. goods and services. 302 TDPs budget, initially only a few million dollars, has grown in recent years. In fiscal year 1990 the $32.2 million in funds included $29.9 million for program activities and $2.3 million for administrative expenses. 303 Of program funds obligated in
PAGE 173
168 Competing Economies: America, Europe, and the Pacific Rim fiscal year 1990, 36 percent were for infrastructure, 17 percent for energy, 15 percent for manufacturing and processing, and 15 percent for communications and computer technology. In contrast to most other U.S. aid, the fast-growing Asian region received the most funding, 33 percent in fiscal year 1990. 304 TDPs fiscal year 1991 budget is $35 million. TDP reports that [e]very major donor nation offers grants for feasibility studies, and the host country almost always has the option of obtaining a grant from another donor government. Typically, these grants are made to promote procurement from the donor nation, thereby virtually locking U.S. firms out of major projects at the implementation stage. 305 Thus, the competition to determin e who performs the feasibility study is sometimes decisive in the competition for exports. As an independent agency since 1981, TDP spent $161.6 million through fiscal year 1990 in program funds. 306 As of December 1990, the associated projects had given rise to documented U.S. sales of $3.2 billion ($2.4 billion to Asia). The sales yield to date is thus 20 times the program funds spent. The annual sales have increased rapidly in recent years, and TDP estimated that activities already funded by December 1990 will provide an additional $18 billion of exports as projects mature. EXPORT CONTROLS Some items, such as computers, have both commercial and military use. The United States controls export of such dual-use items under the Export Administration Act of 1979 (EAA), as amended. 307 The Act is administered by the Commerce Departments Bureau of Export Administration (BXA), in cooperation with the Defense, State, and Energy Departments, and several other Federal departments and agencies. The Control List specifies dual-use items subject to control. This list includes physical products, software, technical data, and knowhow. 308 A license from BXA is necessary to export these items to some foreign countries. The license may be conditioned on assurances that the item will be used for civilian purposes; in some cases, BXA insists on the right of inspections or other constraints to guard against diversion to military use. BXA does not know precisely how much export trade is licensed, but $90 billion seems a fair estimate for 1990. 309 This is about 28 percent of the $316 billion 310 in total U.S. exports of manufactured goods in 1990. This percentage declined from an estimated 40 percent or more in 1985 311 and is expected to decline further in 1991. The purpose of most controls on dual-use exports is to deny militarily strategic technology to potential adversaries. This control, however, comes at a price: requiring a license can hinder commercial exports. Moreover, export controls affect primarily hightechnology industries on which U.S. manufacturing competitiveness especially depends. Some export controls on dual-use items are necessary, and some resulting drag on competitiveness is inevitable. However, there is an emerging consensus that export controls have unduly hindered competitiveness. The export control regime came under intense scrutiny with the publication in 1987 of a report by the National Academy of Sciences. 312 This scrutiny intensified for three reasons. First, political reform swept the nations of Eastern Europe and the Soviet Union. This reduced the military threat from these countries, lessening the need for controls. These nations need technology to hasten their transformation into open, economically viable societies. Sophisticated computers are needed, for example, to run a modem banking system, and fiber optics are crucial for a modem telecommunications network. If export controls prevent the transfer of such technology to these nations, they could increase the risk these nations pose. 313 Second, the changing political climate intensified allies dissatisfaction with the United States hard line position. This led to speculation that CoCom (Coordinating Committee for Multilateral Export Controls), the multilateral forum in which the United States and its allies coordinate export controls (see box 4-C), might break apart. To avert a crisis, in June 1990 the United States agreed to a great loosening of controls. Third, concern has increased over the continuing decline in U.S. manufacturing competitiveness, prompting closer scrutiny of whether the security benefits of particular controls are worth the competitiveness costs. Most observers agree that too many items were controlled under the EAA. Some items were not decontrolled even when advancing technology eliminated their strategic value. Some other controlled items, while still strategic, were available from unrestricted sources in third countries; the controls
PAGE 174
Chapter 4--U.S. Trade Policy 169 Box 4-CEast-West and North-South Controls; Cocom and Other Multilateral Export Control Forums Traditionally, most U.S. export controls on dual-use items (items with both military and commercial use) aimed to keep militarily strategic technology from Communist countries. These controls are issued as national security controls under Section 5 of the Export Administration Act of 1979, as amended (EAA). 1 They are sometimes referred to as East-West controls. East-West controls affect more than just exports to Communist or formerly Communist countries; approximately 85 percent of the individual license applications processed in 1990 were for exports to the free world (not Communist or formerly Communist countries). 2 While individual license applications for exports to the free world are almost always granted, 3 the approvals sometimes contain conditions designed to prevent possible diversion to Communist countries. Inmost cases, the United States militarily strategic technology is shared by other countries; to prevent transfer of such technology, cooperation from other countries is necessary. Internationally, East-West export controls have been coordinated by the Coordinating Committee for Multilateral Export Controls (CoCom), an informal, nontreaty organization with 17 member countries. CoCom consists of the NATO members except Iceland, plus Japan and Australia. 4 CoCom controls those dual-use items on its Core List, a scaled down version of its old Industrial List. Traditionally, the United States has been the most zealous CoCo member. While many CoCo members generally control items only to the extent required by CoCom, the United States controls many items to a greater extent. Lately, the difference between the United States and other members has narrowed. Another group of U.S. export controls is issued under Section 6 of the EAA. These North-South or foreign policy controls aim to prevent proliferation of weapons of mass destruction, including missiles and chemical and biological weapons, to countries that do not have them, especially those countries prone to support terrorism or other reckless use. 6 Such controls are not aimed at the Communist or formerly Communist nations as a whole or the Soviet Union in particular, indeed, the Soviet Union and the Peoples Republic of China possess great capability in these areas. Missile and chemical weapons technology are controlled in multilateral groups other than CoCom, but these groups have limited effectiveness because they do not include the Soviet Union. No international group controls biological weapons. 7 Foreign policy controls under Section 6 of the EAA can be imposed for other reasons as well, such as to sanction nations for human rights abuses. In that case the control is intended as a political statement, rather than to deny particular technology or products. In some cases-such as sanctions against nations that support terrorism the control has a mixture of military and political goals. 8 %0 U.S.C. app. ~. 2~~~1w of Li~ns@ Statistics: CY 1990, n.d. (prepared by BIK*U of ~rt ~ stration (BXA), U.S. Commerce Department). This pementage is by number of applications, not dollar value. The figure includes all individual license applications, not just those required by East-West controls. 3h fW~ Y= 1990,94.2 percent of such applications were approved. BXA 1990 Annual Repo~ ch. 2 (ti press). ~S figure is H on counting the Peoples Republic of China as a k world natiorE otherwise the approval rate would probably be slightly higher.) Only about 0.3 percent were denie~ the rest were returned without action because of insufllcient information or at the applicants request. 4~e complete membershi p list is Australi& BelgiW GIM@ Demark e~ y, France, Greece, Italy, Jap~ Luxembourg, the Netherlands, Norway, Portugal, Spa@ Ihrkey, the United Kingdo~ and the United States. 5~e tem tNofi+ou~~~ & a ~nvenient ~oti to fidicate tit tie ori~@on of ~~e controls is fi~t fiOIIl ht Of ~t-w~t controls. It should not be taken to imply that all or most developing countries give cause for concern. 650 U-S*C. app. ~5. Nu~l~ W=wm ~ con~oll~ by ~o~~ s~tute, f$~tion 309(c.) of tie NUCIW Non-Proliferation Act of 1978, codifkd at 42 U.S.C. 2139a(b). See Section 17(d) of the EAA at 50 U.S.C. App. 2416(d). 7M~ti~te~ cwr~tion of Nofi+ou~ con~~ is discus~ in p~el on tie fiture ~si~ @ Implementation of U.S. National Security Export Controls, Committee on Science, Engineering, and Public Policy, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, Finding Common Ground: U.S. Eqort Controls in a Changed Global Environment (Washington DC: National Academy Press, 1991), pp. 83-85. 8Expofi con~o~ on d~.u~ item cm b be imposed under Section 7 of the EAA during a Amtage of domestic SUPPIY. did not deny technology to the target country but license was submitted and approved. In practice, simply diverted business to foreign suppliers. however, the time and trouble involved discouraged some firms from applying. And even if permission In principle, controlled items could still be to export were ultimately granted, the delay could exported, provided that an application for an export cost the sale. Such delays have been used as a selling
PAGE 175
170 l Competing Economies: America, Europe, and the Pacific Rim point by competing foreign manufacturers who suffered no such hindrances. In addition, export controls occasionally stopped U.S. firms from providing repairs and spare parts to items already in the field; this made U.S. suppliers appear especially unreliable. Another thorn in the side of foreign customers was reexport controls. If controlled U.S. technology or components were incorporated abroad into finished goods, sometimes permission from the U.S. Government was necessary to export those goods to a third country. As a result, many foreign manufacturers avoided using technology and components that originated in the United States. Some U.S. allies resented this extension of U.S. law onto their soil. 314 In recent years this situation has improved. Reform has focused on East-West controls, traditionally the most common type, that are designed to keep technology from former Communist bloc 315 countries (see box 4-C). Reform has proceeded on three interdependent tracks: multilateral negotiations, legislation by Congress, and domestic action by the Administration. In CoCom, the United States has agreed to lessen controls, and other countries therefore appear more willing to maintain not only common rules but a uniform standard of enforcement; this should reduce the advantages to foreign firms resulting from less restrictive control regimes. The Export Enhancement Act of 1988 (1988 Act) 316 amended the Export Administration Act of 1979 to prod the Administration to remove unnecessary controls and streamline the license application process. Further legislation in this direction is under consideration. 317 The Administration has been reducing U.S. controls to track reductions in CoComs controls, and over time (sometimes more time than Congress allowed) has been implementing changes mandated in the 1988 Act. Despite the reforms, export controls remain a live issue. In two sectors, computers and communications, many within industry view the Core List reductions as insufficient. Nor is it clear that procedures and attitudes have changed enough to avoid licensing delays, permit timely decontrols in the future, and facilitate increased governmentindustry cooperation. While reform proceeds on East-West controls, broad use of North-South controls could reverse some of the progress made. North-South controls are implemented for two purposes: to prevent proliferation of weapons of mass destruction; or to further other foreign policy goals, such as to show disapproval of a countrys human rights abuses or support of terrorism (see box 4-C). The recent Gulf War has increased proliferation concerns, sparking the Administrations Enhanced Proliferation Control Initiative. North-South controls are largely unaffected by the recent reforms, and could lead to substantial disruption of legitimate commercial sales. Reducing East-West Controls' Drag on Competitiveness What Is Controlled The United States controls exports of dual-use items bound for either former Communist bloc countries or other countries where risk of diversion to former Communist bloc countries is substantial. This second case has greater commercial impact; in 1990, 85 percent of all license applications were for exports to the free world (not former Communist bloc countries). 318 In fiscal year 1989, computers accounted for 42 percent of the dollar value of all license applications; other electronic equipment and aircraft also accounted for large dollar values. 319 These East-West controls are based largely on CoComs Core List, which replaced the old Industrial List effective September 1, 1991. The Core List includes items in nine categories: advanced materials; material processing; electronics; computers; telecommunications and information security; sensors and lasers; navigation and avionics; marine; and, propulsion. 320 CoCom controls exports at three levels, depending on the item and the proposed destination. At the highest or general exception level, un animous approval by CoCom members is necessary. At the next level, favorable consideration, there is a presumption of approval; the export may be made if no CoCom member objects within 30 days of submission to CoCom. At the lowest level, national discretion (also called administrative exception), a member nation may approve the export on its own, but CoCom must be notified after the fact. In May 1990, following a study of export controls, the Bush Administration found that the Industrial List contained items not strategically critical to U.S. interests. 321 In a June 1990 CoCom meeting, the United States, responding to pressure from its allies, proposed to scrap the Industrial List and build from scratch a new, much shorter Core List containing only truly strategic technologies. 322 As
PAGE 176
Chapter 4--U.S. Trade Policy l 171 an interim measure, the Industrial List was trimmed in summer of 1990, especially in the computers, telecommunications, and machine tools sectors; a revised Core List was agreed upon on May 24, 1991, to take effect September 1, 1991. Controls were lifted on many important items, including satellite ground stations for television, weather, and civil communication. Machine tools were decontrolled to up to accuracies far greater than before. 323 Many personal computers based on Intels 80386 processor chip, which were available throughout the world, were no longer subject to controls as of July 1990, 3X and some more advanced versions based on Intels 80486 chip, also available from unrestricted sources, were decontrolled in May 1991. The significance of the reduced Core List is not only that fewer items are controlled, but also that U.S. and foreign controls will more closely resemble one another. It is widely believed that in the late 1980s many other CoCom members were chafing under controls retained only at the United States insistence, and were sometimes enforcing controls loosely. With the U.S. agreement to eliminate or reduce most of the controls, other members appear headed toward more uniform enforcement. CoCom members have agreed on standards for an effective export control regime, and as of June 1991 BXA anticipated that these standards would be in force in all member countries by the end of 1991. 325 Uniform enforcement standards would mean that the United States could trust the export controls of other CoCom members. Current East-West controls require licenses for some items even when shipped to CoCo destinations. The United States imposes these licenses because it does not fully trust its allies to prevent reexport in violation of CoCom rules. Those few cases in which BXA requires licenses for intra-CoCom trade tend to be cutting edge items, such as the latest computer models, which can be important for winnin g new customers. For years industry has wanted BXA to eliminate the license requirement for intra-CoCom trade. In the 1988 Act, Congress mandated that the Commerce Department annually take stock of which other countries have fully satisfactory export control systems, and eliminate license requirements (under the East-West controls) for exports to these countries. 326 The Administration has not yet found that any countries meet the requirements. In November 1989, President Bush promised to remove intra-CoCom East-West license requirements by June 1991; 327 the target date for the new enforcement standards has since been revised to the end of 1991. 328 Previously, U.S. East-West export controls covered many items not controlled by CoCom, thus putting U.S. producers at a disadvantage relative to producers in many countries that rarely controlled such items. 329 The 1988 Act provided that, with certain exceptions, East-West controls must not be imposed unilaterally for over 6 months. This provision led to decontrol and reduced control of many items in February 1989, 6 months after the 1988 Amendments became law. 330 This should lead to greater parity between U.S. and foreign export controls. The 1988 Act tightened provisions that items must be decontrolled to particular destination countries when they are in fact available to those countries in sufficient quality and quantity to render U.S. controls ineffective. 331 The 1988 Act requires the Commerce Department to rule on alleged cases of foreign availability within 4 months and to publish its determination within 1 more month; if CoCom permission is necessary to export the item, 8 months are allowed for the ruling instead of 4. 332 Under this provision, BXA completed 17 such assessments by May 15, 1991, and found foreign availability in 12 cases. In eight of those cases the item was decontrolled; in one case the item remained controlled based on by the Presidents finding that decontrol would be detrimental to national security, and the decision was pending in three cases. 333 Another development is the effort to induce nations outside of CoCom to institute similar export controls. This will enhance the effectiveness of U.S. and CoCom controls, put U.S. exports more on a par with exports from the other countries, and permit trade with those countries under lessened controls. CoCom has approached countries for this purpose, including those of Eastern Europe. The United States has for years conducted bilateral negotiations to encourage other countries to control exports, resulting in more liberal treatment of exports to Austria, Finland, Ireland, Singapore, Sweden, Switzerland, and South Korea. Other nations, such as Taiwan, have not yet been responsive. 334 A further reform is the reduction of reexport controls. The 1988 Act removed East-West controls for the reexport of items to a destination within CoCom or with similar safeguards. The Act lifted all East-West controls for the reexport of U.S. goods
PAGE 177
172 Competing Economies: America, Europe, and the Pacific Rim and technology incorporated abroad into foreign products, provided that either the controlled U.S. content comprises at most 25 percent of the products total value, or CoCom controls the parts at only the national discretion level. 335 The Commerce Department implemented these changes in July and October of 1989, well after the November 1988 deadline given by Congress. 336 Moreover, BXA did not completely implement this provision because it believed that it would endanger national security, for example, by decontrolling sensitive avionics. BXA changed the 25 percent threshold to 10 percent for reexport to certain countries, including Iran and Libya, and did not modify an existing rule that controlled reexport of U.S. technology incorporated abroad into finished products regardless of the percentage value the technology represented. As a result of these and other reductions in East-West export controls, the amount of licensed trade has been decreasing. The number of applications received for individual licenses dropped from 98,000 in fiscal year 1988 to 85,000 in fiscal year 1989,65,000 in fiscal year 1990, and an annual rate of about 40,000 for the first 6 months of fiscal year 1991.337 The value of individual license applications processed dropped from $132 billion in 1989 to $73 billion in 1990, and the value of applications approved dropped from $123 billion in 1989 to $63 billion in 1990. 338 The new Core List and other recent or pending changes will further decrease the amount of licensed trade. Licensing Procedures In the midand late 1980s, acquisition of export licenses took much time and effort. For the first quarter of 1986, the Commerce Department reported an average processing time of 27 days, with some applications taking several months and a few even taking years. 339 However, a survey of U.S. industry reported an average of 54 days; the difference in the averages is that industry counted time spent preparing applications, time after the applications were sent but before the Commerce Department logged them in, and time after approvals were sent but before they were received. 340 The view from European firms that needed U.S. approval for purchases or reexports was particularly negative. European industry representatives reported widespread disgust with this system, citing: lost applications; technically incompetent questions; and delays caused by the use of surface mail. License turnaround times in other CoCom countries were generally much shorter. 342 When the Commerce Department wanted to approve a license but DoD did not, the interagency dispute resolution process was cumbersome. 343 Both the time required and the nature of the process discouraged industry. The regulations were complex and difficult to fathom; firms seldom got advance indications from the government of the likelihood of approval and the likely delay; and companies were sometimes kept in the dark about concerns delaying license approval, making them unable to help resolve the problem. 344 In one instance, it took a U.S. company almost 3 years to get a license to sell a $450,000 nuclear magnetic resonance (NMR) spectrometer to a medical research unit in Eastern Europe; in the interim a German firm sold several similar systems to Communist bloc customers. 345 The U.S. practice contrasted with that of Japan, where company representatives work[ed] closely with their government counterparts and [a]s a result, export license applications [were] rarely submitted if they [were] not virtually certain to be approved." 346 Similar government-industry consultation occurred in some other CoCom countries. 347 Industry was not told of likely U.S. or CoCom decisions to reduce or eliminate certain controls; advance w arning could have allowed a head stint in exploring new export markets. Other CoCom members kept their industries much better informed. 348 This picture has changed. Average processing times for individual licenses, as reported by the Commerce Department, have decreased from 23 days in fiscal year 1986 to 16 days in fiscal year 1990. 349 This decrease is significant because during this period, with the decontrol of lower levels of technology, the applications complexity and sensitivity increased. Applications for exports to CoCom members take the least time-an average of 3 days in calendar year 1990. Much delay comes from the need for the Department of Commerce to refer applications to other agencies and/or CoCom. In calendar year 1990, applications not requiring referral took an average of 7 days for processing; those requiring referral averaged 67 days. 350 BXA has made the application process more user-friendly. BXA opened branch offices, gave many seminars to industry, and now provides extensive counseling to exporters. Urgent applications get special handling, sometimes gaining ap-
PAGE 178
Chapter 4--U.S. Trade Policy l 173 proval in a few hours. A computerized tracking system permits instant determination of an applications status, and BXAs automated phone-in system gives status updates 24 hours per day, including authorization to ship if the license was approved with no restrictions. Exporters may submit applications by computer (accounting for 29 percent of the applications filed in fiscal year 1990); approvals for those applications are returned electronically. 351 BXA consults with applicants if problems arise in interagency review and tries to work out a solution acceptable to all parties. 352 Remaining Concerns Some concerns remain despite recent progress. One issue is whether there is sufficient political will to overcome bureaucratic gridlock. Traditionally, the Defense Technology Security Administration (DTSA), the part of DoD with primary responsibility for export controls, has been hostile to liberalization. This is to be expected up to a point, since DoDs mission is military security rather than economic competitiveness. However, there is a consensus that in the midand late 1980s the DTSA went too far; it did not take seriously the need to avoid unnecessary drag on competitiveness and often blocked clearly justified liberalization. DTSA stalled for almost a year a 1988 Commerce Department recommendation to decontrol personal computers compatible with IBMs AT models that were based on Intels 80286 processor. These machines, at the time midrange PCs, were available from many producers in seven non-CoCom nations (including Korea and Taiwan, which produced over 1 million units in 1988) and were in wide civilian use in Soviet bloc countries. 353 A combination of circums tances overcame DTSAs resistance in 1990, when President Bush personally backed a drastic reduction in controls. However, many in industry worry that DTSA can impede the actual implementation of promised reforms. This may have happened already. The new Core List, and the U.S. proposals for it, were supposed to be written from scratch 354 with each item specifically justified. However, when it came time to draft the U.S. proposals in meetings of BXAs Technical Advisory Committees (TACs), industry representatives report that some DTSA representatives refused to do so and instead sought only to trim existing lists. 355 With the 1988 Act, Congress strengthened the Commerce Department vis--vis the Defense Department. DoD can no longer delay license approvals indefinitely. When the Secretary of Commerce decides to approve a license, the Secretary of Defense has 20 days to object, and the license may be granted unless the President intervenes in another 20 days. 356 However, interagency coordination below the cabinet level could still be cumbersome, taking over 100 days. 357 CoCom approval might require additional time. The detailed procedures are not publicly known because an Executive Order setting them out is classified. Another possible problem in implementing reform is that dual-use items removed from the Control List might be added to the Munitions List, which is supposed to govern only purely military items. The control regime for items on the Munitions List is much stricter, without the EAAs competitiveness safeguards. While President Bush promised to remove all overlap between the lists unless significant U.S. national security interests would be jeopardized, 358 here is some danger that the Munitions List will be used to maintain or institute controls over dual-use items that could not be justified under the EAA. Another concern is the new Core List. In two important sectors, computers and telecommunications, industry is not completely satisfied. In 1990, the computer sector accounted for 41 percent of the value of all individual license applications. 359 Modern telecommunications systems for the Soviet Union and Eastern Europe will cost many billions of dollars. Hungary plans to spend $6.3 billion over 10 years, Czechoslovakia $3 billion through 1995, and Germany $4 billion on East German upgrades in 1991 alone. 360 The Core List controls computers that these nations need for economic development. In one broad performance range, the United States wanted to control the computers, but most allies did not. As a compromise, the Core List controls them at the national discretion level. Most other CoCom members will probably permit those computers to be exported with either no prior application or a very quick one; some in industry worry that they will suffer the disadvantage of long licensing delays. Current Administration policy is to process applications for items at the national discretion level within 15 days, without DoD review. 36l However, even if BXA keeps to 15 days, that delay, coupled with the need to apply in the first place and the lack of certainty that the application will be approved, could cost some sales.
PAGE 179
174 l Competing Economies: America, Europe, and the Pacific Rim The Core List controls telecommunications equipment these nations need. This includes fiber optic cable for state-of-the-art telecommunications systems, which the Core List controls at the general exceptions level. The United States pushed for that result, but many other members were quite unhappy and will seek substantial loosening of controls in 1992. Poland, Czechoslovakia, and Hungary receive favorable consideration. As long as controls affect all suppliers equally, U.S. companies will suffer no competitive disadvantage. But it is possible that foreign governments will try harder than the United States to obtain CoCom approval on behalf of their fins. There is concern as to whether an effective mechanism exists for updating controls. Todays current list of controls can become tomorrows obsolete list. More timely review will be easier because the U.S. list now has fewer items; it is now within the bounds of CoComs Core List, which is much smaller than the old Industrial List. U.S. reviews will better track CoCom routine reviews, which will occur every 3 years instead of every 4 years as in the past. 362 Also, industry can now better funnel information about advancing technology to the government through the TACs. But list review requires considerable effort and expertise. The basis for decontrol is often a determination that an item is available in sufficient quantity and quality from unrestricted sources. 363 A delay in this determination results in an ineffective control that hinders U.S. exports while foreign manufacturers get the business. These determinations, performed by BXAs Office of Foreign Availability (OFA), require investigating the sources, quality, and quantities of foreign goods; such information can be hard to find. Further, the technologies involved are specialized and complex. OFAs staff of about 25 364 -out of about 530 BXA employees 365 seems too small for such an important and difficult job. The salary levels authorized for OFA--GS-13 (starting salary about $44,000 as of Jan. 1, 1991) for most working engineers 366 -make it difficult to keep top-notch people. In addition, these politically charged determinations are sometimes opposed by other agencies at high levels. One example is the case of AT personal computers already mentioned; another is semiconductor wire bonders. The Commerce Department found foreign availability of wire bonders and recommended decontrol, which was approved by the President in 1987, but was blocked into 1990 by interagency dissension. 367 In 1991, a National Academy of Sciences panel found that BXAs procedure of finding foreign availability has proven largely ineffective, and has been costly and contentious and has rarely resulted in timely decontrol. 3@ A complementary approach to avoiding outdated controls is indexing. For example, if a particular control applies to all computers above a certain processing speed, there might be a presumption, based on projected industry trends, that each year the processing speed threshold be increased by 10 percent. Each subsequent year, those opposing the adjustment would have to justify their position. As another example, there could be a presumption that personal computers (or some other type of equipment) would be decontrolled (or controlled at a lower level) after the model in question has been sold commercially for a certain number of years. Again, those who opposed a particular decontrol would have to justify their position. While Congress has encouraged indexing, 369 so far the Administration has rarely used it. Some in industry are concerned that the government does not welcome its participation and cooperation in administering export controls. In 1991 a National Academy of Sciences panel recommended increased cooperation. 370 BXA did involve industry TACs in drafting and negotiating the Core List. The TACs advised on appropriate technical thresholds for control; industry representatives occasionally spotted foreign proposals that were apparently crafted to decontrol items made by foreign firms while leaving controlled similar items made by U.S. fins. It remains to be seen how BXA and industry will build on this experience. According to some analysts, BXA and industry have been in a vicious circle: BXA not taking the TACs seriously, and industry not devoting sufficient resources to them. One industry observer commented that because the industry representatives are busy corporate officers who barely have time to fly to Washington for a one-day meeting, the TACs can work well only if the Commerce Department does considerable preparatory and follow-up work: preparing memoranda setting out issues to be discussed; writing draft position papers based on the discussion at meetings; gathering facts; and soon. This observer commented that the National Academy of Sciences proceeds in this manner with its panels of experts, but the
PAGE 180
Chapter 4--U.S. Trade Policy l 175 Commerce Department lacks the resources to do this. 371 Industry representatives have had complaints about the use of TACs. For example, a representative of one high-technology firm with experience on several TACs stated that some parts of TA C meetings are needlessly classified, preventing the representatives from reporting back to industry. 372 An experienced representative on the TAC for automated manufacturing stated that DTSA representatives often will not give reasons for their positions, even in a classified session, so that industry cannot address DTSAs concerns. 373 Som e felt that the government used TACs to legitimate its policies: it would pose a limited set of choices, none of which appealed to industry; the industry representatives would make their choice; and then the government would claim that industry had approved that choice. Controls for Reasons of Foreign Policy While much progress has been made in reducing the drag from East-West export controls on commercial exports, that progress could be reversed by the use of North-South or foreign policy controls. Some foreign policy controls are aimed at preventing proliferation of weapons of mass destruction, missiles to carry them, and chemical and biological weapons. Other foreign policy controls, such as those in reaction to human rights abuses, and those against Libya, Syria, and Iran, are intended primarily to make a political statement rather than deny military technology. 374 As explained in box 4-C, Section 5 of the EAA governs East-West controls, while Section 6 governs foreign policy controls. Section 6 controls must be justified annually, but they are not subject to the Section 5 safeguards against unnecessary interference with legitimate commercial exports. The Section 5 prohibitions against unilateral controls and against controlling items shown to have foreign availability do not apply to controls under Section 6. 375 Th u S section 6 can provide an end run around the restrictions in Section 5. For example, in March 1991, BXA imposed unilateral controls for chemicals and chemical manufacturing equipment that could be used to make chemical weapons. 376 A s another example, in June 1991, BXA removed Section 5 controls on certain equipment used to manufacture prepregs (fibers embedded in resin used to make, for example, tennis racquets and aircraft structural components), because of foreign availability. However, the equipment remained subject to foreign policy controls. 377 In August 1991, BXA issued regulations under Section 6 that prohibit all exports that the exporter knows will be used to make chemical or biological weapons. 378 This rule might be interpreted by BXA and the courts to imply a duty on the exporters part to make a reasonable inquiry as to where the exports will be used; if the exports are used in making the prohibited weapons, and the exporter could have discovered that in advance with reasonable inquiry, then the exporter might be deemed in violation of the act and suffer stiff penalties. To be safe, any firm that exports virtually any goods or technology to any country might need to set up a monitoring and control system with careful communication between headquarters and marketing. 379 BXA can inform exporters that exports to a particular consignee require a license because of risk of diversion to a prohibited weapons plant. This gives BXA broad power to prohibit exports of any items to particular destinations. This regulation controls exports unilaterally, with no consideration of whether items have strategic importance and whether they are available from non-U.S. sources. These rules have been fashioned within the rubric of the Administrations Enhanced Proliferation Controls Initiative (EPCI), announced December 13, 1990 380 Heightened concern for proliferation is natural in the wake of the Gulf War and revelations about prior exports to Iraq. However, industry is concerned that the Administration is starting an open-ended export control campaign without serious consideration of the effect on commercial trade. 1 U.S. Congress, Office of Technology Assessment, Making Things Better: Competing in Manufacturing, OTAlTE-443 (Washington DC: U.S. Government Printing Office, February 1990), passim. Making Things Better did not consider domestic policies that affect competitiveness incidental to their main purpose--for example, environmental regulation regulation of workers health and safety, and product liability. 2 Gary Clyde Hufbauer, Diane T. Berliner, and Kimberly Arm Elliott, Trade Protection in the United States: 31 Case Stud-es (Washington, DC: Institute for International Economics, 1986). This report uses the term special trade protection to mean protection under Section 201 and related sections of the Trade Act of 1974, as amended. This protection is based on GATTs so-called escape clause, which permits countries to protect industries seriously injured by foreign competition even in the absence of any unfair trade practice. 3 Labor-Industry Coalition for International Trade, The Uruguay Round: Will It Be a Good Deal for U.S. Manufacturing? June 1990, p. 1. 4 Complete specialization was considered unlikely because of the
PAGE 181
176 l Competing Economies: America, Europe, and the Pacific Rim diminishing marginal productivity of factors of production. This means that, after the point where inputs are combined optimally, any input (in this case, capital and labor) contributes in d iminishing quantities to productivity the more of that input is used. 5 For a more detailed but still simple, explanation of how this works, see David B. Yoffie, International Trade and Competition: Cases and Notes in Strategy and Management (New York NY: McGraw-Hill Publishing Co., 1990), pp. 4-8. 6 The contributions of Linder, Vemon, and Wells are discussed in Yoffie, ibid. 7 Yoffie, op. cit., p. 13. 8 Paul R. Krugman, Rethinking International Trade (Cambridge, MA: The MIT Press, 1990), p. 1. 9 The two commercial aircraft makers in America are Boeing and McDonnell Douglas; the European manufacturer is Airbus, a consortium with four member companies. The engine makers are Rolls Royce of England, and GE and Pratt& Whitney of the United States. In addition, SNECMA of France, MTU of Germany, and Mitsubishi Heavy Industries, Kawasaki Heavy Industries, and Ishikawajima-Harima Heavy Industries of Japan make major subassemblies of engines. See ch. 8 for a description of the worlds large commercial jet aircraft industry. 10 W. Brian Arthur, Positive Feedbacks in the Economy, Scientific American, February 1990, p. 93. Approximate dollar figures are OTA estimates for 1991. 11 Krugman, op. cit., p. 164. 12 Infant industry protection is defined as the protection of a fledgling group of producers from incursions of foreign firms that dominate production and have superior technology in industries with high knowledge intensity, significant positive externalities or increasing returns. 13 Arthur, op. cit., p. 98. 14 Ernest H. Preeg, The American Challenge in World Trade: U.S. Interests in the GATT Multilateral Trading System (Washington, DC: The Center for Strategic and International Studies, 1989), p. 4. 15 Ibid., p. 23. 16 Hufbauer et al., op. cit., pp. 250-251. These limits were exceeded every year of the VRA, rising from actual Japanese imports of 1.78 million units in 1981 to 2.31 million units in 1985. 17 Occasionally limitations on imports are more stringent. In 1981, at the beginning of the negotiation of the third phase of MFA, the European Economic Community cut back the volume of imports from t MFA suppliers (Hong Kong, Macao, Korea, and so-called dominan Taiwan) by nearly 7 percent. Thomas R. Howell and William A. Noellert The EEC and the Third Multifiber Arrangement: A Study Prepared for the Fiber, Fabric and Apparel Coalition for Trade (FFACT) (Washington, DC: Dewey, Ballantine, Bushby, Palmer & Wood, 1986), p. 7. 18 This means that all GATT signatories are to be given the same favorable terms of trade as those given to the nation with the best terms, or the most-favored nation. 19 Preeg, op. cit., p. 31. 20 Some experts in both the United States and the EC argue convincingly that these countries are considering standards other than those used by Japanese producers is for technical rather than nationalistic reasons; but among policymakers, the concern is at least as much for national competitiveness as for technical superiority. 21 U.S. Department of Labor, Bureau of Labor Statistics, BLS Reports on Worker Displacement News Release, Dec. 9, 1988. 22 Michael Podgursky and Paul Swain, Labor Market Adjustment and Job Displacement: Evidence from the January, 1984 Displaced Worker Survey, Final Report, Bureau of International Labor Affairs, U.S. Department of Labor, January, 1986. 23 See chs. 6 and 7, on the industrial policies of Japan Korea, and Taiwan, for further discussion. 24 19 U.S.C. 2411-2420. 25 19 U.S.C. 2251-2254. 26 19 U.S.C. 2411(b); see also 19 U.S.C. 2411(a)(l). 27 GATT Article 12, paragraph 1. 28 GATT allows agricultural quotas in many cases; it also allows quotas by countries with dangerously low foreign currency reserves and for reasons of national defense. GATT Article XI, paragraph 2; Articles XII, XXI. Quotas are also permitted under GAITs so-called escape clause (Article XIX, though in this case the issuing country is obliged to pay compensation. Under GATT's Multifiber Arrangement, many bilateral agreements limit trade in textiles. Finally, sometimes two countries reach an understanding to limit bilateral trade in specified products, even though this understanding is arguably prohibited under GATT. 29 This term, as used here and often elsewhere, refers to barriers that are not tariffs and also are not quotas. Also, as used here, the term refers to both public (government-imposed) and private barriers; some use the term to mean only public barriers. One type of barrier is the failure of foreign counties to protect intellectual property (by means of patent rights, copyrights, trademark rights, trade secret rights, etc.). If a foreign government protects intellectual property, U.S. inventors, authors, etc. will in many cases have the exclusive right to make and sell particular goods in that country. If a foreign government protects intellectual property poorly or not at all, U.S. inventors and authors can lose sales to foreign imitators. Therefore, the lack of protection can be considered a trade barrier, since it hinders sales of U.S. goods abroad. This trade barrier was discussed in U.S. Congress, Office of Technology Assessment Making Things Better, op. cit., pp. 69-71, 211-219. The justifications for protecting intellectual property are discussed in that report and in U.S. Congress, Office of Technology Assessment Intellectual Property Rights in an Age of Electronics and Information, OTA-CIT-302 (Springfield, VA: National Technical Information Service, 1986). 30 See, for example, Office of the United States Trade Representative, 1991 National Trade Estimate Report on Foreign Trade Barriers (Washington DC: U.S. Government Printing Office, Mar. 29, 1991); Office of the United States Trade Representative, Report to Congress on Section 301 Developments Required by Section 309(a)(3) of the Trade Act of 1974 (January-June 1990), n.d. (one of a series of semiannual reports). 31 President Ronald Reagan memorandum of Aug. 1, 1986, Dete rmination Under Section 301 of the Trade Act of 1974, 51 Federal Register 28219 (Aug. 6, 1986). 32 Semiconductor Industry Association Four Years of Experience Under the U.S.-Japan Semiconductor Agreement: A Deal Is a Deal, Fourth Annual Report to the president November 1990, pp. 35-38; Clyde V. Prestowitz, Jr., Trading Places: How We Allowed Japan To Take The Lead (New York, NY: Basic Books, 1988), pp. 62-63; Donald Phillips, Assistant United States Trade Representative, testimony at hearings before the House Committee on Energy and Commerce, Subcommittee on Commerce, Consumer Protection and Competitiveness, May 9, 1990, Serial No. 101-149, pp. 8-9; Semiconductor Industry Association, World Market Sales and Shares for 19821990, n.d. 33 U.S. firms market shares generally declined over the years. As late as 1985, U.S. firms still held 86 percent of the U.S. market, 54 percent of the European market, and 40 percent of the rest of the world market outside of Japan. Semiconductor Industry Association World Market Sales and Shares for 19821990, n.d. 34 The text of the agreement is contained in an exchange of letters on Sept. 2, 1986, between Japans Ambassador Matsunaga and USTR Clayton Yeutter. 35 Prestowitz, Jr., op. cit., pp. 65. (The quoted language is a paraphrase by Prestowitz of the letters contents.)
PAGE 182
Chapter 4--U.S. Trade Policy 177 36 52 Federal Register 10275, (Mar.31, 1987). 37 52 Federal Register 13412, (Apr. 17, 1987). Some other punitive tariffs were levied at the same time in retaliation for Japans failure to stop dumping; in the September 1986 agreement, Japans Government had also promised to stop its firms from dumping in the U.S. and third country markets. 38 Semiconductor Industry Association, World Market Sales and Shares for 1982-1990, op. cit. 39 Ibid. The USTR gives slightly different figures for the share of imports from countries other than the United States. See Office of the United States Trade Representative, United States, Japan Conclude Semiconductor Agreement, press release dated June 4, 1991, ref. No. 91-21, attachment entitled Background Statistics on Semiconductor Sales to the Japanese Market. 40 Jacob M. Schlesinger, Foreign-Chip Shine in Japan is Below Goal, The Wall Street Journal, June 17, 1991. 41 Office of the United States Trade Representative, United States, Japan Conclude Semiconductor Agreement, press release &ted June 4, 1991, ref. No. 91-21. MITI claimed that the imports market share had in fact risen to about 17percent in the first quarter of 1990 and 19 percent in the fourth quarter. MITI used a different measure of market share than the U.S. Government and industry, based on its surveys to 60 Japanese firms. The U.S. approach is based on statistics collected by the World Semiconductor Trade Statistics (WSTS) Program. The two measures differed in three ways. First, MITTs survey included only about 75 percent of the industry, while WSTS statistics were more comprehensive. Also, MITI counted, but the U.S. approach excluded, 1) captive sales-that is, in-house sales of a chip of a type that is not offered for sale to outsiders, and 2) private label sales-that is, sales of a chip made by a Japanese firm with a U.S. fros label. The U.S. industry maintains that the Japanese industry agreed in 1986 to use the U.S. approach. In the new agreement, MITI agreed to drop using its survey and accept the use of WSTS statistics; the two countries still disagree on the other two differences in measuring approach. Under this new framework MITI claims that import penetration reached about 16 percent in 1990, compared with the U.S. figure of about 13 percent. Sources that discuss the m easurement issue include Semiconductor Industry Association, Four Years of Experience, op. cit., Appendix 8; Jacob Schlesinger, op. cit.; U.S. and Japan Sign Semiconductor Pact; Targeting 20 Percent Share, Design-Ins, Regulation, Economics and Law, June 5, 1991 (Bureau of National Affairs, Inc., Washington DC), p. A-15 (substantially the same article is in BNAs International Trade Reporter, June 5, 1991); Tsuyoshi Sunohara Chip Pact Dispute Shifts to Market Share Calculation, Japan Economic Journal, July 14, 1990, p. 1. 42 Ch.6; U.S. Congress, General Accounting Office, U.S.-Japan Trade: Evaluation of the Market-Oriented, Sector-Selective Talks, NSIAD-88-205 (Gaithersburg, MD: U.S. General Accounting Office, July 18, 1988), p. 45; Cray Research, Inc., The Japanese Public Sector: Problems and Prospects for U.S. Supercomputer Vendors (May 1990), pp. 3-1 through 3-23,4-1 through 4-8,5-1 through 5-3. 43 Chapter 6; U.S. Congress, General Accounting Office, ibid., pp. 45-46; Cray Research Inc., ibid., p. 4-10. 44 Chapter 6; Office of the United States Trade Representative, Report to Congress on Section 301 Developments Required by Section 309(a)(3) of the Trade Act of 1974 (January-June 1990), n.d., pp. 5-6; Cray Research Inc., ibid., pp. 4-11 through 4-18. 45 Office of the United States Trade Representative, ibid., pp. 5-7. 46 This course of action is available only if the subject in question is covered by GATT. For example, while the United States is seeking in the Uruguay Round to have GATT require certain levels of intellectual property protection GATT does not now have any such requirement. 47 See GATT Article XXII; Article XXIII, paragraph 1. 48 This procedure is based on GATT Article XXIII, paragraph 2. 49 John Jackson, The World Trading System: Law and Policy of International Economic Relations (Cambridge, MA: MA: Press, 1989), pp. 98-100. A small number of these cases were brought under GATTs supplementary Codes, some of which provide similar dispute resolution procedures. 50 Ibid., p. 100. Of there remaining cases, 4 specifically concerned primary products (from mines, forests, and the like), while 25 had less specific product focus (such as a complaint about a countrys import licensing scheme in general). 48 case were brought against the United States: 15 concerning manufactured goods, 17 concerning agriculture, 2 concerning primary products, and 14 without specific product focus. 51 Ibid., pp. 98-100. 52 19 U.S.C. 2411-2420. 53 19 U.S.C. 2411(b); see also 19 U.S.C. 2411(a)(l). 54 Omnibus Trade and Competitiveness Act of 1988, Public Law 100418, Sec. 1301, enacting new provision at 19 U.S.C. 24ll(d)(3)(B)(i)(III). 55 See 19 U.S.C. 2413(a). 56 19 U.S.C. 2411(c)(1),(3). 57 Office of the United States Trade Representative, Section 301 Table of Cases, computer printout dated Jan. 17, 1991. This classification involved some judgment calls, so other analyses may give different figures. 58 As a quid pro quo for GATT coverage and effective GATT dispute resolution mechanisms, the United States would consider forgoing unilateral retaliation under Section 301. 59 Some of the cases dealt only in part with manufactured goods. Those cases are included in the table, but only those aspects of the case concerning manufactured goods are reported. 60 Normally these findings were made in the context of the Section 301 case. In Case No. 68, dealing with Argentinas patent protection for pharmaceuticals, the petition was withdrawn before such findings were made, but the USTRs annual reports, National Trade Estimate Report on Foreign Barriers, identified Argentinas inadequate patent protection for pharmaceuticals as a foreign market barrier. In Case No. 70, dealing with the ECs export restrictions on copper scrap, there were no published findings, but the U.S. Government took the case to GATT to get the foreign practice stopped. 61 Up to 1988, whenever a GATT dispute resolution mechanism was applicable, Section 301 had an open-ended time limit allowing as much time as needed for the GATT process to conclude. Under that regime, Section 301 cases during 1980-1985 covered by GATT dragged on an average of almost 4 years, compared to about 1 year for other Section 301 cases. U.S. Congress, General Accounting 0ffice, International Trade: Combating Unfair Trade Practices, NSIAD-87-1OO (Gaithersburg, MD: U.S. General Accounting Cffice, Mar. 17, 1987), pp. 18-19. In 1988, Congress imposed an 18-month limit to avoid such long delays. 62 19 U.S.C. 2A14(a)(l),(2) (investigation, negotiation, and determination of sanctions), 2415(a) (application of sanctions). In certain circumstances these deadlines are shortened. For example, 3 to 6 months less time is allowed for cases involving intellectual property. 19 U.S.C. 2414(a)(3). 63 After receiving a petition, the USTR has 45 days in which to decide whether to start an investigation. 19 U.S.C. 2412(a)(2). 64 Nine out of 15 cases initiated in 1988 through 1990 were started without an industry petition. Office of the United States Trade Representative, Section 301 Table of Cases, op. cit. 65 Omnibus Trade and Competitiveness Act of 1988, Public Law 100-418, section 1301, adding a new Section 310 to the Trade Act of 1974, codified at 19 U.S.C. 2420. 66 The retaliation itself could be of some help if it limits imports of products made by the complaining U.S. industry, or downstream products.
PAGE 183
178 Competing Economies: America, Europe, and the Pacific Rim 67 U.S. retaliation under Section 301 has twice provoked a GAIT complaint by other countries, though no GATT panel has yet ruled on the issue. 68 Prestowitz, Jr., op. cit., pp. 67-68. 69 See, for example, ch. 6; Bela Balassa and Marcus Noland, Japan in the World Economy (Washington DC: Institute for International Economics, 1988), chapter 3; Advisory Committee [to the USTR] for Trade Policy and Negotiations, Analysis of the U.S.-Japan Trade Problem (February 1989), ch. 5; Prestowitz, Jr,, ibid., especially pp. 78-81,94-99,281-283. These market barriers occur in the context of Japans overall trade and industry policy, which is described in ch. 6. Other reasons for the bilateral trade deficit-including Japans better financing for long-term investment, better education better diffusion of technology to small business, and more aggressive support for commercially oriented R&D projects--are discussed in U.S. Congress, Office of Technology Assessment Making Things Better, op. cit. 70 Prestowitz, Jr., ibid., pp. 96-99. 71 Back to Business, Business Tokyo, July 1990, p. 40. 72 Sam Besachio, Director, Financial Management, Office of the U.S. Trade Representative, personal communication% Aug. 7, 1991. The USTR office has 160 full-time equivalent positions authorized, but only 151 filled. The USTR estimates that roughly 10 percent of its staff effort concerns Japan. Probably at most half of this effort concerns manufactured goods. 73 Prestowitz, Jr., op. cit., pp. 276-277. 74 Ibid., pp. 257-260. 75 Report of the Advisory Committee for Trade Policy and Negotiations, op. cit., pp. xvi-xvii. 76 U.S. Congress, General Accounting Office, U.S.-Japan Trade, op. cit. passim. 77 Ibid., pp. 20-21; Ryuji Katayama et al., Motorolas on the Line: Will Japan Pick Up? Business Tokyo, vol. 4, September 1990, pp. 39-40. 78 Joint Report of the U.S.-Japan Working Group on the Structural Impedimenta Initiative, June 28, 1990, Introduction, second paragraph. 79 Ibid., passim. 80 Ibid., Introduction. 81 First Annual Report of the U.S.-Japan Working Group on the Structural Impediments Initiative, May 22, 1991, comments of the U.S. Delegation, pp. 1-2. 82 Rudiger Dombusch, The SII Talks Are a Joke, The International Economy, vol. 4, February/March 1990, pp. 47-90. 83 U.S. Government Task Force on the EC Internal Market, through Office of the U.S. Trade Representative, EC 1992: An Assessment of Economic Policy issues Raised by the European Communitys Single Market Program (Washington DC: U.S. Government Printing Office, May 1990), pp. 31-33; Youri Devuyst, The United States and Europe 1992, World Competition; Law and Economics Review, vol. 13, No. 1, Sept. 1989, pp. 30-31. 84 19 U.S.C. 1677(5)(A), 1303(a)(l). 85 Final Affirmative Countervailing Duty Determination and Countervailing Duty Order; Carbon Steel Wire Rod From New Zealand, 51 Federal Register 7972 (Mar. 7, 1986). The payment was in the form of a tax credit; if the credit exceeded the amount of tax due, the difference was paid to the firm. 86 GATT Article XVI; GATT Subsidies Code, Article 9. (GATT Codes are explained in box 4-A.) 87 19 U.S.C. 1677(5). 88 GATT Subsidies Code, Article 11, paragraph 1. 89 Ibid, paragraph 2. Accordingly, the Codes signatories have agreed to seek to avoid causing such effects through the use of subsidies. 90 GATT Article VI, paragraph 1. 91 See GATT Antidumping Code, Article 2, paragraph. 92 GATT Article VI, paragraph 1. Under U.S. law, foreign producers are assumed to have a minimum of 10 percent administrative overhead and 8 percent profit. 19 U.S.C. 1677b(e)(l)(B). 93 Mark Koulen, Some Problems of Interpretation and Implementation of the GATT Antidumping Code, in John Jackson and Edwin Vermulst (eds.), Antidumping Law and Practice: A Comparative Study (Ann Arbor, MI; University of Michigan Press, 1989), pp. 367-369. 94 GATT Article VI 95 A price difference between the two markets that is greater than shipping costs from the low-priced market to the high-priced one is not sustainable absent market barriers. If such a difference existed and no market barriers were present, third parties would buy the good in the low-priced market and resell in the high-priced market, hinging down the price there. 96 The protection could be due to a variety of factors, including explicit government regulation informal administrative guidance given to firms, and firms own habits of doing business. 97 The law actually covers subsidies from private as well as government sources. See 19 U.S.C. 1303(a)(l). 98 For this plan to succeed, it is not necessary for the firm to drive its competitors out of business (which would constitute predatory pricing); it can be sufficient to capture substantial additional market share. One special case of this approach is life cycle pricing, or charging a constant price throughout a products whole life cycle so as to yield a normal profit over the cycle as a whole. This typically involves below-cost pricing at the start of the cycle, when the manufacturing costs are highest. 99 U.S. Congress, Office of Technology Assessment, Making Things Better, op. cit., ch. 3. 100 Thomas R. Howell et al., Steel and the State: Government Intervention and Steels Structural Crisis (Boulder, CO: Westview Press, 1988), pp. 6,4145,94-98,207-215, 254-256,301-302,332-334, 347-349, 360-361 101 If fixed costs are particularly high the incentive to dump may be particularly strong. This might be the case, for example, in Germany and Japan where employees are not as readily laid off during slack times as in the United States. 102 Jackson, op. cit., p. 244 (emphasis in original). 103 The U.S. law is found at 19 U.S.C. 1303, 1671 -1677k. 104 The final action is normally an order to impose antidumping or countervailing duties; but sometimes the additional duties are not imposed in exchange for an agreement by foreign exporters to raise prices or discontinue receiving subsidies, or an agreement by the foreign government to discontinue subsidies or limit quantities exported. 105 These figures and tables are based on semiannual filings that signatories to the GATT Subsidies Code and the GATT Antidumping Code are required to file. It is possible that other countries also instituted cases during this period. Also, the signatories are required to include only cases against other signatories, and some countries filings omit cases against non-signatories. For additional analysis of global usage of autidumping cases, see U.S. Congress, General Accounting Office, International Trade: Use of the GATT Antidumping Code, NSIAD/90238FS (Gaithersburg, MD: U.S. General Accounting Office, July 25, 1990). 106 See for example Robert Ford and Wim Suyker, Industrial Subsidies in the OECD Economies (January 1990), table 2 (covering years 1970-88), paper No. 74 in OECD Department of Economics and Statistics Working Papers; Gary C. Hufbauer and Joanna Shelton Erb,
PAGE 184
Chapter U.S. Trade Policy l 17 9 Subsidies in International Trade (Washington DC: Institute for International Economics, 1984, distributed by MIT Press, Cambridge, MA), p. 3 (covering years 1952-80). 107 U.S. Congress, General Accounting Office, International Trade: Use of the GAIT Antidumping Code, op. cit., p, 4. 108 GATT Article VI, paragraph. 109 19 U.S.C. 1671d(b)(l), 1673d(b)(l). This difference in wording can be important. Suppose that dumped or subsidized goods caused material injury to U.S. firma but would have caused substantially the same injury if they had been sold in the U.S. market without dumping. Them at least arguably, the dumped goods did cause injury but the dumping did not. 110 See 19 U.S.C. 1303(a),(b), 1671(a),(b). In some cases, U.S. law denies the injury test to some GATT signatories that have not signed the GATT Subsidies Code. This provision of U.S. law was enacted before the GATT agreement, and is allowed under a GATT grandfather clause found in GATTs Protocol of Provisional Application paragraph (l), item (b). 111 See 19 U.S.C. 1671c(b),(c). 112 See 19 U.S.C. 1673c(b),(c). 113 In early 1987 Torrington faced stiff price competition for every major account. The competition came from foreign firms that were slashing prices despite exchange rate shifts that should have made foreign goods much more expensive. To meet price competition, the firm took substantial losses. David Gridley, Industry Marketing Manager, The Torrington Company, personal communication, Aug. 6, 1991. The Commerce Department found large dumping margins. 54 Federal Register 20900-20910 (May 15, 1989). 114 Gridley, ibid. 115 In some cases, petitioners could purchase the required data. Some economic consulting firms specialize in collecting and analyzing data for subsidies and dumping cases. 116 United States International Trade Commis sion, New Steel Rails From Canada: Determination of the Commission in Investigation Nos. 701-TA-297 (Final) and 731-TA-422 (Final), USITC Pub. No. 2217, September 1989. 117 William Hoppe, General Manager, Industry Affairs, Bethlehem Steel Corp., personal communication, Aug. 29, 1991. 118 Gridley, op. cit. 119 See U.S. Congress, General Accounting Office, op. cit., pp. 13-15. 120 Public Law 100-418, Sec. 1614, amending 19 U.S.C. 1339. 121 Bureau of National Affairs, Auto Parts Industry Panel Requests Section 301 Probe of Japanese Imports, International Trade Reporter, vol. 8, No. 26, June 26, 1991, pp. 978-979. 122 The statutory time limits for the Commerce Departments preliminary determina tion are 160 days after the filing of the petition for dumping cases and 85 days for subsidy cases. 19 U.S.C. 1673b(b)(l)(A) (dumping), 1671b(b)(l)(A) (subsidies). The Commerce Department usually takes all or virtually all of the time allowed. 123 19 U.S.C. 1671 b(d),(e), 1673b(d),(e). 124 The statute provides another 75 days for the final determination and another 7 days to publish an order reflecting the new subsidy value or dumping margin. 19 U.S.C. 1671d(a), 1671e(a) (subsidies), 1673d(a), 1673e(a) (dumping). While the difference between the preliminary duty and the final duty is refunded on importations during the interim if the new duty is lower, the difference is not collected if the new duty is higher. 19 U.S.C. 1671f(a) (subsidies), 1673f(a) (dumping). 125 In the 1980s, petitioners often had to wait several years for annual reviews. 126 45 Federal Register 30,618 (May 9, 1980). 127 52 Federal Register 1504 (Jan. 14, 1987). This annual review, which covered imports during the period May 1981 through April 1982, was several years late. Such delays in annual reviews were common in the 1980s, but recently the Commerce Department has worked to reduce the backlogs. 128 Smith Corona Corp. v. United States, 915 F.2d 683 (Fed. Cir. 1990). The courts opinion summariz es the history of the case. 129 The customs entries for the intervening imports had been liquidated, meanin g that it was no longer possible to collect antidumping duties retroactively. The Court of Appeals for the Federal Circuit pointed out that in February 1989, the Court of International Trade, a lower reviewing court, should have ordered liquidation suspended; had it done so, antidumping duties could have been collected retroactively to that date. 130 Statement of Senator Carl Levin before the Senate Committee on Governmental Affairs, Subcommittee on Oversight of Government Management, on Oversight and Enforcement of Antidumping and Countervailing Duties, June 13, 1991. 131 See "Request for Antidumping Monitoring Program Covering Certain Ball Bearings and Parts Thereof, submitted to the U.S. Commerce Department International Trade Adminis tration,on Apr. 20, 1990, by Stewart and Stewart, Washington, DC, on behalf of the Torrington Co. In 1989, Torrington had obtained antidumping orders on ball bearings from nine countries. Torrington requested that the Commerce Department monitor imports from 14 countries not included in the 1989 orders to determine if those imports were also dumped. when the Commerce Department denied this request, Torrington filed antidumping petitions for the 14 countries. Torrington never got the chance to prove dumping, because the ITC found at a preliminary stage that there was no reasonable indication that the imports complained of had caused or threatened material injury to the U.S. industry. U.S. International Trade Commis sion, Ball Bearings, Mounted or Unmounted, and Parts Thereof from Argentina, Austria, Brazil, Canada Hong Kong, Hungary, Mexico, The Peoples Republic of China, Poland, The Republic of Korea, Spain, Taiwan, Turkey and Yugoslavia, USITC Pub. 2374, April 1991. (One Commis sioner dissented.) Torrington has appealed this ruling, and maintains that the U.S. industry is worse off than in 1989 when the ITC did find injury. Gridley, op. cit. The 1989 antidumping orders are found at 54 Federal Register 20900-20910 (May 15, 1989). 132 Howell et al., op. cit., pp. 526-534. 133 Presidents Steel Decision, pp. 1-2, attached to Office of the United States Trade Representative, Brock Announces Presidents Steel Decision, Sept. 18, 1984 (press release). As requested, Congress then empowered the President to negotiate bilateral agreements to limit other countries total steel exports to the United States. 134 A previous order concerning only the imported finished product would not apply if the parts are imported instead (although sometimes cases initially cover both a product and its components); a previous order concerning only imported parts would not apply if the finished product is imported instead; and an order applying to a product assembled in one foreign country would not apply to the end product assembled in another foreign country. 135 These provisions are found in Public Law 100-418, Sec. 1321, codified at 19 U.S.C. 1677j. As of June 1991, the Commerce Department had twice ruled on whether these provisions applied and both times decided that they did not. 56 Federal Register 9667 (Mar. 7, 1991), 55 Federal Register 52066 (Dec. 19, 1990) (color picture tubes from Canada, Japan, Korea, and Singapore); 55 Federal Register 6028 (Feb. 21, 1990), 54 Federal Register 50,260 (Dec. 5, 1989) (fork-lift trucks from Japan). The anticircumvention provisions apply only if the difference in value between the components and the end product is small. In both cases, the Commerce Department found this require ment not satisfied; in the fork lifts case, the value added in the final assembly, including profit, was appro ximately 25 to 40 percent. Another limitation is that the provisions do not appear to apply when parts from one foreign country are initially assembled in another foreign country,
PAGE 185
180 Competing Economies: America, Europe, and the Pacific Rim as often happens today in the Pacific Rim. The anticircumvention provisions also address cases, such as Smith Coronas electric typewriters case, in which merchandise subject to an order is slightly altered or is replaced with a later-developed product. In such cases, the previsions might be more effective. 136 A March 1990 GATT panel decision found the ECs anticircumvention provisions to violate GATT. EEC Regulation on Imports of Parts and Components: Report by the panel, GATT Dec. L/6657, Mar. 22, 1990. While not directly relevant to the U.S. provisions, that decision does indicate that GATT will seriously scrutinize anticircumvention provisions if the matter arises. 137 19 U.S.C. 1671(a)(2), 1673(2), 1677(7). 138 Gridley, op. cit. 139 Final Determina tion of Sales at Les S Than Fair Value: Digital Readout Systems and Subassemblies Thereof From Japan, 53 Federal Register 47,844,47,847 (Nov. 22, 1988). 140 United States International Trade Commission, Digital Readout Systems and Subassemblies Thereof From Japan: Determination of the Commission in Investigation No. 731-TA-390 (Final), Jan. 1989 (USITC Publication 2150), pp. 16-22. The Commis sioners noted various indicators of health for the two industries. Sales in both industries had declined from 1985 to 1986, but rose in 1986, 1987, and the first three quarters of 1988. During this period, pre-tax profits as a percentage of net sales in the console industry rose steadily from 2.6 percent to 9.2 percent; in the transducer industry, pre-tax profits were 14.0 percent in 1985, fell to 10.3 percent in 1986, rose to 13.5 percent in 1987, and rose an amount not publicly disclosed in the first 3 quarters of 1988 as compared with the first 3 quarters of 1987. Other indicators of health considered by the Commissioners included production volume, capacity utilization, inventory levels, number of employees and hours worked, and investment in R&D. 141 Ibid., pp. 42-53, 125-126. 142 Ibid., pp. 42-53, 126-136. 143 By statute, the injury requirement is satisfied if the ITC finds that the U.S. industry is threatened with material injury, even if the U.S. industry is not yet materially injured. However, actual injury must be imminent, and a finding of a threat of material injury may not be made on the basis of mere conjecture or supposition. 19 U.S.C. 1677(7)(F)(2). Thus, projecting current trends to a conclusion of future long-term injury to the U.S. industry will not normally justify finding a threat of material injury. 144 Ibid., p. 32. 145 U.S. International Trade Commission Pressure-Sensitive PVC Battery Covers from West Germany, Investigation No. 731-TA-452 (Preliminary), USITC Pub. No. 2265 (March 1990). 146 Ibid., pp. A-2, A-6, A-7. 147 Ibid., pp. 17-18. 148 Ibid., pp. A-38, A-39; Post-Conference Brief of the Petitioner, National Label Co., in the matter of Pressure-Sensitive PVC Battery Covers from West Germany, Investigation No. 731-452 (Preliminary), Feb. 13, 1990, pp. 17-21. 149 U.S. International Trade Commission op. cit., pp. A-39, A-40. 150 U.S. International Trade Commis sion op. cit., p. 20. Evereadys statements were confidential and not revealed by the ITC. The ITC found that Zweck form did not bid on any Eveready purchases at the meeting, and the discussions between the two parties (had not] reached the point where they were in sufficient detail and scope to provide an indication of price suppression. National Label argued that, even if Zweckform had not solicited Evereadys business, Zweckforms dumped sales to other firms in the U.S. market made National Label reasonably fear that Zweckform would seek Evereadys U.S. business. Post-Conference Brief of Petitioner, National Label Co., op. cit., pp. 38-41. Therefore, according to National Label, Zweckforms dumping had in fact caused National Label to suffer injury by lowering its prices to Eveready. The Commission did not address this argument. 151 Post-Conference Brief of the Petitioner, National Label Co., op. cit., pp. 63-66; U.S. International Trade Commission op. cit., p. A-46. 152 U.S. International Trade Commission, op. cit., pp. 20-21. 153 The Commissions general practice is to permit outside attorneys and consultants to see other parties confidential business information, but not to permit the parties own personnel to see such information, for fear the information would be used improperly to gain business advantage. 154 James Shacklett, president, National Label Co., personal communication, Oct. 11, 1990. 155 The text of the agreement is contained in an exchange of letters on Sept. 2, 1986, between Japans Ambassador Matsunaga and USTR Clayton Yeutter. An account of the U.S. negotiating strategy is given in Prestowitz, Jr., op. cit., p. 62. When a foreign country sends dumped or subsidized merchandise to a third country, the United States could ask the third country to conduct its own countervailing duty or antidumping investigation and under GATT Article VI paragraph 6(b) relief could be granted if the dumping or subsidies injured the U.S. industry. Under U.S. law, firms may petition the USTR to make such a request to foreign governments. 19 U.S.C. 1677k (added in 1988, Public Law 100-418, Sec. 1317). However, this has never been tried, and there is no guarantee that the third country would honor such a request. 156 See ch. 6. Cray Research, the principal U.S. supercomputer manufacturer, lacks the size and product diversity to similarly crosssubsidize. 157 Marie Anchordoguy, Computers Inc: Japans Challenge to IBM(Cambridge, MA: Harvard University Press, 1989), pp. 72-75. 158 Ibid., pp. 68,69 (table 3, total of gross sales minus trade-ins), 228-230 (app. C). 159 Ibid., pp. 141-147. 160 See ch. 6. 161 These firms are Fujitsu, Hitachi, Mitsubishi, Toshiba, and NEC. Anchordoguy, op. cit., p. 143. The 1989 computer, electronic devices, and consumer electronics revenues of Fujitsu, Hitachi, and NEC alone were $13.7 billion, $14.6 billion, and $15.0 billion for the respective fins, totaling $43.3 billion. See tables 6-1 and 6-3. 162 Subsidies law was designed on the premise that subsidies distort trade but do not change the underlying comparative advantage, so that discontinued subsidies have no more effect. This assumption has become increasingly mistaken. In the modem world, comparative advantage is often purchased by governments; for example, government money bought the United Kingdom a modem steel industry that could sell steel competitively in the world. Howell et al., op. cit., pp. 171-173. 163 GATT Article VI, paragraph 3. 164 Section 19 U.S.C. 1677b(e). Other countries have argued that these floors are not proper and disadvantage fins, for example, in countries such as Japan where profit margins are often lower than 8 percent. 165 In July 1990, a panel of the GATT Committee on Antidumping Practices ruled that the Antidumping Code requires the investigating government to verify that an antidumping petition is supported by a majority of the domestic industry, and that the U.S. Commerce Department had not done so in an antidumping case involving stainless steel pipes and tubes from Sweden. As countries have often done with adverse GATT rulings during the Uruguay Round, the United States has said that it will bring its behavior in line with GATT norms, but only after the Uruguay Round is completed, when it is clear what the new norms will be. 166 These and other problems are discusscd in N. David Palmeter, The Capture of Antidumping Law, Yale Journal of International Law, vol. 14, No. 1, pp. 182-198 (reviewing Jagdish Bhagwati,
PAGE 186
Chapter 4--U.S. Trade Policy l 181 Protectionism (Cambridge, MA: The MIT Press, 1988)). 167 In light of this, the EC accuses the United States of a double standard in complaining about government subsidies to the members of Airbus Industrie, a European consortium challenging U.S. dominance. 168 Omnibus Trade and Competitiveness Act of 1988, Public Law 100418, section 1301, enacting new provisions at 19 U.S.C.2411(d)(3)(B)(ii), 2411(d)(3)(E). This use of Section 301 is not completely new. While Section 301 has most often addressed foreign countries barriers to sales of U.S. goods in their markets, is has sometimes included foreign subsidies as they affect sales of U.S. goods in third country markets; those subsidies could fit the definition of export targeting. 169 Normally, GATT prohibits protection with the exception that tariffs maybe imposed as agreed upon in negotiating rounds. Aside from the escape clause, GATT permits additional protection in certain other contexts, such as national defense, agriculture in some cases (this exception was made at the United States insistence), to meet a balance of payments crisis, and in response to subsidies and dumping. Protection is also allowed when compensation is negotiated under Article XXVIII (ch. 2). 170 GATT Article XIX. 171 Ibid. 172 19 U.S.C. 2251-2254. 173 19 U.S.C. 2252(b)(l)(B). 174 See 19 U.S.C. 2252(b)(l)(A). 175 19 U.S.C. 2252(c). 176 A petition may be filed by any entity, including a trade association firm, certified or recognized union, or group of workers, which is representative of an industry. 19 U.S.C. 2252(a)(l). The great majority of investigations are triggered in this manner. An investigation can also be triggered by the President, the USTR, the House Ways and Means Committee, the Senate Finance Committee, or the ITC itself. 19 U.S.C. 2252(b)(l)(A). 177 19 U.S.C. 2252(b)(2). 178 19 U.S.C. 2252(e). The ITC can recommend protection adjustment assistance, or both. Adjustment assistance consists of: 1) payments, retraining and/or job placement services for workers, and/or 2) financial help and/or technical assistance to firms. Even if adjustment assistance is recommended by the ITC and the recommendation is approved by the President, the petitioners must still apply for it to the Departments of Commerce and/or Labor. Generally, Section 201 petitioners desire import relief rather than adjustment assistance, for which they could apply independently without going through a Section 201 proceeding. Recommendations for relief are made only by those Commissioners who voted to find injury. If the vote on injury was tied, those Commissioners finding injury recommend relief; the President then has the choice of treating the Commissions vote as a vote to find no injury or a vote to find injury. 19 U.S.C. 1330(d)(l). 179 19 U.S.C. 2252(e). 180 19 U.S.C. 2252(f). 181 19 U.S.C. 2253(a)(4). 182 19 U.S.C. 2253(a). 183 19 U.S.C. 2253(a)(3). Congress can in principle override the Presidents decision and reinstate the relief recommended by the ITC, but never has done so. See 19 U.S.C. 2253(c). 184 19 U.S.C. 2253(e). 185 19 U.S.C. 2254(b). 186 U.S. General Accounting Office, International Trade: Activity Under Section 201 of the Trade Act of 1974, NSIAD-87-96FS (Gaithersburg, MD: U.S. General Accounting Office, Mar. 30, 1987), pp. 3, 6-9. This report lists all Section 201 cases brought during these years, except for cases TA-201-61 and IA-201-62, which both ended with a finding of no injury as shown in the ITC Annual Reports for 1988 and 1990. Some form of protection was granted by the President in eases No. (201 -TA-) 5, 16, 18, 19,29,35,36,37,39,43, 47,48,51, and 56. In cases 29, 37, 39,47, and 56 the protection was by its terms less than that recommended (for example, lower tariffs, or shorter duration of the same tariffs). While the President sometimes granted a different type of relief that is not directly comparable with the protection recommended the President never granted protection that was by its terms greater than the relief recommended. 187 Section 201 has a higher standard of injury, and in that regard Section 201 relief is harder to get. However, before 1980, antidumping and countervailing duty cases had no time limits and no court review, and the Treasury Department, which handled those cases then, often put cases on indefinite hold. Section 201, in contrast, promised a decision within well under a year, and the required public hearing could help muster political support for relief. 188 The ITC found that the auto industry, while substantially injured by increasing imports, was injured more by a general economic recession in the country, and was thus ineligible for relief. U.S. International Trade Commission, Certain Motor Vehicles and Certain Chassis and Bodies Therefor: Report to the President on Investigation TA-201-44 under Section 201 of the Trade Act of 1974, USITC Publication 1110, December 1980, pp. 21, 67, 134, 173 (3-2 vote). In 1988, Congress specified that the ITC may not aggregate the causes of declining demand associated with a recession or economic downturn in the United States economy into a single cause of serious injury or threat of injury. 19 U.S.C. 2252(c)(2)(A). 189 19 U.S.C. 2251. 190 In theory, the government could encourage a new industry to form by other means, such as R&D support, low-interest loans, and the promise of protection in the future if needed. When the new industry started to sell products, it would then exist and could be protected under Section 201. However, to use Section 201 in this way could violate GATT. GATTs escape clause requires that the injury be due to unforeseen increased imports. If the dominan t foreign suppliers were already established when the U.S. industry started to form, it could be bard to argue that any increased imports were unforeseen. While U.S. law does not require that the increase in imports be unforeseen, an application of Section 201 to imports that were anticipated could provoke a GATT challenge, (However, the GATT provision requiring an unforeseen increase has fallen into disuse and might be eliminated in the Uruguay Round.) 191 The statute speaks in absolute, rather than relative, terms, directing Commissioners to consider, among other economic factors, the significant idling of productive facilities in the domestic industry, the inability of a significant number of firms to carry out domestic production operations at a reasonable level of profit, and significant unemployment or underemployment within the domestic industry. 19 U.S.C. 2252(c). See, for example, U.S. International Trade Commission Nonrubber Footwear: Report to the President on Investigation No. TA-201-50 Under Section 201 of the Trade Act of 1974, USITC Publication No. 1545 (July 1984), pp. 11-19 (views of Chairwoman Stern, Vice Chairman Liebeler, and Commissioner Rohr). 192 U.S. International Trade Commission, ibid., p. 19 (views of Chairwoman Stem, Vice Chairman Liebeler, and Commissioner Rohr) (quoting legislative history). 193 See U.S. International Trade Commission Television Receivers, Color and Monochrome, Assembled or Not Assembled, Finished or Not Finished, and Subassemblies Thereof, Report to the President on Investigation TA-201-19, USITC Pub. 808, March 1977, pp. 13-16; U.S. International Trade Commis sion, Citizens Band (CB) Receivers, Report to the President on Investigation TA-201-29, USITC Pub. 862, February 1978, pp. 7, 28. These cases are also discussed in Morris Morkre and David Tarr, The Effects of Restrictions on United States Imports: Five Case Studies and Theory, Staff Report of the Bureau of Economics to the Federal Trade Commission June 1980, chs. 4, 5.
PAGE 187
182 l Competing Economies: America, Europe, and the Pacific Rim 194 U.S. International Trade Commission, Nonrubber Footwear, No. TA-201-50, op. cit., pp. 9-10,23. 195 Ibid., p. 22. 196 Ibid., pp. 9-10. 197 U.S. International Trade Commission, Nonrubber Footwear, No. 221-201-55, op. cit., pp. 11-12 (views of chairwoman stern). 198 As used in this discussion of what the U.S. public sector purchases, defense purchases mean purchases classified as defense in Congressional budget authorizations. To a first approximation this includes all of the Defense Department (except the civilian side of the Army Corps of Engineers), plus defense programs of the Department of Energy and NASA. 199 Department of Commerce, Bureau of Economic Analysis, Government Division Table PMB-llB, National Defense Purchases by Subcategory in Constant Dollars, Computer generated on Feb. 28, 1991. 200 Computer and Business Equipment Manufacturers Association. 201 Department of Commerce, Table PMB-llB, op. cit. 202 U.S. Department of Commerce, International Trade Administration, 1991 U.S. Industrial Outlook, S/NO03-009-00586-8 (Washington, DC: U.S. Governm ent Printing Office, January, 1991), pp. 37-2. 203 Eric Nelson, Director, Market Research, North American Telecommunications Association, personal Communication Apr. 1, 1990. 204 Public purchasing reportedly accounts for 90 percent of the output of telecommunications equipment, measuring equipment, and electro-medical equipment. Commis sion of the European Communities (CEC), Communication from the Commission on a Community Regime for Procurement in the Excluded Sectors: Water, Energy, Transport and Telecommunications, COM (88) 376 (Brussels, Belgium: CEC, Oct. 11, 1988), table 4, telefaxed from Anna Snow, Delegation of the CEC, Washington, DC, May 29, 1991. 205 U.S. Department of Commerce, Bureau of Economic Analysis, Government Transactions Methodology Papers: U.S. National Income and Product Accounts, BEA-MP-5 (Washington DC: U.S. Government Printing Office, November 1988), table III-6. 206 Robert LaucL, Library of Congress, Congressional Research Service, American Law Division, The Buy American Act: Legislative History and Differential Favoring Domestic Products, Oct. 25, 1975, p. 2. For a recent description of Buy American legislation, see Robert Lauck, Library of Congress, Congressional Research Service, American Law Division, Buy American Legislation, Mar. 28, 1990. 207 Other laws include the Davis-Bacon Act, the Walsh-Healy Act, and the Miller Act. See Commis sion on Government procurement, Report of the Commission on Government Procurement, Stock No. 5255-00002 (Washington, DC: U.S Government Printing Office, December 1972), vol. 1, app. G, p. 167. 208 Inclusive of duty. 209 48 C.F.R. 25.105. 210 48 C.F.R. 25.101 (definition of domestic end product). Costs other than components, such as direct labor and overhead for final assembly, are not counted at all for this test. Also, the test considers only major components- those articles, materials, and supplies incorporated directly into the final product. Ibid. (definition of components) (emphasis added). Thus, the origin of subcomponents, which are first incorporated into major components, is not considered. 211 48 C.F.R. 25.104,.105. 212 48 C.F.R. 225.105. 213 U.S. General Services Admini stration, Federal Procurement Data Center, Federal Procurement Report, Fiscal Year 1990 through Fourth Quarter (Washington, DC: U.S. General Services Adminis tration), pp. 74-75. Additional detail on some of these preferences is given in U.S. Department of Defense, A Report to the United States Congress by the Secretary of Defense, The Impact of Buy American Restrictions Affecting Defense Procurement (Department of Defense, July 1989), app. E, p. 13; Section 8(a) of the Small Business Act of 1953, as amended, 15 U.S.C. 637(a); and 48 CFR 13.105, 13.106 (Oct 1, 1990). 214 Alfred Volkman, Director, Foreign Contracting, 0ffice of Deputy Assistant Secretary of Defense (Procurement), testimony at hearings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27,1989, pp. 64, 72; see also ibid., p. 266. In this testimony, DoD said merely that less than half of defense procurement is opened under the MOUs. However, a DoD official clarified that the estimate was close to 50 percent. Personal communication July 19, 1991. 215 Joseph Kelly, Director, Security and International Relations Issues, U.S. General Accounting Office, testimony at hearings before the House Armed Services Committee, Subcommittee on Investigations, Apr. 30, 1991, p. 4. 216 For example, the United States does not waive restrictions on defense purchases of foreign machine tools. U.S. Congress, General Accounting Office, Defense Procurement: DoD Purchases of Foreign Made Machine T OO l S GAO/NSIAD-91-70 (Gaithersburg, MD: U.S. General Accounting Office, February 1991), p. 22. Restrictions on defense purchases of supercomputers also are not waived. 217 The GATT Procurement Code took effect Jan. 1, 1981. It is described in U.S. Department of Commerce, International Trade Administration, Office of Policy, Government Procurement (July 1981) (prepared by Nancy E. Morgan) (vol. 2 of a series entitled The Tokyo Round Trade Agreements), which also gives the annexes specifying the Codes scope of application to particular countries. 218 Public Law 96-39. 219 The test here is the nature of the good rather than which agency buys it. 220 An SDR is a currency unit consisting of a basket of currencies from nations that are members of the SDR Department of the International Monetary Fund (IMF). It is also the unit of account for the IMF. In January 1991, one SDR was worth about $1.40. Periodically the USTR publishes an updated dollar equivalent for 130,000 SDRs, to be used in assessing the Acts applicability to purchases by the U.S. Government. 221 Public Law 96-39. The Presidential authority to waive Buy American and other preferences comes from Section 301 of the Act, codified at 19 U.S.C. 2511. 222 These provisions are found in 19 U.S.C. 2511-2512 and 48 CFR 25.40.406. Generally, for purchases covered by the Trade Agreements Act, bids from non-covered countries will not be considered at all. 19 U.S.C. 2512; 48 CFR 25.402(c). Also, under the Trade Agreements Act a goods origin is determined by where it underwent its last substantial transformation instead of by the fifty-percent-ofcomponents rule under Buy American. 48 CFR 25.401 (definition of designated country end product). 223 Robin Zee, Purchasing Preference Practices: A 50 State Overview, Document No. C-130 (Lexington, KY: National Association of State Purchasing Officials (NASPO) and The Council of State Governments, 1989), pp. 22-24. 224 Bethlehem Steel Cop. v. Board of Commissioners of the Dept. of Water and Power of the City of Los Angeles, 276 Cal. App. 2d 211, 224,80 Cal. Rptr. 800,802 (1%9). 225 Zee, op. cit. 226 Ibid., pp. 8-10. 227 The preference of 12 percent for goods produced in labor surplus areas raises prices paid by the government, but may save the country money by helping to reduce the costs of unemployment. Preferences for small and minority business raise prices paid by the
PAGE 188
Chapter 4--U.S. Trade Policy l 183 government but they serve definite social goals. 228 For example, bidders and agencies sometimes mistakenly count the cost of domestic labor for final assembly, plus associated overhead and general administrative expense, as part of the cost of domestic components. U.S. Congress, General Accounting Office, Report to the Secretary of the Interior, Bureau of Reclamation: Misapplication of the Buy American Act, GAO/NSIAD-90-32 (Washington DC: U.S. General Accounting Office, November 1989); U.S. Congress, General Accounting Office, Defense Procurement. DOD Purchases of Foreign Made Machine Tools, GAO/NSIAD-91-70 (Gaithersburg, MD: U.S. General Accounting Office, February 1991), pp. 25-27. 229 See, for example, U.S. Department of Defense, Office of the Secretary of Defense, The Impact of Buy American Restrictions Affecting Defense Procurement (U.S. Department of Defense, July 1989), p. 35. 230 Public Law 100-202 (FY-1988 Department of Defense Appropriations Act), Sec. 8112. This prohibition could be waived if the Secretary of Defense certified to Congress that a foreign acquisition is required for national security. This provision is still in effect. Public Law 101-511, Sec. 8034. 231 U.S. Congress, House Committee on Gov ernment Operations, International Procurement and Waivers of the Buy American Act: U.S. Business at a Disadvantage, House Report No. 101-989, Nov. 29, 1990, p. 22. 232 Preferences for small business, minority-owned business, and businesses in labor surplus areas have not drawn as much criticism, probably because such programs and the underlying social goals are shared by some other nations. 233 U.S. Congress, Committee on Government Operations, op. cit., p. 7. 234 At times foreign firms were treated even better than U.S. firms when bidding for defense contracts, because foreign firms costs were subject to less stringent audits. Ibid., pp. 13-17. 235 Allan I. Mendelowitz, Director, Trade, Energy and Financial Issues, United States General Accounting Office, testimony at hearings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27, 1989, p. 30; and answers by the Defense Departments Office of Foreign Contracting to questions from the Subcommittee, same hearings, pp. 266-267. The Defense Department cautioned against comparisons of such data because of differing national org anization and procurement systems. Defense spending is also higher as a proportion of GDP in the United States than for U.S. allies. For example, despite roughly equal GDP, the market for defense procurement in the EC is only 40 percent of that in the United States. U.S. Congress, Office of Technology Assessment, Arming our Allies: Cooperation and Competition in Defense Technology, OTA-ISC-449 (Washington DC: U.S. Government Printing office, May 1990), pp. 47-50. 236 U.S. Congress, General Accounting Office, European Initiatives: Implications for U.S. Defense Trade and Cooperation, NSIAD-91167 (Gaithersburg, MD: U.S. General Accounting Office, Apr. 4, 1991), pp. 36, 39-42. 237 The 19 are Australia (nonreciprocal), Belgium, Canada (nonreciprocal), Denmark, Egypt Germany, France, Greece, Israel, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain Sweden, Switzerland, Turkey, United Kingdom. 238 U.S. Congress, General Accounting Office, European Initiatives, op. cit., p. 37. 239 Matthew B. Coffey, President, National Tooling and Machining Association, testimony at h earings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27, 1989, pp. 3840. 240 U.S. Congress, House Committee on Gov ernment Operations, op. cit., p. 21, citing U.S. Congress, General Accounting Office, Industrial Base: Adequacy of Information on the U.S. Defense Industrial Base, NSIAD-91-167 (Gaithersburg, MD: U.S. General Accounting Office, November 1989). The Defense Department estimated that the ratio of defense exports to MOU partners to imports from those partners dropped from almost 8:1 in the late 1970s and early 1980s to about 2:1 in 1987; in 1987, exports were estimated at 8 billion, imports at 4 billion. These estimates are based on unverified procurement data collected by DoD and foreign defense ministries. For that reason, and because of specific problems with the data which GAO has identified, GAO considers these estimates unreliable. Testimony at hearings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27, 1989, p. 266 (written answers to questions included in the hearin g record); U.S. Congress, General Accounting Office, European Initiatives, op. cit., pp. 39-43, 61-62. 241 U.S. Congress, General Accounting Office, European Initiatives, op. cit., pp. 43-44. 242 House Committee on Government Operations, op. cit., pp. 11-13, 17-21. 243 U.S. Congress, General Accounting Office, European Initiatives, op. cit., p. 44. 244 The House Committee on Government Operations reported that the Admini stration had classified U.S. data derived from publicly available information and had classified data from other Code signatories after they declared the data no longer restricted. U.S. Congress, House Committee on Government Operations, op. cit., pp. 20-21. 245 U.S. Congress, U.S. General Accounting Office, The International Agreement on Government Procurement: An Assessment of its Commercial Value and U.S. Government Implementation, NSIAD-84117 (1984), pp. 11-17. As used in this chapter, Code-covered procurement means procurement that under the Code must be both transparent and nondis criminatory. 246 Allan I. Mendelowitz, Director, Trade, Energy and Financial Issues, United States General Accounting Office, testimony at hearings before the House Subcommittee on Legislation and National Security, Sept. 27, 1989, pp. 13-14. (This testimony gave a brief, qualitative unclassified summary of the classified data.) 247 In 1985, the total central gov ernment procurement expenditures (including goods and services) of all other Code signatories combined was at least roughly equal to that of the United States. Ibid., p. 24. 248 W. Douglas Newkirk Assistant U.S. Trade Representative for GATT Affairs, testimony at hearings before the House Committee on Government Operations, Subcommittee on Legislation and National Security, Sept. 27, 1989, p. 50; U.S. Congress, General Accounting Office, The International Agreement on Government Procurement, op. cit., p. 16. Probably almost half of defense procurement is now covered under MOUs. By 1981, 11 out of 19 countries had already signed MOUs. 249 However, the Code might have made it easier in the long term for foreign suppliers to form and maintain such relationships with the U.S. Government, thus increasing U.S. imports over the years. 250 U.S. Congress, U.S. General Accounting Office, The lnternational Agreement on Government Procurement, op. cit., pp. 15-17. 251 Ibid, pp. 13-17. 252 For the 25 percent estimate, see U.S.International Trade Commis sion, The Effects of Greater Economic Integration Within the European Community on the United States, USITC Publication 2204 (Washington, DC: USITC, July 1989), p. 4-14 [one page]. Some sources cite estimates as high as 50 percent, such as Aviva Freudmann, EC Takes Giant Step to Open Public Procurement Markets, Atlantic Trade Report, vol. 2, No. 5, Mar. 1, 1990, p. 1. 253 Statement of Bernard H. Falk, president National Electrical Manufacturers Association, on the Annual Report on Foreign Discrimination Against Procurement of U.S. Goods and Services, before the
PAGE 189
184 l Competing Economies: America, Europe, and the Pacific Rim House Committee on Government Operations, Subcommittee on Legislation and National Security, U.S. House of Representatives, May 1, 1990, p. 2. 254 Steve Thomas and Francis McGowan, The World Market for Heavy Electrical Equipment (Surrey, UK: Nuclear Engineering International Special Publications, 1990). 255 In 1987, digital switching systems in the EC reportedly cost between $225 and $500 per line, compared to an estimated price of around $150 that would have resulted from competitive procurement. (The price in the United States was around $100.) Paolo Cecchini with Michel Catinat and Alexis Jacquemin, The European Challenge, 1992: The Benefits of the Single Market (Aldershot, UK: Gower Publishing Co., 1988), p. 22. 256 U.S. Congress, General Accounting 0ffice, International Restraints to Competitiveness of the U.S. Heavy Electrical Equipment Industry, NSIAD-83-51 (Gaithersburg, MD: U.S. General Accounting Office, Sept. 14, 1983), p. 14. These agencies were the Army Corps of Engineers; Tennessee Valley Authority; Western Area, Southwestern, and Bonneville Power Admini strations; and Bureau of Reclamation. 257 Negotiations to amend GATTs Pr ocurement Code have been going on in parallel with the Uruguay Round negotiations to amend the basic GATT agreement. 258 See U.S. International Trade Commission The Effects of Greater Economic Integration Within the European Community on the United States: Third Followup Report, USITC Publication 2368 (Washington, DC: USITC, March 1991), p. 6-4; U.S. International Trade Commission, The Effects of Greater Economic Integration Within the European Community on the United States, USITC Publication 2204 (Washington, DC: USITC, July 1989), p. 4-15. 259 Beverly Vaughan, Director of International Government Procurement Policy, Office of the United States Trade Representative, personal communication Mar., 25, 1991. 260 Public Law 100-418, Title VII. 261 Public Law 100-418, Sec. 7003, codified at 19 U.S.C. 2515 (see especially paragraph d(2)). 262 19 U.S.C. 2515(d),(f). For practices covered by the Procurement Code, the Codes dispute resolution procedures are to be tried for a year before the United States retaliates. 263 Office of the United States Trade Representative, Information on Certain Foreign Government Procurement Markets (Washington DC: Office of the USTR, Apr. 27, 1990). 264 Office of the United States Trade Representative, Title VII Report (Washington, DC: Office of the U.S. Trade Representative, Apr. 26, 1991). 265 Congress passed Title VII in the belief that a strategy of toughness, not patience, was more likely to open the foreign markets. This fundamental dispute--Congress advocating toughness on international trade issues while the Adminis tration prefers patience-has long been a theme of U.S. trade policy. 266 The agencies represented are the Departments of Commerce, Agriculture, State, Treasury, Defense, Labor, Transportation, and Energy; the Office of Management and Budget; the Office of the U.S. Trade Representative; the Council of Economic Advisers; the Environmental Protection Agency, the Small Business Administration; the Agency for International Development the Export-Import Bank of the United States; the Overseas Private Investment Corporation; the U.S. Trade and Development Program, and the U.S. Information Agency. U.S. Department of Commerce, Export Programs: A Business Directory of U.S. Government Resources, May 1991. 267 U.S. Commerce Department International Trade Admim stration, Office of Financial Management, personal communication, Aug. 2, 1991. 268 Negotiations with foreign governments are led by the Office of the U.S. Trade Representative (USTR), which is within the Executive Office of the President. 269 Export Promotion Activities of Major Competitor Nations, May 1988. While the report was not released by the Commerce Department, it was provided under the Freedom of Information Act to the authors of Government Export Promotional Programs in Nine Countries: The Cornell University/XPORT Report on Government Export Promotion Programs and Policies in Canada, the United Kingdom, France, West Germany, Japan, Taiwan, Singapore, South Korea, and Brazil (Ithaca, NY: Cornell University, Johnson Graduate School of Management 1989); see page 7 of that report. The figures in the text and table 4-10 are drawn from table A on page 5 of that report, and app. I on page 54. 270 U.S. Congress, General Accounting Office, Export Promotion: Problems in Commerces Programs, NSIAD-89-44 (Gaithersburg, MD: U.S. General Accounting Office, Jan. 26, 1989), pp. 16-17. 271 Ibid., p. 18. Similarly, Trade Development usually lacked funds to print attractive, professional-looking brochures to advertise trade shows. Ibid., p. 41. 272 U.S. Commerce Department, International Trade Adminis tration, office of Financial Management, personal communication July 30 and Aug. 2, 1991. The combined budget for USFCS, TD and IEP increased from $146 million in fiscal year 1987 to $170 million in fiscal year 1991. The fiscal year 1987 figures are funds actually spent. 273 The program depended on ITA providing free publications. This stopped in 1981, causing program participation to wane. These two examples are taken from Ronnie L. Goldberg, BSI Consulting, Inc., Federal programs for the Promotion of Manufactured Exports, contractor report prepared for the Office of Technology Assessment, June 1987, ch. 2, pp. 6-7. 274 U.S. Congress, General Accounting Office, Export Promotion, op. cit., passim. 275 Information about the USFCS strategic review is taken from U.S. Department of Commerce, U.S. Foreign and Commercial Service, Strategic and Technical Reviews Working Paper Export Assistance Needs: Working Group Recommendations SR 90-7, Mar. 1990; and interviews with USFCS staff. 276 See Organisation for Economic Co-operation and Develop ment, The Export Credit Financing Systems in OECD Member Countries (Paris, France: OECD, 1990, 4th cd.). 277 Ibid., ch. 23. 278 Eximbank 1990 Annual Report, p. 27; Organisation for Economic Cooperation and Development, op. cit., p. 229. Eximbank was capitalized in 1945 with $1 billion from the U.S. Treasury. James Jackson, Library of Congress, Congressional Research Service, Economics Division, Export-Import Bank Financial Issues, IB88013, July 27, 1989, p. 2. 279 The Arrangement on Guidelines for Officially Supported Export Credits requires countries to announce contemplated export loans, to permit other countries to match the terms. The Arrangement is signed by 23 countries, including the major economic powers. The text of the Arrangement is found in Organisation for Economic Cooperation and Development op. cit., ch. 25. Notification is treated in sections 14-17. 280 Chief Accountant, Commodity Credit Corporation, U.S. Department of Agriculture, personal communication, July 22, 1991. 281 This figure of $40 billion does not include some products eligible for financing assistance, such as wood products, seafood, and leather. 282 Public Law 91-181, amending Section 9 of the Export-Import Bank Act of 1945, codified at 12 U.S.C. 635g(d). The amendment also directed Eximbank to report on activities to enhance the opportunity for growth and expansion of small businesses and entrepreneurial enterprises. 283 Eximbanks 1990 Annual Report(p. 14) merely lists the dollar
PAGE 190
Chapter 4-U.S. Trade Policy 185 value of products exported with Eximbanks help that 1) meet the descriptions in the text (key linkage industry, high value added, etc.), or 2) are related to products meeting the descriptions in the text. This information by itself does not show how strategic concerns influence Eximbrmks decisions, or what particular measures Eximbank has taken to meet the goals listed in the text. 284 Export-hnport Bank of the United States, Report to the U.S. Congress on Export Credit Competition and the Export-Import Bank of the United States for the Period Jan. 1, 1989 through Dec. 31, 1989 (August 1990), p. 4. 285 Ibid. 286 Organisation for Economic Cooperation and Development op. cit., ch. 25. 287 These were satellite earth stations, turbines, locomotives, and rail signaling equipment. ~ ExPort-~Port Bank of the United States, Report to the U.S. Congress on Tied Aid Credit Practices, Apr. 20, 1989, ch. IV. The subseetors were central office switching equipmen~ satellite earth stations, radio communications, boilers, turbines, rail signaling equipment, and earth moving equipment. Earth moving equipment was not divided into subsectors, so it counts both as a sector and a subsector. 289 Ibid., ch. III. 290 Export-Import Bank of the United States, Report to the U.S. Congress on TiedAid Credit Practices, op. cit., pp. 6, 14-15. The other major donors are Canam France, Germany, Italy, Japm and the United Kingdom. 291 Ibid., p. 16. 292 In 1988 Japan granted $9.1 billion, the United States $9.8 billion. Japans Sk has been growing relative to the United States. U.S. Congress, General Accounting Office, Economk Assistance: Integration of Japanese Aid and Trade Policies, NSIAD-90-149 (Gaithersburg, MD: U.S. General Accounting OffIce, May 24, 1990), p. 6. 293 Ibid., p. 15. 294 Ibid., pp. 6-7, 295 Ibid., pp. 14-15; Export-Import Bank of the United States, Report to the U.S. Congress on Tied Aid Credit Practices, op. cit., p. 8. 296 U.S. Congress, Generrd Accounting OffIce, Economic Assistance, op. cit., pp. 11-13. 297 Export-Import Bank of the United States, Report to the Congress under Section 15(g) of the Export-Import Bank Act of 1945, As Amended, p. 1, attached to letters horn John Macomber, President and Chanman, l@ort-Import Bank of the United States, to the President of the Senate and the Speaker of the House, Dec. 3, 1990. 298 Export-Import Bank of the United States, Report to the Congress under Section 15(g) of the Export-Import Bank Act of 1945, As Amended pp. 2-3, attached to letters from John MaComber, President and ChairmarL Export-Import Bank of the United States, to the President of the Senate and the Speaker of the House, July 8, 1991. 299 These $53.8 million counted against @baoks ceiling on direct loans of $612 million for fwcal year 1990. 300 Export-Import Bank of the United States, Report to the Congress under Section 15(g) of the Export-Import Bank Act of 1945, As Amended, pp. 2-3, Attachment 2 (this report was attached to letters from John Macomber, President and ~ Export-Import Bank of the United States, to the President of the Semte and the Speaker of the House, hdy 8, 1991). 301 Export-Import Bank of the United States, Report to the U.S. Congress on Tied Aid Credit Practices, op. cit., ch. 11 U.S. Congress, General Accounting Office, Econowc Assistance: Integration of Japanese Aid and Trade Policies, op. cit., p. 6. 302 United States Trade and Development Program Annual Report, 1990, pp. 6-7. 303 Ibid., p. 35. 304 Ibid., pp. 19,21, 305 Ibid., p. 3. 306 Jomthan Raymon@ Director, Congressional Affairs, U.S. Trade and Development progr~ personal communicatio~ Aug. 2, 1991. TDPs total budge~ which includes administrative expenses, has run 5 to 10 percent higher. 307 150 U.S.C. App. 2401-2420. This Acts authorization lapsed as of Oct. 1, 1990, but President Bush continued its provisions in force by invoking the International Emergency Economic Powen Ac4 50 U.S.C. 1702. See Executive Order 12730, Continuation of Export Control Regulations, Sept. 30, 1990. The export of purely military items is controlled by other statutes. 308 The Control List is found at 15 CFR 799 (Jan. 1, 1991). Additional controls on software, technical daa and know-how are specif3ed in a separate list found at 15 CFR 779. These lists are found within the Export Administration Regulations, 15 CFR, ch. VII, subchapter c. 309 According to BXA statistics, in 1990 BXA granted approvals to individual license applications for $63 billion worth of goods. However, this figure represents the amounts of goods authorized for shipment; not all amounts authorized were necessarily shipped. Also, this figure does not include five types of special licenses covering many shipments at once. The most important of these is the distribution license, which enables exporters to ship a defiied product line for up to two years. BXAS fiscal year 1990 annual report (ch. 2, in press) gives an informal estimate of $17.9 billion for products shipped in fiscal year 1990 under distribution licenses. Acknowledgeable BXA official guessed that goods shipped during calendar year 1990 under all special licenses perhaps totaled between $30 billion and $40 billion. Personal communication Apr. 25, 1991. Given this uncertain information a fair estimate for calendar yea 1990 might be $55 billion for goods shipped uncles individual licenses and $35 billion for goods shipped under special licenses. The total then would be $90 billion for goods shipped under license in 1990. 310 U.S. Commerce Departrnen~ International Trade A&mm stratioq Oftlce of Trade and Investment Analysis, personal communication Aug. 2, 1991. 311 NAS Z (complete citation given in the next footnote), pp. 116-117. 312 Panel on the Impact of National Security Controls on International Technology Transfer, Committee on Science, Engineering, and Public Policy, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, Balancing the National Znterest: U.S. National Security Export Controls and Global Econonuc Competition (Washington DC: National Academy Press, 1987). In 1991 the National Academy of Sciences published a second report. Panel on the Future Design and Implementation of U.S. National Security Export Controls, Committee on Science, Engineering, and Public Policy, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, Finding Common Ground: U.S. Export Controls in a Changed Global Environment (Washingto~ DC: National Academy Press, 1991 ). These reports are sometimes caUed the Allen Schmitt Reports, respectively, after panel chairmen Lew Allen and Roland Schmitt. These reports will be referred to here as NAS Z and NAS H. 313 This calculus of the costs and benefits of export controls will no doubt be affected by the attempted coup in the Soviet Uniom and that Unions partial disintegratio~ in August 1991. As this report goes to press, it is unclear where these events will lead; they have the potential to both increase the urgency of technology transfer to promote social change and increase the risk of military conftict. 314 NAS 1, op. cit., pp. 187, 191; NASH, op. cit., pp. 100-101. While some allies also imposed some reexport controls, they were less intrusive. 315 This term is an imprecise shorthand. The countries at issue 292-889 0 91 7 QL: 3
PAGE 191
186 l Competing Economies: America, Europe, and the Pacific Rim include former members of the defunct Warsaw Treaty Org anizatiorL the Peoples Republic of China, North Kor~ and Vietnam, among others. Within these countries, CoCorn for years has distinguished China as requiring less control than the Soviet Uniorq and recently CoCorn has moved to reduce controls to Eastern European countries whose new governments represent a lesser strategic tbrea~ provided that those countries institute export control systems of their own to keep technology from the Soviet Union and other, unreformed bloc members. So far, Czechoslovakia, Hungary, and Poland have qualified for this special treatment. 316 Title II of the Omnibus Trade and Competitiveness Act of 1988, Public hiW 100-418. 317 S. 320. This bill passed the Semte on Feb. 20, 1991, and was referred jointly to the House Committee on Foreign Affairs and the House Committee on the Judiciary on Feb. 26, 1991. This bill resembles the conference vmsion of H.R. 101-4653 and S. 101-2927, as passed by the House and Semte in 1990, but omits certain provisions that President Bush found objectiomble in vetoing that bill. See President George Busk Memorandum of Disapproval for the Omnibus Export Amendments Act of 1990, Nov. 16, 1990. 318 Offkeof Licensing Statistics: CY 1990, n.d. (preparedly Bureau of Export AdmmI stratiou U.S. Commerce Department). This percentage is by number of applications, not dollar value. (This figure includes all individual license applications, not just those required by East-West controls.) 319 lkstimony of Dennis E. Kloske, Undersecretary for Export Administration, U.S. Department of Commerce, at hearings before the House Committee on Banking, Housing, and Urban Affairs, subcommittee on International Finance and Monetary Policy, Mar. 7 ad 28, 1990, Serial No. 101-1030, p. 19. (The figures include all individual license applications, not just those required by East-West controls.) 320 CoCorn Document CL (90) 42 (listing categories adopted on May 22, 1991 Preparatory Meeting). 321 Glennon HarrisorL Library of Congress, Congressional Research Semice, The Bush Administrafions New Approach to Export Controls and CoCom, 9@316E, June 26, 1990, p. 4. 322 Ibid., pp. 4-5. 323 Senate Committee on Banking, Housing, and Urban Affairs, Export Administration Act Amendments of 1990, Report to Accompany S. 2927, House Report No. 101-399, pp. 41-43. 324 Ibid., p. 41. 325 William Clements, Director, 0fi3ce oflkchnology and Policy Analysis, Bureau of Export Administration U.S. Department of Commerce, personal communicatio~ June 4, 1991. 326 Public Law 100-418, sec. 2415, codifkd at 50 U.S.C. App. 2404(B)(2). 327 President George Bush, Memorandum of Disapproval for the Omnibus Export Amendments Act of 1990, Nov. 16, 1990. 328 There is some question as to whether the economically less advantaged CoCorn members can implement a snftlciently reliable system of controls. J.fnot, a two-tiered CoCorn system could be required, with countries in the fm~ more reliable, tier still controlling exports to countries in the second tier. The EC, for which trade barriers between its member states are anathema, is helping less advanced nations to establish reliable control systems so as to avoid the need for a two-tiered CoCorn. 329 The Allen Panel studied the export control systems of Canada, France, Jap~ the United Kingdom, and Germany and concluded that [w]ith limited exceptions, none of the[se] countries imposed controls that extend beyond the CoCorn lists. NAS 1, op. cit., p. 99. 330 Thestatutoryprovision is found at50U.S.C. App. 2404(c)(6). The 6-month period may be extended 6 months at a time based on a finding that the item is not sufficiently available from other countries to render the U.S. control ineffective; the control may also be extended by at most 12 months to conduct negotiations to bring the item under multilateral control. An order removing East-West controls on many items 6 months after the Act is found at 54 FederaZRegister 8281 (Feb. 28, 1989). 331 50 U.S.C. App. 2404(f). The President could maintain the controls for at most an additional 18 months if he deemed it necessary for mtional security, provided that he negotiated with other countries to remove the foreign availability. 332 Public Law 100-418, sec. 2418, codifkd at 50 U.S.C. App. 2404(f). 333 BXA, Of.tlce of Foreign Availability, Completed Assessments (Based on claims under Trade Act-As of May 15, 1991). Sixteen of the assessments were requested by interested firms; one was requested by a Ikchnicxd Advisory Committee. 334 BXA 1989 Annual Reports, pp. 29-31, 34; William Clements, Director, Office of lkchnology and Policy Analysis, BXA, personal Communication Aug. 9, 1991. South Korea achieved this status only in December 1990; it and Singapore enjoy only the f~st of severrd levels of liberal treatment. Indus@ would like to see the level increased and to see other Southeast Asian mtions brought in. 335 Public Law 100-418, section 2414, cdfkd at50 U.S.C. App. 2404(a)(4),(5). 336 BXA 1989 Annual Repoz p. 26. Congress had instructed that these changes be implemented within 90 days of the Ac~ or by Nov. 21, 1988. 337 BXAAnnualReports: 1990, ch.2(inpress); 1989, p.22; 1988, p. 22. The 1991 figure is from William Clements, Director, Ofllce of lkchnology and Policy Analysis, Bureau of Export Adrnuu stratiou U.S. Department of Commerce, personal communicatio~ June 4, 1991. 338 The 1989 figures are fromNASZZ, pp. 80, 82; the 1990 figures are from Office of Export Licensing Statistics: CY 1990, n.d. (prepared by BXA). 339 341 342 343 345 346 347 348 NAS 1, op. cit., p. 114. Ibid., p. 113. Ibid., p. 191. Ibid., p. 113. NAS II, op. cit., pp. 98-99. NASl, op. cit., pp. 114-115. Ibid., p. 114, footnote. Ibid., p. 214. Ibid., p. 114. U.S. Comzress. General Accounting OffIce, Export Controls: Advising U.S. Busi~ess of Policy Changes, NSIAD-90-201 (Gaithersburg, MD: U.S. General Accounting Office, May 18, 1990), pp. 4-9. 349 BXA Annual Reports: 1990, ch. 2 (in press); 1989, p. 16. 350 Offke of Export Licensing Statistics: CY 1990, n.d. (Jl~piUed by BXA). 351 BXA 1990 Annual Repo@ ch. 2 (in PIIXS). 352 William Clements, Director, Ofilce oflkchnology and Policy Analysis, personal communicatio~ Aug. 22, 1991. 353 Alan Macmillan and Michael Andrews, BXA, ~lce of Foreign Availability, Foreign Availability Assessment: ATCompatible Microcomputers (December 1988); AT Assessment Chronology, n.d. (prepared by BXA). 354 U.S. Department of State, CoCom High Level Meeting: Fact Sheets, reprinted in Senate Report No. 101-399, p. 40. 355 Charlie Carter, Vice President for lkchnology, National Machine Tool Builders Association and industry member of the TAC for automated manufacturing, personal communication, Apr. 10, 1991; Representative of a high technology firm with representatives on several TACS, personal communicatio~ Apr. 11, 1991.
PAGE 192
Chapter -.S. Trade Policy l 187 356 50 U.S.C. App. 2409@. 357 Fact Sheet on Export Control Procedures, White House Press Release, Dec. 13, 1990. 358 President George Bush, Memorandum of Disapproval for the Omnibus Export Amendments Act of 1990, Nov. 16, 1990. 359 William Clements, Director, Offlceoflkchnology and Policy Analysis, BXA, personal communication, Aug. 9, 1991. 360 Ibid. 361 William Clements, Director, Gfilce oflbchnology and Policy Analysis, BXA, personal communication, June 4, 1991. 362 To decontrol an item controlled by CoCorn, CoCorns permission is required. CoCorn routinely reviews one-third of its list each year; to get CoCorn to consider an item out-of-turn is possible, but requires special effort. 363 The only other basis for decontrol would be that an item is no longer strategic. Often foreign availability occurs before strategic value is lost. 364 Steven Goldman, Director, Offke of Foreign Availability, personal communicatio~ June 26, 1991. 365 This figure is for FY 1989. BXA 1989 Annual Report, p. 1. 366 Steven Goldman, op. cit. 367 NASII, p. 98. 368 Ibid., p. 96. 369 50 U.S.C. App. 2504(g)(l). 370 NASIZ, pp. 194-195. 371 Personal communication Aug. 27, 1991. 372 Personal communicatio~ Apr. 11, 1991. 373 Personal commu.uicatio~ Apr. 10, 1991. 374 See U.S. Department of Commerce, Bureau of Export Administration, Office of lkchnology and Policy Analysis, 1990 AnnuaZ Foreign Policy Report to the Congress (Jan. 21, 1990-Jan. 20, 1991), February 1990, chs. I-III. 375 Under Section6 the President is supposed to consider foreign availability and the positions taken by other countries, and the impact of these factors on the controls likely effectiveness, However, how these factors are weighed is up to the Presidents discretion. See 50 U.S.C. App. 2405(b),(h). In additioq for many years the interagency review of license applications have km different for the two types of controls. However, in December 1990 the President direeted that the procedures in place for Section 5 controls also be used for Section 6. Fact Sheet on Export Control Procedures, White House Press Release, Dec. 13, 1990. 376 56 Federal Register 10,756, 10,760 (Mar. 13, 1991). 377 56 Federal Register 25053 (June 3, 1991). BXA originally found foreign availability on Sept. 25, 1989. See also Semiannual Report to Congres+Report on the Operations of the Office of Foreign Availability: Apr. 1, 1990-Sept. 30, 1990, p. 12. 378 56 Federal Register 40494-40502 (Aug. 15, 1991). The regulation will later be extended to cover use in making missiles. 379 Clements, personal communication Aug. 27, 1991, op. cit. What duty of inquiry fwill have is unclear. However, Mr. Clements has strongly urged exporters to ask their lawyers what their legal exposure might be and what precautions they should take. 380 See Fact Sheet on Enhanced Proliferation Control Initiative, issued by the White House Dec. 13, 1990, reprinted in Bureau of National Affairs, International Trade Reporter, Dec. 19, 1990, p. 1933.
PAGE 193
Chapter 5 EC-92: Trade and Industry Policy
PAGE 194
CONTENTS Page THE EUROPEAN SINGLE MARKET . . . . . . . . . . . . 193 EC-92 AND U.S. BUSINESS: ECONOMIC GROWTH AND TRADE POLICY . . . 195 EconomicGrowthDuetoE~-92 . . . . . . . . . . . . . 196 Opportunities for U.S. Exports . . . . . . . . . . . . . . 196 Opportunities for U.S.-Owned Companies in Europe . . . . . . . . . 202 EC Technology Policy . . . . . . . . . . . . . . . 209 Framework Program.. . . . . . . . . . . . . . . . 210 EUREK A . . . . . . . ., . . . . . . . . . . . 211 Emphasis on Cooperation . . . . . . . . . . . . . . . 215 EC-92: COMPETITION POLICY AND EFFICIENT FIRMS . . . . . . . 226 European Antitrust Policies . . . . . . . . . . . . . . 226 Subsidies and Nationalized Companies . . . . . . . . . . . . 228 Box Box Page 5-A. The Politics and Institutions in the European Community . . . . . . . 194 Figures Figure Page 5-1. U.S. and EC Tariff Duties for Selected Commodities, 1990 . . . . . . . 198 5-2. Automotive Market Shares in Western Europe in 1989.. . . . . . . . 202 5-3. Scientists and Engineers per 10,000 Labor Force . . . . . . . . . 210 5-4. U.S. Patents by Nationality of Inventor . . . . . . . . . . . 211 5-5.JESSI Program: Functional overview and Structure of the European Microelectrionic Program . . . . . . . . . . . . . . . . . . 215 5-6. ESPRIT I Funding Allocated by Sector . . . . . . . . . . . 221 5-7. Size of Partners in RACE Program . . . . . . . . . . . . 221 5-8. Europes Trade Deficit in Electronics in 1987 . . . . . . . . . . 225 5-9. Total Merger Activity, All National, Community, and International Mergers by Combined Takeover . . . . . . . . . . . . . . . 227 Tables Talbe Page 5-1. Macroeconomic Consequences of EC Market Integration for the Community, 1992-98 196 5-2. U.S. Balance of Trade With the EC,1985 and 1989 . . . . . . . . 197 5-3. Direct Investment Outflows From Japan . . . . . . . . . . . 203 5-4. Japanese Production Capacity in Europe . . . . . . . . . . . 203 5-5. Motor Vehicle Production in the EC,1973-1989.. . . . . . . . . 206 5-6. Japanese Share of the European Car Market, 1989 . . . . . . . . 206 5-7. Funding and Duration of the Framework Program . . . . . . . . . 212 5-8. Glossary and Description of the Framework Program . . . . . . . . 213 5-9. Funding of EUREKA Projects, by Area . . . . . . . . . . . 214 5-10. EUREKA Projects: Number of projects by Funding Category . . . . . . 214 5-11. Industrially Significant ESPRIT I Results . . . . . . . . . . 225 5-12. Shares of Manufacturing Industry Held by Large Firms in Japan, the United States, and the EC, l986 . . . . . . . . . . . . . . . +.. 227 5-13. State Aids in the EC, 1981-86 . . . . . . . . . . . . . 229
PAGE 195
Chapter 5 EC-92: Trade and Industry Policy The European Community (EC) is entering a new stage in its efforts toward greater unity. Most European history for the past thousand years centered on states establishing, defending, and expanding national borders. Now the power of these states will be altered in fundamental ways. This stems from the commitment to make the EC into a single market by the end of 1992 (EC-92). The goal of completing the internal market was first adopted in the Treaty of Rome, which established the European Economic Community in 1957. However, progress was slow until, in 1985, the Commission of the EC 1 proposed a long stride forward, in the form of a White Paper that listed 300 specific policy actions, with the goal of removing barriers to the free movement of goods, services, people, and capital throughout the EC. While EC-92 will not be completed by the end of 1992, the reforms already introduced and the growing commitments of governments, industry, and the public mean that this process is not reversible. Just what that means is debatable. According to some observers, businesses will become more efficient when the barriers to trade and cross-border operations fade and competition increases. The more optimistic forecasts estimate that growth rates could increase by 50 percent or more, while reducing inflation and unemployment. Although the size of the growth spurt is subject to debate, it is likely that the economies of the EC will get at least a temporary shot in the arm from the single market. If EC countries are to open their markets fully to each other, it will be difficult for the individual nations to continue to create special advantages for their own firms. But that is precisely the aim of most existing national industrial policies in Europe. Thus, EC-92 could lead to either the elimination of some kinds of national industrial policies, or their control and supervision by the Commission. Some policies, such as the control of credit favored in France until the mid1980s, could be swept away entirely, in this case by the removal of barriers blocking the flow of capital across borders. Other policies, such as controls over large mergers, will be transferred to the Commission. Many, such as subsidies policy, will remain in the hands of the member states but under the Commissions supervision. Not all the important differences among member nations have been ironed out. Subsidies are at the heart of a significant remaining dispute. Some member nations, led by Britain, oppose government intervention to save EC-owned firms threatened by Japanese or U.S. competition. Others, with France in the lead, are determined to keep alive European companies-often their own national companies in key industries such as electronics and motor vehicles. Although some national industrial policies will likely be constrained under EC-92, the forces that generated them remain strong. Government intervention in cases of market failure is widely accepted in Europe. Hence, the EC is under pressure to develop an industrial policy of its own, one that does not favor industry in one EC country over that in another but seeks to improve the competitiveness of European industry. In principle, the EC is committed to free trade. In practice, that is only partly true. While some early fears of Fortress Europe were exaggerated, the EC governments do use trade policy to protect European industry. Some defensive policies are longstanding (auto quotas dating back to the 1950s); some are new (rules of origin for semiconductors that favor manufacture within Europe); and some are now pursued more aggressively (antidumping). Whatever the sources, current policy is not set in stone. EC officials seem to have considerable latitude in their interpretation of trade policy. And the outcome of the GATT 2 Uruguay Round, ex pected to conclude in 1993, may change existing policies considerably; for example, the EC has put its discrimin atory public procurement policies on the table. Overall, however, EC trade policy does have significant protectionist components, e.g., high tariffs, quotas, aggressive use of antidumping laws, dis criminatory public procurement policies, protective rules of origin. EC trade policy is less multilateral than it appears. It is aimed more at Japan and other Asian countries than at the United States. Key trade policies
PAGE 196
192 l Competing Economies: America, Europe, and the Pacific Rim notably antidumpting--have been used to dampen Japanese competition in strategic sectors, such as consumer electronics. This country-specific selectivity reappears in reciprocity clauses that extend specific benefits to trading partners only when reciprocal access is available. EC efforts to protect European industries have encouraged importers to become local manufacturers through foreign direct investment, 3 and there has recently been a large influx, especially from Japan. U.S. firms have had a major manufacturing presence in Europe for decades, and many (e.g., IBM, Ford) are now regarded as European in many respects. Other U.S. companies, however, are developing or improving their presence in Europe, not only to take advantage of the EC single market but also to expand sales into Eastern Europe. This strategy could prove tough for smaller companies. This accelerated foreign direct investment offers a new set of challenges to the EC. For example, while the EC has used trade policy to protect its semiconductor producers, and through technology policy has spent billions of dollars trying to accomplish the same end, a strong Japanese manufacturing presence in the EC challenges both. The Commission has responded by moving toward policies that encourage specfiic kinds of investment likely to bring high wage jobs, technology transfer, work for European suppliers, and development in depressed areas. The EC also seeks to improve the competitiveness of existing European firms through other policies. One method is technology policy aimed at supporting generic research through R&D consortia; another is support for large multinational projects like Airbus; and a third is competition policy (analogous to U.S. antitrust policy), which will discourage subsidies and state ownership. Other policies may exert indirect effects. For example, regional policy is aimed not only at redressing the inequalities of income and wealth within the EC, but also at helping failing industries and underdeveloped regions to adjust. It is too early to assess the results of many of these policies on European competitiveness. The consequences of encouraging local production, especially in high-technology industries, remain unclear. Technology policy has been expensive and the payoffs for competitiveness remain indistinct at best, although it may have encouraged the development of an integrated European economy. Cross-border mergers that result in the formation of giant firms may improve competitiveness in some industries but not others. Progress toward deregulation of national public procurement markets could encourage EC firms to become more competitive, as could efforts to control national subsidies and limit their duration. Trade between the EC and the United States is no small affair. Taken together, the 12 nations of the EC are the largest market for U.S. exports; in aggregate they are the biggest U.S. trading partner. Individually, 4 members of the EC (Germany, the United Kingdom, France, and Italy) are among the top 10 in volume of total trade with the United States. In 1990, the United States ran a small merchandise trade surplus with the European Community, although the range of variation with individual countries was wide; Germanys $8 billion trade surplus with the United States in 1989 was fourth highest among all U.S. trading partners, and Italys $4.7 billion was eighth. Europe is also the location of the bulk of U.S. foreign direct investment. The effects of EC-92 policies on trade between the EC and the United States and on U.S.-owned companies are not entirely clear either, but are not likely to be large. The most direct effects of EC-92 on the United States will come through trade policy, public procurement policy, and new mechanisms for setting technical standards. EC trade policy differentiates between goods produced within the EC and those imported into it, disc ruminating against the latter. The new EC public procurement policy does the same, providing large benefits for goods where at least half the content is added in the EC. And as Europe develops new procedures for setting ECwide standards to replace existing national standards, U.S. firms have worried both that they will be frozen out of the process and that EC firms will use the standards to keep out foreign-made goods. 4 These policies will have different effects on U.S. exporters and U.S. firms that manufacture in Europe. Many of the policies that accompany EC-92, e.g., tariffs, quotas, antidumping enforcement, rules of origin, and public procurement, will encourage exporting firms to relocate production to the EC. Firms that do not relocate could face significant trade barriers, more so in important sectors such as computers, semiconductors and telecommunications equipment. Exporters should benefit somewhat from the macroeconomic effects of EC-92, for growth
PAGE 197
Chapter 5--EC-92: Trade and Industry Policy l 193 brings increased demand that should expand U.S. exports. It is likely that any resulting boost to exports will be modest. It is too early for a conclusive judgment on U.S.-EC trade, especially as the EC has indicated that it will negotiate over many protectionist policies. U.S. manufacturers in the EC should encounter expanded opportunities as national barriers and preferences are removed. Conversely, they may also face stiller competition from other foreign competitors into Europe (mostly Japanese firms), or from European firms themselves. Besides its effects on U.S. exports and U. S.owned firms operating in Europe, EC-92 is interesting as a source of policy ideas and options for the United States. In many ways, EC-92 is responding to a set of issues this country is also grappling with in the face of formidable international competition. Europes solutions may not be those we would choose, but it is a very large working laboratory for policy experiments. As its efforts take effect over the next few years, they will bear close scrutiny as to their possible relevance to the United States. THE EUROPEAN SINGLE MARKET The push to create a unified market in Europe began in 1957 with trade policy and the slow reduction of tariff barriers among members, which reached zero in 1968. Nonetheless, barriers remained. Even the process of crossing the border was inordinately complex and time-consumin g, sometimes taking days for commercial vehicles. 5 In 1985, the Commission of the European Communities (CEC) introduced a package of radical reforms in a White Paper. 6 This ambitious program aimed to remove barriers blocking the free flow of people, goods, services, and capital among EC countries. The package boiled down to 287 specific actions, each of which had to be negotiated in detail with the 12 member governments, affected interest groups, the European Parliament, and even third parties like the United States. This would have been impossible without a key institutional change, the Single European Act (SEA), which allows most agreements to be pushed through by modified majority vote among the member countries; a single country can no longer block adoption. These changes take place in the institutional and political context described in box 5-A. Progress in removing barriers among EC nations is substantial but incomplete. The process of approving and implementing measures is complex and time-cons uming. Most measures have been agreed on in principle, but as of mid-1991, only about one-third had been finalized, and only a handful have been adopted as law throughout the Community. Serious obstacles remain, such as harmonization of monetary policy. Yet there is little doubt that the process is irreversible. Member states have made too many crucial commitments and central reforms have already been implemented. For goods, the hundreds of documents needed to cross borders have been reduced to one, and the Commission aims to remove all border formalities in the next few years. Quantitative restrictions are formally illegal among members, and the EC is working to resolve some signicant exceptions (e.g., autos). 7 Taxes are being reformed so that indirect taxes do not dis criminate against imported items. 8 The free flow of goods can still be restricted on health and safety grounds, but the European Court is interpreting this loophole very narrowly. For capital, three directives remove national controls on deals in stock and mutual shares: 1. 2. 3. long-term commercial credits and securities not traded on exchanges; admission of securities to capital markets; and investments in short-term securities, current and deposit accounts, and loans. Once the three directives are fully irnplemented, 9 there will be no barriers to the free movement of capital within the EC. Services have always been tightly regulated in Europe; even the right to establish a business has been limited. That will now change, as companies operating in any EC country will have increasing access to services markets in all the other EC countries. Banks, for example, will operate under the key principles of the single banking license, mutual recognition, and home country control. These will create a single market for banking services, regulated by the authorities of the banks home country. This also sets off a race toward deregulation, as banks will tend to setup shop in the country with the least restrictive regulatory regime. 10
PAGE 198
194 l Competing Economies: America, Europe, and the Pacific Rim L Box 5-AThe Politics and Institutions in the European Community 1 Institutions-There are four institutional power centers in the EC: the Council of Ministers, the Commission of the European Communities (Commission), the European Parliament, and the European Court of Justice (European Court). The Council of Ministers represents national political power within the EC and is the strongest of the four centers. European Council meetings, 2 are attended by heads of government for the 12 countries. More specialized topics are discussed by Councils of particular ministries-e. g., the ECs Science and Technology Framework Program is considered by Ministers of Science, Technology, or Research. The Foreign Ministers Council is the most important of the specialized Councils. Until 1985, legislation considered by the Council of Ministers had to be passed unanimously, so that even the smallest state held an absolute veto over the affairs of the EC. In 1985, the EC passed the Single European Act (SEA), which applied majority voting to all decisions affecting the implementation of the single internal market, except fiscal decisions and those relating to the free movement of people and workers rights. The decisions needed to implement the single market will be taken largely by qualified majority voting. 3 The Council is chaired by the President of the EC, an office that rotates biannually among the member states, in alphabetical order. The Commission is the executive branch of the EC and has the sole right to make proposals to the Council of Ministers. Under its current President, Jacques Delors of France, the Commission has been a dynamic force pushing member states to give up sovereignty on the way to the single market, and then economic and even political union. The Commission is divided into 24 Directorates General (DGs), the equivalent of Ministries. Each is led by a Commissioner, of which there are 17 (the President included), one nominated by each small member state and two by each large one. 4 There are powerful tensions between DGs and Commissioners who strongly support free market ITMS ~t~ is &am from VSMUS (JE!C documents and Alberta Sbragi~ The European Community and rnstitutiomd Development: Politics, Money, and Law, Brookings Institution forthcoming. 2Some~w called EC s ummit meetings. 3Votes in the Comcil are weighted roughly by population: the United Kingdom, France, Germany, and Italy hold 10 votes eack SP@I 8; Belgium, Greece, the Netherlands, and Portugal 5 each Denmark snd Ireland 3; and Luxembourg 2. Proposals pass underqualified majority voting when they receive at least 54 out of the 76 total votes. Hence a blocking veto requires at least two major countries plus one medium sized one. Britain has been the country most consistently opposed to decisions but overruled under the SEA. 4~ m~~r states with tWO CO missioners are France, Gc rmany, Italy, Spa@ and the United Kingdom. Article 48 of the Treaty gives workers the right to to its free market principles, allowed the Japanese move freely within the EC to take up offers of electronics giant Fujitsu to buy the premier British employment. 11 This right has been expanded so that workers can now also seek work and bring family and some dependents. 12 There are still difficulties, such as the mutual recognition of professional qualifications, but EC citizens are now in essence free to work anywhere in the EC. Going further, France, Germany, the Benelux countries, and Italy have recently signed the Schengen agreement, which will eliminate all border checks on people. Spain and Portugal also intend to participate. Other aspects of EC policy are not as settled. Disagreements remain, especially over the issue of subsidies and trade protection for firms whose ultimate owners are mainly European. This rift among member countries was highlighted in disputes over how to treat companies in the electronics sector. Britains conservative government, sticking computer firm, ICL, in 1990. At the opposite pole, the French Government announced in April 1991 that it would provide $1.5 billion in subsidies to its two state-owned electronics companies-the computer manufacturer Groupe Bull SA and the defense and consumer electronics group Thomson SA, both of which experienced large losses in 1990. The justification, given by French Minister of Industry Roger Fauroux, was that the future competitiveness of entire sectors of European industry depended on the survival of European electronics firms. 13 The EC Commission, which has supervisory power over industry subsidies, was reported to be divided over whether to approve the French subsidy. It would be a mistake to view the progress toward EC-92 as a flood propelling all before it in one direction. In tone and even intent, various EC
PAGE 199
Chapter 54X-92: Trade and Industry Policy 195 principles (notably Leon Brittan of the United Kingdom) and those who support a more interventionist industrial policy (notably Pandolfi of Italy and Delors). Until recently, the Parliament has been relatively weak essentially a talking shop. Direct elections strengthened the Parliament; its role was also enhanced by the SEA. It can now delay but not block proposals. The powers of the Parliament will be extended. The European Court acts somewhat like the U.S. Supreme Court. The Commission occasionally uses the Court to define its own rights and to interpret the key legal statutes of the EC. The Treaty of Rome and subsequent amendments, notably the SEA, provide the written constitution on which the Court bases its decisions. The Court may become increasingly important, as the EC seems to be moving away from a governmental system based on administrative discretion toward a more rule-driven model of government. Politics-Industrial policy in Europe has attracted political backing from three key points on the political spectrum. While the left has abandoned a fully socialist economy in favor of a mixed model after World War II, it seeks to mitigate market outcomes and prevent market failures. The mix varies, as do the mechanisms chosen. 5 Christian Democratic parties have developed corporatist ideas based on notions of reciprocal rights and obligations between the state and the major social groups. These parties usually have tight relations with major producer groups-both employers and unions-and see negotiation with these groups as the basis for economic and industrial policies. Nationalist tendencies springing from the right also support industrial policies. Gaullism stresses the importance of strategic industries and national independence, for national security reasons. That implies national control of the technology base. Together, these three philosophies have given industrial policy the respectability that it lacks in the United States. Still, industrial policy is not universally viewed in Europe as either successful or appropriate. The tide of political opinion in Europe turned against it following the election of Margaret Thatcher as British Prime Minister in 1979. 6 Thatcherites view industrial policies as expensive, inefficient, and corrupt. They replaced nationalization with privatization. EC-92 itself in many respects rejects traditional European industrial policy; it is committed to make markets work and reduce the role of the government in the economy. sFOr a good ex~ple of moderate left ~ gin the post-War period, see Anthony Crosland, The Future of Socialism (Westpo~ CT: Greenwood Press, 1977). 6Even in Sweda, tie s~i~ Dernomatic party has reeently been forced into a mainstream agenda influenced by ~tcherism. directives are at loggerheads with others, reflecting and standards give EC-made goods an advanthe differing views of the members on how close government-industry cooperation ought to be. Even where several policies converge toward one result, they are not entirely consistent and probably were not designed to be so. Rather, they evolved in that direction. EC-92 AND U.S. BUSINESS: ECONOMIC GROWTH AND TRADE POLICY tage over U.S. exports; and a longer term competitiveness effect, in which EC support of technology advances and collaboration among European firms might improve their performance compared with U.S. fins. Nearly everyone expects a boost in economic growth from EC-92. The questions are how great it will be and whether it will outweigh any losses U.S. firms may suffer. EC-92 policies will affect U.S. exports and sales by U.S. firms located in Europe EC-92 will affect U.S. companies in three ways: differently. The distinction is important from the standpoint of U.S. national interests. Success in a growth effect, in which an expanded EC exports tends to strengthen the national economy market will offer greater sales opportunities to and raise standards of living. Success by U.S.-owned l both European and American fins; firms operating in foreign countries is not so closely a near-term protectionist effect, as some EC tied to the Nations interests, though it may be policies governing trade, public procurement, indirectly supportive if profits from the foreign
PAGE 200
196 l Competing Economies: America, Europe, and the Pacific Rim Table 5-lMacroeconomic Consequences of EC Market Integration for the Community, 1992-98 Trade Public Financial supply Total barriers procurement services effects value Change (percent) GDP . . . . . . . 0.4 0.5 1.5 2.1 4.5 Consumer prices . . . . .0 -1.4 .4 .3 -6.1 SOURCE: Paolo Cecchini, Michel Catinat, and Alexis Jacquemin, The European Cha//enge: 1992: The Benefits ofa Sing/e Market (Aldershot, U.K.: Gower Press, 1989). Paolo Cecchini was Chairman of the Cost of Non-Europe Steering Committee, and Special Advisor to the Commission of the European Communities. ventures come back to enrich American citizens or enhance domestic investment. Economic Growth Due to EC-92 EC-92 is still in early stages of implementation, so its effects on growth are not yet discernible. However, most analysts expect EC-92 to increase the ECs productivity by removing intra-European trade barriers, encouraging restructuring, promoting competition, and boosting investment. All this is expected to result in faster economic growth. The original estimates on the medium-term (199298) growth effects of EC-92 were produced by Paolo Cecchini for the Commission. Relying extensively on macroeconomic analysis, Cecchini used OECD and EC economic models to produce a set of estimates for the medium-term impact of EC-92 on the European economy over 6 years, as described in table 5-1. This estimate from the Commission strongly influenced European policymakers. It suggested that there was a specific policy route available that would increase the rate of economic growth by around 50 percent every year for 6 years, with further effects later on (the Cecchini estimate of 4.5 percent extra growth over 6 years is the average of a range of estimates, from 3.2 to 5.7 percent). According to the report, still more growth might result with changes in macroeconomic policy, such as increased public investment and reduced income taxes. In that case, gross domestic product (GDP) gains might be as high as 7 percent over the 6 years, and employment gains could rise from about 1.8 million new jobs to 5 million, at the cost of a smaller reduction in consumer prices (4.5 instead of 6.1 percent). The Cecchini report has not been accepted unchallenged. While Richard Baldwin sees dynamic gains from EC-92 that dwarf the estimates made by Cecchini, 14 most other American (and several European) economists regard the Cecchini estimates as optimistic. Merton Peck, for example, concludes that 1.5 to 2.5 percent additional growth in GDP is a more justified figure, although he concedes that his estimate is itself only a guess based on previous experience of GDP gains with removal of trade barriers, and that his work excludes the dynamic gains stressed by Baldwin. 15 Keynes long ago pointed out that a primary determin ant of economic activity was the animal spirits of entrepreneurs, and EC-92 already has been a shot of adrenalin for Europe. Europhoria may be no more appropriate than Europessimism, but it does have the benefit of making the business community more optimistic and more likely to invest, hence improving chances for the economy to grow faster. How much faster is, for now, an unanswerable question. Opportunities for U.S. Exports If EC-92 is successful in making European firms more competitive, both imports and exports would be likely to increase. This growth in trade, added to the growth of the EC market, could help U.S. exports, ass uming they are able to take advantage of the growth. An obstacle to increased U.S. exports could lie in the ECs trade and procurement policies. Since most European countries already have some protectionist trade and procurement policies, only EC policies that move in the direction of greater protection will make it harder for U.S. exporters to sell in Europe. But the early fears that EC-92, while dissolving trade barriers within the Community, would erect a Fortress Europe against outsiders have abated. An overall move toward more protectionism against non-EC traders now appears unlikely. However, there are important exceptions affecting specific industries. One of these is the electronics complex, including computers, semiconductors, and telecom-
PAGE 201
Chapter 5--EC-92: Trade and Industry Policy 197 munications equipment. This is the largest and most important category of U.S. exports to the EC. Table 5-2 shows exports and trade balances for the industries accounting for most U.S. manufacturing exports to the EC in 1989 (the last full year for which detailed trade figures are available). It seems that changes in trade and procurement policy will make little difference to two of the three industries that are our biggest exporters to Europe--chemicals and aircraft. The electronics situation is murkier. Without question, EC trade, procurement, and technology policies are all targeted toward helping European firms compete more vigorously in the computer, semiconductor, and telecommunications industries. Although this effort is unlikely to cripple U.S. exports, it could dampen them. U.S.-EC trade in chemicals is large and two-way; the $9.9 billion in U.S. exports to the EC was almost matched by U.S. imports from the EC. This reflects the fact that both the United States and the EC are home to strong, competitive chemicals industries (though EC companies are larger and have a bigger share of world sales 16 ). There are no new departures in EC-92 trade or procurement policy likely to hinder expanded U.S. chemicals exports to the EC as that market grows. The new, more unified system of standards and testing could make it easier for U.S. firms to export to the EC, because their products will have to pass only 1 set of tests and requirements, not 12; however, it is unclear whether U.S. exports could be tested in the United States or would have to be tested in Europe (see the discussion of standards below). Firms just beginnin g to export to the EC may find it more difficult, because competition within the EC market will be keen. 17 The U.S. aircraft industry is not only a big exporter but also a big generator of trade surpluses for the United States, in both EC and world trade. It is getting increasingly tough competition from Airbus, however. The large, sustained subsidies to Airbus from European governments are no small part of its success (see ch. 8), but those subsidies have nothing to do with new policies under EC-92. Table 5-2U.S. Balance of Trade With the EC, 1985 and 1989 1985 1989 (billion dollars) (billion dollars) Share of Us. Us. Us. Us. U.S. exports Product category exports imports Balance exports imports Balance (in percent) Total trade . . . . . . . . . . $49.0 $67.8 --------. .-. Manufactures . . . . . . . . . 36.8 Chemicals . . . . . . . . . . 5.9 Basic manufactures . . . . . . . 2.5 Paper . . . . . . . . . . . 0.3 Textiles . . . . . . . . . . . 0.4 Iron and steel . . . . . . . . . 0.2 Nonferrous metals . . . . . . . . 0.3 Metal manufactures . . . . . . . 0.6 Machinery and transport equipment . . . 20.6 Computers and office machines . . . . . 6.5 Electric machinery and parts . . . . . . 2.7 Power generating machinery . . . . . . 2.4 Non-consumer telecom equipment . . . . 0.9 Aircraft and parts . . . . . . . . 3.4 Construction and supplies purpose machinery . 1.7 Agriculture machinery and tractors . . . . 0.2 Machine tools and metalworking . . . . . 0.3 General industrial machinery . . . . . . 1.4 Motor vehicles . . . . . . . . . 0.1 Auto parts . . . . . . . . . . 0.5 Miscellaneous manufactures . . . . . 4.6 Science and control instruments . . . . . 2,2 Photography, optical, and time equipment . . 0.6 Clothing . . . . . . . . . . 0.1 Footwear . . . . . . . . . . 0.0 Toys and games . . . . . . . . 0.1 56. 8 6.2 10.8 0.6 1.4 3.0 1.3 1.2 27.1 1.8 2.6 3.2 0.3 2.0 2.7 0.9 0.8 2.7 7.9 1.5 9.7 1.1 0.9 1.8 1.1 0.3 $-18.8 -20.0 -0.3 .3 -0.3 .0 -2.8 -1.0 .6 .5 4.7 0.1 -0.8 0.6 1.4 .0 -0.7 -0.5 -1.3 -7.8 -1.0 .1 1.1 .3 -1.7 -1.1 -0.2 $86.6 71.7 9.9 5.4 0.7 1.0 0.4 0.7 1.0 42.7 11.2 5.5 5.4 1.7 9.4 2.6 0.6 0.7 2.9 1.0 1.0 11.0 4.1 1.4 0.3 0.1 0.4 $85.1 75.5 9.1 13.4 0.8 1.7 3.1 1.3 1.7 36.2 2.7 4.3 4.6 0.6 3.5 4.8 0.9 1.1 4.5 6.7 2.5 13.2 2.0 1.4 1.7 1.3 0.4 $ 1.5 -3.8 0.8 -8.0 -0.1 -0.7 .7 -0.6 --0.7 6.5 8.5 1.2 0.8 1.1 5.9 .2 -0.3 -0.4 .6 -5.7 1.5 .2 2.1 0.0 .4 -1. 2 0.0 24% 26 27 19 16 25 12 15 21 27 44 21 37 23 39 23 25 26 22 10 8 31 36 40 14 24 25 a 1985 data includes Spain and Portugal. SOURCE: National Association of Manufacturers, from Department of Commerce data.
PAGE 202
198 l Competing Economies: America, Europe, and the Pacific Rim None of the EC-92 trade and public procurement policies should change the prospects for sales of U.S. aircraft in Europe very much; developments in the competition among Airbus, Boeing, and McDonnell Douglas are likely to affect U.S.-EC trade in aircraft much more than EC-1992 (see ch. 8). 18 EC Trade and Public Procurement Policies New EC-92 trade and procurement policies do affect U.S. electronics exports directly. Europe is the leading destination for computer and electronics exports from the United States, and an important market for these industries. As shown in table 5-2, computers and office machines alone accounted for $11.2 billion of U.S. exports to the EC in 1989, generating an $8.5 billion surplus for the United States. Electrical machinery and parts-mostly semiconductors-added another $5.5 billion in exports and nonconsumer telecommunications equipment $1.7 billion; together, these two sectors generated a $2.3 billion trade surplus. As shown in figure 5-1, EC tariffs on computers and telecommunications equipment are the same as or a little lower than U.S. tariffs on the same itemsaround 5 to 8 percent. The largest tariff in the group is on semiconductors, at 14 percent; the United States has had no tariff on semiconductors since the early 1980s. The EC semiconductor tariff has not stopped U.S. exports. In fact, the Commission can waive the tariff when a certain chip is needed by an EC producer and is not made there or is in short supply. According to one estimate, about 20 to 30 percent of semiconductor imports faced no tariff in 1988. 19 What is new is the ECs rule of origin for semiconductors, introduced in February 1989. Under this rule, the country of origin is defined not by the testing and assembly of chips, as had been the case, but by the location of wafer fabrication (where the diffusion process occurs) .20 Wafer fabrication is the most important and technically demanding part of semiconductor manufacture. The final step, testing and assembly, adds only 10 to 15 percent of the chips total value; wafer fabrication constitutes about 60 percent. 21 Semiconductors that do not qualify as EC-made (i.e., are not fabricated in Europe) face the 14-percent tariff. But more important is the combination of the semiconductor rule of origin with strong preferences for EC-made goods in public procurements and the settlement of antidumping actions. EC directives allow public purchasers in four sectorswater, energy, transport, and telecommunications-to reject bids that have less than half EC content by value. (These sectors are excluded from GATT rules that govern public procurement, Figure 5-I U.S. and EC Tariff Duties for Selected Commodities, 1990 Passenger cars Heavy trucks VCRs TVs Telephone sets Switching equipment Parts: sets/switches Computers Computer monitors Semiconductors Photocopiers Photocopier parts 14 14 0 5 10 15 20 25 = U.S. = E C NOTE: Japan had no tariffs for these products. SOURCE: Office Of Technology Assessment, 1991, derived from published tariff schedules for the United States, EC, and Japan.
PAGE 203
Chapter 5--EC-92: Trade and Industry Policy 199 and therefore are designated the excluded sectors.) Even if the purchasers choose to consider non-EC bids, they are required to give a 3-percent price advantage to goods and services of EC origin. The four sectors account for at least one-quarter of all public procurement in the EC 22 (some estimates are as high as one-half 23 ) and for many high-technology purchases. Public procurement, as defined by the EC, covers not only governments but also firms that benefit from exclusive geographical rights or barriers to entry in their industry and businesses that governments control through the granting of licenses. This means, for instance, that British Telecom, a private firm licensed by the British Government, follows the public procurement rules. The U.S. International Trade Commission reckons that public sector purchases account for as much as 90 percent of U.S. companies telecommunications equipment sales in the EC, and up to one-third of EC sales by computer companies. 24 Semiconductors are essential components of the computers and telecommunications equipment covered by the public procurement rules. On the other hand, U.S. exporters could benefit from some changes in public procurement under EC-92, especially as they apply to purchases outside the excluded sectors. The new rules require greater openness and less discrimination in public purchases. Assuming tough enforcement, this could break the hold that many national companies now enjoy as favored suppliers-an important EC goal, for favoritism is costly. One of the main sources of the economic growth and reduced prices expected in EC-92 is lowered costs arising from competition in public procurement. The Commission has a measure of control over the national governments in this regard. Public purchasers must disclose annual projected procurement figures and report on contracts awarded in the past year. 25 So far, however, the Commissions enforcement powers are weak. For example, when the Danish Government specified that only Danish labor and materials could be used in building a bridge, the EC stopped short of taking court action to halt the project and reopen bids. Instead, it agreed to an out-of-court settlement in which Denmark promised to change its ways in the future and allow the excluded bidders on the bridge project to sue for compensation. Nevertheless, even this much is progress. The excluded sectors also fall under new nondiscrimination rules, but with more exceptions and looser criteria for openness. The preference for EC suppliers in these sectors remains one of the biggest worries for U.S. exporters facing EC-92. GE, for example, noted that it was hard to justify spending upwards of $200,000 on bids to provide power plant equipment when the bids need not even be considered. Other U.S. firms have made the same point, emphasizing that the EC rules differ from the U.S. Governments Buy American preferences, which give U.S.-based bids a price advantage of 6 percent (12 percent for small businesses), but then require that all bids be considered on an equal basis. The uncertainty under the EC rules may be enough to deter bids from any company that cannot confidently claim 50-percent EC content. Some U.S. producers of semiconductors are fearful that the restrictions on public sector purchases could spill over into the private sector (or public purchases outside the excluded sector). A firm selling computers, say, in more than one market might not want to make different product lines for each, and so would have reason to buy chips from EC sources to help satisfy the 50-percent requirement. Thus the public procurement rule could set the standard for the whole market. These fears may be exaggerated. U.S. firms have done quite well exporting semiconductors, computers, and telecommunication equipment to the EC, despite the European tradition of making public purchases from national champion firms. 26 U.S. exporters stand to benefit from the new rules for greater openness to all bidders. Furthermore, there is a possibility that even in the excluded sectors, U.S. exports may eventually win the same treatment as EC goods. In the Uruguay Round of GATT negotiations, the EC has shown interest in an expanded GATT Government Procurement Code, which would extend to signatories equal treatment in all public procurement (except defense). Meanwhile, however, the combination of new rules of origin and discrimin atory public procurement is pushing foreign firms to manufacture in Europe rather than export to Europe goods formerly made at home. While that may be good news for U.S. multinationals, it could be bad news for exporters, especially small ones without the scale of operations to justify locating in Europe. Another EC trade policy that affects U.S. exports, especially in electronics, is its recent aggressive pursuit of antidumping actions. From 1985 through
PAGE 204
200 Competing Economies: America, Europe, and the Pacific Rim 1990, the EC initiated 209 antidumping actions, concluded 190, imposed penalties in 67, and made agreements to remedy the injury to EC producers in 51, often by shifting high-value-added production to the EC. 27 The most significant actions were against Japanese, other Asian, and Eastern European companies; hardly any were directed against U.S. firms. 28 But aggressive antidumping actions and penalties add much weight to the factors that encourage foreign firms to locate and manufacture in Europe, and that has consequences for U.S. exports. The EC has often settled cases against Japanese producers with agreements that, in the future, the goods in question can avoid antidumping penalties (i.e., additional duties) by including at least 40 percent non-Japanese content. This doesnt necessarily mean EC content, according to the Commission. However, there are reports that some manufacturers of machines using semiconductors assume that it does mean EC content, or at least it might, and to be on the safe side they design in EC-made chips rather than U.S.-made chips. 29 In the well-known Ricoh case, the EC did go after photocopiers made in the United States. In November 1988, Ricoh copiers, made in Japan, were found to be dumped. The EC assessed 20-percent antidumping duties (the regular EC duty is 7.2 percent). Three months later, the Commission charged that Ricoh had increased production in California, where it was simply an assembly operation, and that the real manufacture was occurring in Japan. Therefore, the EC ruled, the California-assembled copiers should face the same duty as if they were allJapanese. Moreover, an EC rule adopted in July 1989 extended new rules of origin to copiers. It stated that the country of origin would be determined by where technically sophisticated components, such as the various printed circuit boards, lenses, various motors and high-voltage generators were manufactured. 30 Nowhere is it explicitly stated that these high-technology components would have to be made in the EC in order for a photocopier to escape penalties in an antidumping case, or perhaps to be considered for public procurement. But again, that may well be the practical effect. EC Standards, Testing, and Certification A central part of EC-92 is the creation of a unified standards and testing system in Europe. This means that goods that pass muster in 1 of the 12 member nations should be accepted in all of them. PanEuropean acceptability should be a boon for European as well as foreign firms, since it relieves them of the need to meet varying standards in different countries. For many U.S. companies that positive effect will outweigh any negative ones. Still, there are worries about negative effects. One has to do with the content of the standards and another with the means of testing them. Initially, U.S. companies were concerned that the new EC standards were being written behind closed doors, and that they might be designed so restrictively as to handicap all but European firms. This fear has receded somewhat as EC standards-making bodies have allowed U.S. companies a more substantial voice in the process, for example, by letting them comment on proposed standards at an early stage, before it is too late to make any difference. Concern about the content of the standards has not vanished, however. U.S. firms find that they must push their own interests quite aggressively in the standardsetting process. Even for large, well-organized companies this requires a great deal of vigilance, and for smaller ones it maybe impossible. For many U.S. exporting firms, however, the biggest remaining worry is not so much the content of the EC standards as how their products will be tested to make sure they conform to the standards. A new testing system is part of ECs new unified approach to standards. Before EC-92, all 12 nations had to agree before any European standard could be adopted. The process was glacially slow and produced standards that were sometimes immediately obsolete. EC-92s new approach is based on a two-pronged strategy. First, the EC has taken direct jurisdiction over regulated products that involve health and safety risks to consumers and the environment. The Commission writes broad essential requirements for these products into directives, 31 which then become EC law. Nonregulated products will follow national standards, but with the crucial proviso of mutual recognition; that is, any good that can be legally sold, manufactured, and marketed in one member nation should be equally salable in another. 32 For regulated products, manufacturers can meet the essential requirements by submitting their products to testing by an independent laboratory, which is itself licensed as a notfiled body by a member government. 33 Another route to meeting the requirements is known as self-certification. The ECs
PAGE 205
Chapter 5--EC-92: Trade and Industry Policy l 201 standard-setting bodies, CEN, CENELEC, and ETSI, 34 are writing detailed standards that will meet the requirements for regulated products. Manufacturers may choose to follow these detailed standards. This involves testing their products themselves for the various characteristics required (or getting an outside lab to do so), following approved quality assurance methods, and keeping documentation; they can then put the official CE stamp of approval on their own products. Self-certification is something of an honor system. It can be challenged by rival producers, in which case the self-certifying company would have to show proof of its claims. And the EC may decide to limit self-certification to products in which environmental and safety risks are not great. There are advantages to the other route for approval of regulated productssubmission to testing-since it allows innovative departures from the detailed standards set by CEN, CENELEC, and ETSI so long as the ECs broad essential requirements are met. However, the problem for U.S. firms (as for other non-EC companies) is to get their products tested in their home territory. Considering that parts and components, as well as final products, may require testing, shipping these overseas could be totally impractical, especially for small companies. U.S. industries and government agencies are pushing the EC to license U.S. test laboratories as notified bodies, or else allow EC notified bodies to subcontract testing to U.S. labs. 35 Such devices have precedents. For example, Canada accredits U.S. labs that have passed the inspection of a Canadian official. As of mid-1991, the EC had initially refused to accredit U.S. labs and was taking a narrow view of possible subcontracting; for example, it appeared that quality assurance audits could not be conducted by outsiders. Possibly, testing may be done in affiliated labs that EC notified bodies set up in the United States. There is already a move in this direction, and while it might solve manufacturers testing problems, it could also put some U.S. labs out of business. 36 Mutual international recognition of standards is also possible, but here there are difficulties, too. The EC will deal only with a national authority, and the U.S. system is highly decentralized. Standards for various products are set by hundreds of public and private bodies at the National, State, and sometimes regional levels. Testing and certification matters were far from settled in 1991; all were under negotiation. Discussions were held in June 1991 between U.S. Secretary of Commerce Robert Mosbacher and EC Commission Vice-President Martin Bangemann. These yielded promising prospects for conformity assessment facilities in the United States to receive EC notified body status, allowing them to carry out the full range of required conformity assessment procedures under the EC directives. This would be conditional on the successful conclusion of a mutual recognition agreement allowing, in return, EC bodies to have full participation in U.S. conformity assessment systems. 37 The Commission, however, has not yet received a mandate from the Council to negotiate such agreements, and it is still not clear what the terms of such an agreement would be, except that third country governments would have to ensure the technical competence of the facilities concerned. Both sides also renewed their commitments to the promotion of international standards. Standards for nonregulated products are relatively unaffected by EC-92. Although the EC standardssetting bodies are working on harmonizing detailed standards for these products too, that is a long-range task. Meanwhile, the mutual recognition rule still applies to these products, as it has since 1979; no EC member nation can exclude a product that meets another members standards, solely on the grounds of having a different standard. Finally, if the EC gains wide adoption of its standards throughout the world, U.S. exports could lose some luster. The EC has already announced its intention to give Eastern Europe a special place at the standards-setting table, and to give technical assistance on standards to non-European countries in the Mediterranean, South America, India, and Southeast Asia. 38 Some of the ECs draft policies on standards go further in this direction. For example, some standards include patented technologies. Standard-setting bodies in most countries, including the United States, require that the owner of the patent must license it, on fair and equal terms, to anyone wishing to produce to the standard and sell in that countrys market. However, one draft EC policy would require the patent holder to license it equitably only to EC producers, or to producers in countries that adopt the EC standards. 39 This would not only exclude firms in the United States and Japan from the required licensing and equitable treatment
PAGE 206
202 l Competing Economies: America, Europe, and the Pacific Rim (since these countries are not necessarily going to adopt the EC standards for their own markets), but would be an incentive for other countries to sign on to the EC standards. Quite possibly, this last draft policy will not be adopted. Certainly, U.S. exporting firms oppose it. It does illustrate the fact that American and other foreign firms must carefully watch out for their interests in the standard-setting process and express their concerns forcefully. overall, it appears that U.S. exporters will gain from the unified EC-92 standards and testing regime, and the problems it raises can be manageable, though not without effort. Opportunities for U.S.-Owned Companies in Europe EC-92 could confer some special benefits on American-owned companies already established in Europe, at least in the short run. Ford opened its frost auto plant in Europe 80 years ago, and both Ford and General Motors have been operating as European firms for decades. They have plants in several European countries, along with widespread sales and service networks, and they are leading sellers in European markets (figure 5-2). IBM dominates the European computer market. Other U.S. firms have longstanding alliances with European firms, e.g., GE Aircraft Engines with the French firm SNECMA. Many of the U.S.-based multinationals have strong ties with local suppliers, and some have world-class R&D labs in their host countries (e.g., hightemperature superconductivity was discovered in IBMs R&D facility in Berne, Switzerland). Such companies will have no trouble qualifying their products as EC-made goods, and thus will escape tariffs and quotas. Many EC-92 reforms are specifically designed to disrupt cozy relationships between national governments and national champion fins; in some areas of public procurement, U.S. multinationals will be able to compete on a more equal footing with EC-owned companies. The advantage some U.S. firm have in already being there will not last forever. Japanese firms are latecomers as foreign direct investors in Europe (as they are in the United States), but that is changing fast (table 5-3). In 1989, for example, there was just 1 Japanese-owned semiconductor wafer fabrication plant in Europe (there were 12 U.S.-owned), 40 but at least 3 more were under construction in 1991. 41 Three Japanese auto plants in EC countries were Figure 5-2Automotive Market Shares in Western Europe in 1989 (bn=billion) Fiat $1.99bn 15% GM -. $1.39bn 11% SOURCE: Kevin Done, W Europe Car Sales Continue Rising: Volkswagen Group Keeps Place at Top of League, )%wmial 7imes, Jan. 22, 1990, p. 3. turning out 295,000 cars per year in 1990, but 6 new Japanese plants will be up and running by 1994, producing 934,000 units a year (table 5-4). Tariffs, quotas, anti-dumping actions, and public procurement policies are driving Japanese firms to establish branches in Europe, with the prospect of a growing market under EC-92 as the reward. Foreign Direct Investment in Electronics in the EC EC policies that favor goods made in Europe are most prominent in the electronics sector, covering everything from semiconductors to computer printers to compact disk players. Many of these policies are longstanding, predating the passage of the Single European Act of 1985 and the decision to push forward with EC-92; some are national policies of member states. Notably, European computer firms have long enjoyed favored status in government purchases; public sector purchases make up about 15 percent of the European computer market, and most of that goes to each nations own champion firms. Groupe Bull in France and Siemens-Nixdorf in Germany each make about one-third of their domestic sales to their own governments. 42 The stateowned Bull has received about 7 billion francs (more than $1 billion) in subsidies since 1983, 43 and the French Government proposed to give it still more in 1991.44 The EC is committed to ending member states support for national champion companies, a support that never achieved its purpose. The heavily subsi-
PAGE 207
Chapter 5--EC-92: Trade and Industry Policy l 203 Table 5-3--Direct Investment Outflows From Japan Total Europe North America (billion dollars) (percent of total) (percent of total) 1980 . . . . . $4.7 12.3% 34.O% 1985 . . . . . 12.2 15.8 45.0 1986 . . . . . 22.3 11.5 46.8 1987 . . . . . 33.4 19.7 46.0 1988 . . . . . 47.0 19.4 47.5 1989 . . . . . 67.5 21.9 50.2 NOTE: Fiscal year from April to March. SOURCE: Organization for Economic Cooperation and Development (OECD), OECD Economic Surveys 1989/1990: Japan (Paris, France: 1990), table 14, p. 52. Table 5-4-Japanese Production Capacity in Europe a (000s of units) 1990 1994 Toyota Burnaston, U.K. . . . . . . O Lisbon, Portugal . . . . . O Honda Swindon, U.K . . . . . . o Nissan Washington, U. K. . . . . . 200 Barcelona, Spain . . . . . O Mitsubishi b Boom, Netherlands . . . . . O Suzuki Linares, Spain . . . . . . 50 Esztergom, Hungary . . . . O IBC (Isuzu-GM) Luton, U. K. . . . . . . . 45 Total . . . . . . . . 295 200 15 200 200 74 100 50 50 45 934 a The capacity figures given are incremental capacity beyond that currently existing. The Barcelona, Linares, Boom, and Lutton plants are facilities already being operated on a smaller scale by European or American companies. They will continue to produce European designs exclusively for their European or American joint-venture partners while adding the capacity shown in the table to produce Japanese designs to be sold by both partners. b No final announcement of Mitsubishis plans has been made. Negotiations are understood to be held up over the Dutch Governments asking price for its ownership stake in Volvos Dutch subsidiary. Mitsubishi would take this stake to form a joint venture with Volvo to produce and market a Mitsubishi designed car. SOURCE: James P. Womack and Daniel T. Jones, European Automotive Policy: Past, Present, and Future, contractor report prepared for the Office of Technology Assessment, July 5, 1991. dized national computer companies have not done well in world markets. 45 Nor have European semiconductor producers achieved international success, despite trade protection and government assistance. Their share of the world market declined from about 25 percent in the early 1970s to around 10 to 12 percent in the late 1980s, and today their chips are sold largely in Europe. The weakness of the European computer industry has meant limited demand for high-performance, leading-edge integrated circuits. 46 In the electronic sector overall, the ECs trade deficit was about $35 billion in 1990. 47 While disavowing support for national champions, the EC nevertheless has policies that strongly favor EC-made electronics goods. The hefty EC tariff of 14 percent applies not only to semiconductors (see figure 5-1) but to several consumer electronics items (e.g., color TVs). And protection has increased in the 1980s. The EC tariff was raised from 9.5 to 19.5 percent for compact disk players in 1983, and from 8 to 14 percent for VCRs in 1985. As discussed above, public procurement in the excluded sectors under EC-92 will give European goods strong advantages. Perhaps most forceful of all has been the ECs aggressive pursuit of antidumping actions in the electronics sector. 48 As we shall see, all these EC policies favoring locally made electronics goods have powerfully encouraged foreign firms to locate more of their production in Europe. Several EC antidumping actions in recent years have been directed against the Japanese and other East Asian electronics producers. Of 149 actions initiated in the 4 years 1987 to 1990, 16 were against Japan, 18 against Korea, 6 against Hong Kong, and 6 against Taiwan. 49 Most of these actions concerned electronics products, including both office equipment and consumer products: typewriters, photocopiers, computer printers, video cassette recorders and tapes, audio cassettes and tapes, compact disk players, and small color TVs. 50 There were 63 antidumping actions against China (16), Turkey (14), Yugoslavia (11), and other Eastern European countries (22 in all), but these mostly involved low-technology products such as building materials, textiles, and apparel. Only three actions were taken against U.S. firms (two in 1987, one in 1990). 51 During the 1980s, European electronics firms protested strongly against what they saw as aggressive Japanese pricing. Many Japanese companies employ product cycle pricing, in which early models of a product are priced as high as the current cost of
PAGE 208
204 l Competing Economies: America, Europe, and the Pacific Rim producing them would warrant but lower in relation to the average production cost over the expected product cycle-which would include the period after economies of scale and learning have been achieved. Japanese firms are also known for their commitment to building market share over short term profittaking. These strategies led to accusations that Japanese firms were dumping, for which there are remedies under GATT if it causes injury to the recipient countrys industry. The EC tightened its antidumping enforcement so aggressively as to be accused in some quarters of protectionism. The EC has interpreted its antidumping regulation so as to make the finding of dumping and the proof of injury to EC producers much more likely, and the size of the injury greater. 52 EC mechanisms for determini mg the dumping margin tend to exaggerate the fair value of a product; 53 higher fair prices help to ensure both that dumping is found and that the margin between the fair and dumped price is large. The EC also tends to assume that there is a causal link between dumping and injury to home producers, rather than having to prove the link; for example, even low market shares captured by imports have been used as evidence of injury. 54 Other practices also put exportiers accused of dumping at a disadvantage. The Commission itself verifies complaints; limits access to the case file; holds informal hearings in accordance with Continental law practice, which allows ex parte communications and does not require a written record; and grants no automatic right of review. Most important, the EC does not set antidumping duties individually for each importer. Rather, one rate is imposed on all imports from the country as a whole, affecting dumped and non-dumped imports alike. In principle, importers can get their own antidumping duties reduced or eliminated by showing that they did not dump or dumped by a lesser amount, but the procedures to do this are slow, uncertain, and burdensome. 55 The EC forgoes duties in some cases in return for undertakings by foreign exporters to charge higher prices. This was the case in the EC semiconductor agreements with Japanese producers in 1989 (for DRAMs) and 1990 (for Erasable Programm able Read-Only Memories, or EPROMs). Like the U. S.Japan semiconductor agreement of 1986, these agreements set price floors, with the goal of encouraging European producers to invest in production facilities without fear of cutthroat below-cost competition from the Japanese. 56 Foreign exporters on whom antidumping duties have been imposed have also avoided paying these duties by shifting the last part of their production offshore, usually to the EC. An order directed to imports of finished goods from the exporters home country would then no longer apply. Very often, only the final assembly was shifted. The EC responded with an anticircumvention, or screwdriver assembly, rule. Under this rule, the Commission has the authority to levy the same duty as before on the finished goods if the parts and materials used come at least 60 percent (in total value) from the country or countries subject to the previous order. To avoid this provision, firms have decided to use over 40 percent EC-origin parts and materials, even though, according to the letter of the EC rule, those components and materials could be produced in third countries. 57 Japan, the target of many important EC antidumping actions, took the EC rule as it applied to Japanese plants in Europe to GATT, where it was found by a GATT panel to be illegal. 58 The EC did not abjure its anticircumvention rule, but after the GATT ruling in March 1990, antidumping actions came to a temporary standstill; however, a new action on components of disposable lighters is reported to be under way. The antidumping rules, the new rule of origin for semiconductors, and the preference for EC-made goods in public procurement all add up to powerful incentives for foreign electronics firms to locate production facilities in the EC, which they did. By 1988, Japanese manufacturers of photocopiers, electronic typewriters, and printers had 10 subsidiaries in Britain, 6 in Germany, and 4 in France. 59 But these did not suffice. They were labeled screwdriver plants. To comply with the ECs anticircumvention rule the companies then had to scramble to find European parts suppliers-not always with great success since the components sector, especially in the United Kingdom, was weak. According to a Japanese source, Japanese manufacturers looking to procure parts locally had to start by training and helping parts makers themselves. 60 Similarly, when Japanese companies started making audio cassettes in Europe, they were hit with an antidumping action that imposed duties on the magnetic film going into the cassettes. The Japanese response was to shift production of the film into the
PAGE 209
Chapter 5--EC-92: Trade and Industry Policy .205 EC. 61 As for semiconductors, the indispensable electronic guts of all these products, the ECs 1989 rule of origin is forcing still more inward investment. Fujitsu, Mitsubishi Electric, Hitachi, and Sony have started to build or have announced plans for new wafer fabrication plants (fabs) in Europe; NEC is already there. 62 It is putting it too strongly to say that the EC now has a coherent, unambiguous policy of forcing inward investment, even in the electronics sector, where the forces are most powerful. For example, when Fujitsu bought out the British computer company ICL in 1990, the company was promptly expelled from the industrys most influential lobbying group and was partially excluded from JESSI, the semiconductor research consortium heavily supported by the EC. This suggests that it could take more than local investment for a Japanese company to be treated like a European company. All the same, many EC leaders are strongly and explicitly committed to encouraging inward investment in forms that promise to create well-paid knowledge-intensive jobs and to transfer valuable technology to local supply fins. For U.S.-owned companies already well established in Europe, 63 the new Japanese presence could mean greatly intensified competition in the European market. For U.S. companies that export to Europe but have no wafer fab there, the rule of origin for semiconductors could cause problems, especially for smaller companies that can hardly afford the $200 to $300 million investment in a European plant. Intel (not a small company but not a giant either) was one of those faced with the alternative of laying out hundreds of millions for an EC plant or losing EC sales. At congressional hearings in early 1989, an Intel representative protested against the ECs domestic content policy, saying that it forced U.S. companies to transfer jobs, technology, and investments to Europe regardless of competitive considerations. 64 In October 1989, however, Intel announced a decision to build a $425 million wafer fab in Ireland, with substantial investment aids (i.e., subsidies) from the Irish Government. Foreign Direct Investment in Autos in the EC If there is some ambiguity in the ECs policy of encouraging foreign inward investment in electronics, there is more in motor vehicles. Longstanding quotas and informal arrangements that restricted Japanese auto imports, combined with newer local content requirements, spurred Japanese investment in auto plants in the EC in the late 1980s. However, in April 1991, the European Commission proposed a scheme that, reportedly, would allow only a small increase in the Japanese share of the EC market through 1998, Counting imports and transplant production as the Japanese share. At the end of July 1991, the Japanese Government reached an agreement with the CEC on auto exports and possibly production in the EC through the end of 1999. Although all the details were unclear as of August 1991, the pact apparently caps the Japanese producers shares of the overall market at 16 percent (about half imports and half domestic production) by the end of 1999. 65 Though this is a higher share than they now hold, the agreement apparently permits no growth (or a slight decline) in Japanese exports, compared to 1990, with little additional production in Europe above that already announced and under construction. Until quite recently, many policymakers in Europe considered their motor vehicle industry to be a competitive success. European automotive trade showed a positive balance through 1989, and for a remarkable 6-year period, from 1984 through 1989, European motor vehicle sales were extraordinarily robust, far higher than analysts had forecast (table 5-5). This prosperous period followed one of stagnation, in which the number of car companies shrank and employment greatly declined, but the result was a high level of capacity utilization and profits for every surviving company. 66 This seeming strength was deceptive. First, national auto champions have been protected and encouraged through a panoply of industrial policy measures in European countries throughout this century. Quotas have quite effectively controlled Japanese imports into the major auto-producing countries, as is evident by comparing their imports with those of European countries with no domestically owned motor industry (table 5-6). Italy and Japan agreed in the 1950s on a limit of 2,000 imported Japanese cars annually, and the limit stood until the July agreement, which will permit Japan to export 138,000 cars to Italy by the end of 1999. France imposed a 3-percent market share limitation in 1977; Britain and Japan reached an informal agreement in 1975 limiting Japanese imports to 11 percent of the U.K. market. Germany has no formal agreements with Japan on market share limitations. However, in 1981, when the U.S. voluntary restraint
PAGE 210
206 Competing Economies: America, Europe, and the Pacific Rim Table 5-5-Motor Vehicle Production in the EC, 1973-1989 (millions of units) Year FRG U.K. Spain Italy France Belgium Holland Total 1973 ...............3,94 9 1978 ...............4,18 6 1981 ...............3,89 7 1982 ...............4,06 2 1983 ...............4,15 4 1984 ........,......4,04 5 1985 .. .............4,445 1986 .. .............4,578 1987 .. .............4,634 1988 .. .............4,625 1989 .. .............4,852 2,164 822 1,607 1,144 1,184 987 1,156 1,070 1,289 1,288 1,134 1,309 1,311 1,418 1,203 1,307 1,389 1,704 1,545 1,866 1,626 2,046 1,958 1,656 1,434 1,453 1,575 1,601 1,573 1,913 1,913 2,111 2,221 3,569 3,508 3,019 3,149 3,336 3,062 3,016 3,195 3,493 3,678 3,920 299 303 237 278 285 249 267 295 352 398 389 107 85 101 109 122 129 128 142 152 149 149 12,868 12,489 10,859 11,277 12,049 11,529 12,158 12,633 13,637 14,372 15.203 a Estimated NOTE: Spanish production is included aithough Spainwas not amember ofthe ECuntil 1985. Because the European motor vehicle production system is highly integrated but as yet has no integrated production totals available, it isnecessaryto estimate production forthewholeof the EC by totaling production country bycountry. This method may Ieadtosome double counting, in particular of French and German vehicles assembled in Belgium and French vehicles assembled in Spain. SOURCE: Forallcountries except Holland: Automotive News, Market Data Book, 1990 edition, p. 3. For Holland: Motor Vehicle Manufacturers Association of the United States, Wodd Vehic/e Datq various years. Table 5-6-Japanese Share of the European Car Market, 1989 (in percent) Finland . . . . . . . . . . Ireland... . . . . . . . . . . Norway . . . . . . . . . . Denmark . . . . . . . . . . Austria . . . . . . . . . . . Greece . . . . . . . . . . Switzerland . . . . . . . . . The Netherlands . . . . . . . . Sweden . . . . . . . . . . Belgium and Luxembourg . . . . . . West Germany . . . . . . . . . United Kingdom . . . . . . . . . Portugal . . . . . . . . . . France . . . . . . . . . . . Italy . . . . . . . . . . . Spain . . . . . . . . . . . 39.6%. 39.4 37.8 31.8 30.3 29.9 29.4 25.9 24.7 19.2 14.8 11.3 6.1 2.9 1.6 1.1 NOTE: Sweden is an exception in having a large domestically owned motor industry and a high level of Japanese imports. However, the Japanese imports are almost entirely in the smaller size classes where no Swedish products are offered. For comparison puposes, the Japanese share of the U.S. car market was 27.4 percent in 1989. SOURCE: Financial Times, Feb. 5, 1990. For the United States, Earl Kreher, Motor Vehicle Manufacturers Association, personal communication, Aug. 7, 1991. agreement covering imports of Japanese autos was announced, it was widely reported that the Japanese Government gave these assurances to the German Government: exports would not be diverted to Germany, and Japanese companies would not increase their sales in the German market by more than 10 percent per year (about 1 percentage point of market share). 67 While the high level of trade protection and its long duration might not be proof of fundamental weakness in European auto manufacture, there is more direct evidence. The first independent global survey of auto company performance in plants, R&D operations, and supply chain management indicates that European companies, both the mass producers (Volkswagen, Fiat, Renault, and PSA) and the high-priced specialists (Mercedes, BMW, Volvo, Jaguar, and Rover), lag behind Japanese and even U.S. performance significantly. 68 They score worse in productivity, product quality,@ and responsiveness to changing demand. Finally, by the end of 1990, the European market had run out of steam; sales were falling in all markets but Germany. Exports to the United States had already dropped with the decline of the dollar, and firms with the heaviest dependence on the U.S. market (Jaguar, Saab, and Volvo) had been forced to find buyers or joint venture partners. At the same time, Japanese firms were making major investments in new production capacity within Europe, gearing up to produce 935,000 units per year by 1994 and as much as 1.2 million units in 1998. Thus it began to occur to European policymakers that the European motor industry might face in the 1990s what the Americans faced in the 1980sa catastrophic loss of market share at the hands of the Japanese. This prospect is probably much exaggerated because of the protective measures the EC has taken for its auto companies. Even so, European auto producers face an unaccustomed challenge, as do the well-entrenched U.S. auto makers in Europe. The frost mitigating factor is the extension of quotas on Japanese imports, under the aegis of the
PAGE 211
Chapter 5--EC-92: Trade and Industry Policy l 207 EC. Originally, the EC Commission intended to remove all member states quotas, beginning in January 1991 and completing the process by January 1993. 70 But this idea was conceived when the European auto boom was at its height. As the industrys sales weakened, leaders of the German industry abandoned the free-trade camp they had traditionally occupied, joining the French and Italians to urge Europe-wide share limits on Japanese imports along with a transition period for dropping individual country restrictions. Terms of the agreement made with the Japanese by the EC in April 1991 were not all public, but trade sources reported that they would allow Japanese companies to increase their sales of passenger cars in the ECincluding both imports and transplant production-horn about 1.2 million in 1990 to 2.3 million in 1999, or from about 10 percent to somewhere around 16 percent of the EC market. 71 EC estimates of additional production from the Japanese transplants by 1998 range from 900,000 to 1.4 million cars per year, which implies that virtually all of the growth in Japanese car sales would come from transplants, and that imports would stay flat. Also, these numbers imply sizable growth in the EC market-over 2.5 percent per year. 72 The EC scheme also states that the very restrictive formal national quotas on Japanese imports in France, Italy, Spain, and Portugal will be abolished at the end of 1992, but that the Japanese will voluntarily limit imports into those markets under EC and Japanese supervision, to account for unexpected circumstances [that] might arise after January 1, 1993. After 1998, the EC market would be open to the Japanese, according to the proposal. There are still parts of the agreement that are not entirely transparent. In particular, it is unclear how long the Japanese will be willing to abide by informal limitations on transplant production in the EC. We cant accept this, said Moriharu Shizume, Paris representative of the Japan Automobile Manufacturers Association. They set up local content rules-60 percent, 70 percent, then 80 percent. Now that weve got almost 80 percent, they still dont accept it. 73 Domestic content requirements for Japanese cars were set first in Britain, which was in the vanguard for Japanese auto transplants in Europe. In negotiating terms with the Japanese companies, the British Government demanded high levels of European content, both as a condition for investment aids provided by the national and local governments and in order not to count the transplant vehicles against Britains 11 percent quota on Japanese imports. 74 The result was a commitment for 60 percent European content in the initial production runs at the three largest Japanese transplants (Nissan, Toyota, and Honda, all in the United Kingdom). What is more, these facilities are committed to 80-percent European content by 2 to 3 years after startup, a level not currently contemplated by any North American transplant even after several years of production. 75 This level of local content requires that the body, most major mechanical components, and either the engine or the transmission be fully manufactured in Europe. The local content requirements have probably slowed the rate of Japanese assembly plant investment in Europe. Also, under EC urging, the British have ceased offering investment aids to the transplants, and that, too, may have slowed the rate. This does not mean that the Japanese have no further designs on producing in Europe. Another important but often overlooked reason for slow growth, both of Japanese imports into Europe and of Japanese transplant production, is the European distribution system for motor vehicles. It differs greatly from the U.S. system. The United States has always been a large market without internal barriers where similar vehicles are sold for the same price in every region. Moreover, since the late 1940s, U.S. antitrust laws have forbidden assembler firms from requiring that their dealers sell only one brand from a single site. By contrast, EC rules allow the established European auto assemblers to limit the freedom of the dealer and the customer, to their own distinct advantage. The assemblers are able to sell the same car for vastly different amounts in different countries, both in Europe and elsewhere. Their usual approach is to price very low in their own home market to establish a volume base and keep others (e.g., Japanese imports) out, and to sell at much higher prices in other markets. At the same time, European (and Europe-based American) assemblers forbid their dealers to dual, that is, to take on Japanese brands at established dealerships. It was precisely the aggressive use of dualing that allowed the Europeans (in the late 1950s) and then the Japanese (in the early 1970s) to rapidly build
PAGE 212
208 Competing Economies: America, Europe, and the Pacific Rim distribution networks in the United States with practically no investment. For all these reasonsan EC-wide quota at least for Japanese imports (and possibly an unwritten agreement governing transplant production as well), high local content requirements for transplant vehicles, and the existing distribution system-a surge in Japanese share of the European auto market is not likely over the next few years. The current EC rules allowing selective pricing and exclusive selling come up for renewal in 1995, and the result may well be rules more like those in the United States. Even so, EC-wide quotas and local content requirements would continue to provide very substantial protection for European producers. The gradual relinquishment of voluntary national quotas could cause problems for the French, Italian, and Spanish industries, but the Japanese firms have practically no distribution and service networks in any of these countries and cannot build them overnight. The likely prospect is for only a gradual increase in competitive pressure on European automakers from Japanese imports and Japanese transplants combined. 76 That goes for U.S.-owned auto firms in Europe as well. Whatever pressure they feel will not compare to the hammer blows suffered by the Big Three in the United States in the 1980s. In just the 5 years from 1985 to 1990, cars made by Japanese companies (imports and transplants) rose from 22 to 33.5 percent of a shrinkin g U.S. market (overall sales declined from 11.0 million to 9.3 million units); in the same 5 years, the market share of Big Three cars sank from 72 to 59 percent. There may even be an opportunity for some modest increase in U.S. auto exports to Europe in the 1990s, in all likelihood coming from the Japanese transplants in this country, not the U.S. Big Three. 77 Later in the decade, this eastward flow across the Atlantic might be balanced by exports to the United States and Canada from Japanese transplants in Europe. Each Japanese company might decide to produce certain kinds of vehicles at only one site in its global production system, in Europe or North America, for high-volume sales in the region of production and for exports in lower volumes to fill market niches in other regions. 78 A final note: The pressure of Japanese competition, mitigated though it may be, has finally made bedfellows of European auto producers and the American-owned firms in Europe. For 80 years, since Ford established an assembly plant in Britain in 1911, American companies have been treated as outsiders by the Europeans. For example, the organization formed in the 1970s to promote the interests of European motor vehicle fins, the Committee of Common Market Motor Vehicle Constructors (CCMC), resisted all efforts of the Americans to join. However, in the debate about how to deal with the Japanese in 1992, tensions in the CCMC became so great that 11 of the 12 members resigned in December 1990. Left behind was Jacques Calvet, PSA chairman, who was intransigent in demanding curbs on Japanese imports and transplant production. A new organization is being formed that will include the Europeans and the Americans, but will exclude the Japanese. Thus the course of events has strengthened the perception that the two American firms, with their top-to-bottom production systems spread across the continent, are now European, after 80 years on probation. It seems most unlikely that the EC will apply anything like the import or investment barriers on these fins, or their U.S. parents, that are contemplated for Japanese firms, even if the European market should prove disappointingly flat. EC-92 and the United States: Conclusions If EC-92 does substantially increase the growth of the Communitys GNP, as many Europeans believe it will, that should also increase demand for the products of U.S. companies. As internal trade barriers are cleared away, U.S. firms should find it easier to sell and distribute their products to customers throughout the EC. Market growth could help some American exporters to Europe, especially in industries whose base of operations already is global, such as the chemical industry. On the other hand, some of the ECs most effective protection is focused on computers and electronics, the biggest and arguably most important of U.S. exports to Europe. Even if demand for U.S. products does grow, EC trade and public procurement policies in some sectors will encourage U.S. firms to supply the demand from a production base in Europe. No one policy alone might force a U.S. firm to manufacture in Europe. But companies will judge their effects in the context of other factors: the new opportunities to be found in both Eastern and Western Europe; the shift of competitors, especially the Japanese, into Europe; advantages for manufacturers in getting
PAGE 213
Chapter 5--EC-92: Trade and Industry Policy .209 closer to suppliers and customers in a bigger, richer, more unified market; and uncertainties about more forceful protection in the future. These factors, combined with steady pressures of protectionism, especially in electronics, seem to be encouraging a wave of foreign direct investment, which implies some shift in the manufacturing base into the EC. Thus from the U.S. perspective, the message of EC-92 is mixed for exports to the EC but more positive for those planning or able to produce there. For businesses that would profit from locating production in Europe, not all are equally well-placed to make the shift. EC-92 could create problems for small U.S. exporting firms, although there has been no flood of complaints so far, according to the Small Business Administration. 79 It might also have ill effects on small firms that are not now exporters. As some large U.S. firms shift production into the EC, they may not make an effort to bring their traditional suppliers with them, and many of the suppliers are themselves simply too small to follow. EC Technology Policy European firms face a difficult time competitively. In several industries, years of protection and government support have sapped firms competitiveness. Now those firms are newly vulnerable, as the national policies that formerly pampered and protected them give way to EC policies that forbid subsidizing one member nations firms at the expense of anothers. Another form of competitive pressure is the trend in EC policies to treat foreignowned firms operating in Europe much like locally owned fins. As we have seen, the trend is by no means complete or universal (witness the continued EC favoritism toward European auto companies), but it is strong enough that many foreign fins, especially the Japanese, are shifting production into the EC to evade trade restrictions. A particular worry is the new challenge from Japan, a more serious matter than the longstanding American dominance of European computer and semiconductor markets. Europeans viewed with alarm Japanese companies incursions into U.S. markets where, in some industries, they wiped out the domestic competitors altogether. Europe was the next target. This new competitive pressure on European firms, and the increased intensity of international competition generally, especially in high-technology fields, spurred the Community to develop policies meant to improve the competitive performance of EC-owned firms. First, the EC is vigorously attacking the entrenched structure of European industry through a reinvigorated competition policy (more on this below). Second, it is trying to develop mechanisms to close what it sees as a crucial technology gap between Europe and both the United States and Japan. The new technology policy is an ambitious attempt to use R&D consortia, partly funded by the EC, as a means of reaching the technical frontier in some strategic sectors, notably information technology, telecommunications, and energy. Another major goal is to create a supranational, European orientation among EC fins. Technology is key to improved competitiveness, in everything from product innovation and design to manufacturing quality and efficiency. Even before the 1980s, European efforts to improve technology were afoot. In 1979, the EC Commissioner for Industry, Etienne Davignon, created a task force to study the long-term needs of the telematics sector (telecommunications and information technologies) and to draft a strategy for the EC as a whole to revitalize high-technology electronics industries. 80 Eventually, this effort spawned two new European technology development and implementation programs, ESPRIT (the European Strategic Programme for Research and Development in Information Technology), begun in 1984, and a few years later, RACE (Research in Advanced Communications for Europe). By the end of the decade, several other EC R&D programs, brought together under an umbrella called the Framework Program, had begun: BRITE, or Basic Research in Industrial Technologies for Europe; EURAM (European Research in Advanced Materials), which was merged with BRITE in its second phase; several biotechnology programs; and a host of small programs aimed at promoting research and development, science, and cooperation across the borders of the EC nations. In addition, 19 countries (the 12 of the EC plus Austria, Finland, Iceland, Norway, Sweden, Switzerland, and Turkey) participate in a project called EUREKA (European Research Coordination Agency). The total effort represented by the Framework Program and EUREKA is prodigious. The Third Framework Program (1990-94) is funded at ECU 5.7 billion, and EUREKA projects announced between 1985 and 1990 came to ECU 7.4 billion. As the
PAGE 214
210 Competing Economies: America, Europe, and the Pacific Rim Second and Third Framework Programs overlap, exact annual finding is hard to estimate, but at a rough estimate, public funding of European cooperative research in both the Framework and EUREKA programs comes to about ECU 1.6 billion ($1.9 billion) per year. 81 There is a new spirit of optimism in Europe now, and the technology programs begun in the 1980s are apart of it. But will the programs have a real effect in changing the European technological landscape? This is not answerable now, because many of the initiatives launched in the 1980s are too new to evaluate fairly; the oldest have been going less than a decade. Many of Japans more impressive technical accomplishments, such as the birth and development of microelectronics manufacturing, required far more than a decade to yield conclusive results (though when Japan began its industrial targeting and technology initiatives it was technically much further behind than Europe). But there is enough information to make tentative conclusions about the likely performance of Europes new technology efforts. So far, the programs have achieved some things-a greater degree of transborder and multiinstitutional cooperation, a greater sharing of technical and scientific information, and generally positive reviews on the part of participants about the contribution to technology-but they have not yet made much difference in competitiveness. Framework Program The broad purpose of the Framework Program is to improve living standards, which in turn means boosting European industrial competitiveness and technology .82 It was apparent to European companies and governments in the 1980s that many of the more important technological developments of the past decades had skipped Europe. Several indicators show Europes failure to keep up. A 1988 EC study disclosed that Europe was the world leader in only 2 of 37 technologies identified as economically important. (The same study concluded that 31 were dominated by the United States and 9 by Japan.) 83 Europe was also behind the United States and Japan in the concentration of R&D workers in the labor force (figure 5-3), and Japanese inventors streaked past Europeans in obtainin g U.S. patents (figure 5-4). Moreover, the areas in which Japanese and European patents were granted tell the story of European decline in electronics, communications, and information technology in particular. The Japanese patents were heavily concentrated in electronics, while Germanys were mostly in chemicals and materials, and Frances were a diverse list, promiFigure 5-3-Scientists and Engineers per 10,000 Labor Force Per 10,000 workers 8 0 70 60 50 40 30 20 10 0 I I I I I 1 I I I I I I I I 1 I I I I 1 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 ~ Franc e + W. German y + Japa n + U.K. + U.S.A. SOURCE: National Science Board, Nationa/ Science and Errgineeting Indicators, NSB 69-1 (Washington, DC: U.S. Government Printing Office, 1989), p. 262.
PAGE 215
Chapter 5-EC-92: Trade and Industry Policy .211 Figure 5-4-U.S. Patents by Nationality of Inventor Thousands 6 0 5 0 4 0 3 0 2 0 10 I ,,. A A A &h u% cm v-s 94 -f I 0 I I 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 + France West German y ++ Japan ~ U.K. ~ U.S.A. SOURCE: National Science Board, Nationa/ Science and Engineering Indicators, NSB 89-1 (Washington, DC: U.S. Government Printing Office, 1989), p. 365. nently including nuclear energy and industrial machinery. 84 The Framework Program was set up to overcome the European weaknesses. European industry remains strong in some high-technology fields chemistry, certain energy technologies, and aeronauticsbut is weaker in electronics and most information technologies (hardware and software), motor vehicles, and biotechnology. The biggest program, ESPRIT, tackles information technologies. RACE aims at improvement in telecommunications hardware and service and manufacturing technologies. BRITE/EURAMs research includes, in descending order of funding: new materials; reliability, wear, and deterioration; CAD/CAM and mathematical modeling; new production technologies for textile products; new testing methods; laser and joining technologies; membranes; and new testing methods. 85 High-technology sectors are a particular focus, probably for the same reason that they are a focus of industrial policy elsewhere: they are viewed as making disproportionate contributions to wellbeing. They are knowledge intensive, often have application in many other sectors, and contribute disproportionately to value added. In some cases, such as telecommunications, they are also a key part of the infrastructure of developed nations. The lions share of the ECU 5.7 billion in funding for the Framework program goes for projects designed to improve European competitiveness in hightechnology industries and services. 86 The combined finding for medicine and health, information and communication technologies, biotechnology, and nuclear fusion is ECU 2.7 billion, more than half the total. 87 ESPRIT II (European Strategic Programme for Research and development in Information Technologies), the biggest program, is funded at ECU 1.6 billion from 1988 through 1992. Tables 5-7 and 5-8 show the funding, purpose, and duration of the individual programs within Framework. 88 Another characteristic of the Framework Program is that it is designed to do precompetitive research. The definition of precompetitive is comfortably loose (the CEC defines it as being at a stage prior to industrial development) yet it still exerts a real effect on projects, making officials wary of approving those that seem too close to the market. Further, all Framework projects are collaborative, not just among companies but across national borders. Only projects that have participants from more than one country can be considered. EUREKA EUREKA, the European Research Cooperation Agency, was begun in 1985 at the initiative of
PAGE 216
212 Competing Economies: America, Europe, and the Pacific Rim Table 5-7Funding and Duration of the Framework Program Funding Main area Program (million ECU) Duration Quality of Life Information and Communication Technologies Industrial Technologies and Advanced Materials Biotechnology Energy Science and Technology for Development Marine Technologies Improvement of European Science/ Technology Cooperation Medicine-Health Human Genome Analysis STEP/EPOCH ESPRIT RACE DRIVE DELTA AIM EURET II BRITE BRITE/EURAM Raw materials B.C.R. Biotechnology (revision) BRIDGE ECLAIR FLAIR Agriculture Radioactive waste Decommissioning TELEMAN Fusion JOULE S.T.D. MAST FAR SCIENCE SPES Large-scale facilities MONITOR DOSES EUROTRA VALUE 65 15 115 1,600 550 60 20 20 25 60 499.5 45 59.2 20 100 80 25 55 79.6 31.5 19 551 122 80 50 30 167 6 30 22 4 7 38 1987-91 1989-91 1989-92 1987-92 1987-92 1988-91 1988-90 1988-90 1990-94 1988-89 1989-92 1990-92 1988-92 1988-89 1990-94 1988-93 1989-93 1989-93 1990-94 1989-93 1989-93 1988-92 1989-92 1987-92 1989-92 1988-92 1988-92 1989-92 1989-92 1989-93 1989-92 1989-90 1989-93 SOURCE: Commission of the European Communities, EC Research Funding: A Guide for Applicants (Brussels, Belgium: January 1990), p. 117. France and West Germany. Although it is not an EC project, the EC countries and the CEC itself are members, along with the European Free Trade Association (EFTA) nations and Turkey. Its aims are similar to those of the Framework Program-to promote cross-border collaboration in research and technology and to improve competitiveness. As a result, there is overlap between EUREKA projects and parts of the Framework Program in scope and funding. However, the programs do differ, principally in sources of funding. For EUREKA projects, funding is mostly private. National governments (and the EC) can fix their own contributions at any level they desire up to 35 percent of total project costs, so public funding is generally low compared with the Framework Program. According to one source, of the ECU 7.4 billion allocated to 388 EUREKA projects between November 1985 and June 1990, less than 10 percent was public funding. 89 EUREKA does not limit projects to precompetitive R&D. EUREKA projects can be closer to the market than projects done under the Framework Program, although not as close as some participants might wish. While some EUREKA projects involve basic research, most are on topics of immediate commercial interest. 90 Large companies involved in both Framework projects and EUREKA sometimes prefer EUREKA as a result.
PAGE 217
Chapter S-EC-92: Trade and Industry Policy l 213 Table 5-8-Glossary and Description of the Framework Program 1. Quality of Life Medicine/health: Coordinates and improves efficiency of medical and health research Genome analysis: Studies use and improvement of biotechnology in the study of the human genome Radiation protection: Improves knowledge of human exposure to radiation and effects of radiation on humans and their environment STEP/EPOCH( Science and Technology for Environmental Protection European Programme On Climatology and natural Hazards): Provides scientific and technical support for environmental policies of the EC ESPRIT ii (European Strategic Programme for Research and development in information Technologies): Develops basic technologies for European information technology industries RACE (Research and development in Advanced Communications technologies for Europe): Contributes to introduction of Integrated Broadband Communication taking into account Integrated Services Digital Network DRIVE (Dedicated Road infrastructure for Vehicle safety in Europe): Develops information technologies to improve road transportation efficiency and safety DELTA (Development of European Learning through Technological Advance): Applies advanced information technology to education and develops tools to support (long) distance learning AIM (Advanced Informatics in Medicine): Develops information technologies for improving health care services at reasonable cost EURET (REcherche clans le Transport en Europe, or Research on Transportation in Europe): Develops a Community transport system to respond to increases in demand for all types of transport necessitated by the single market 2. Industrial Technologies and Advanced Materials BRITE/EURAM (Basic Research in industrial Technology for Europe, and European Research in Advanced Materials): intended to strengthen the competitiveness of European manufacturing industry (including particularly small and mediumsized enterprises) in world markets Raw materials (Raw materials and recycling): Enhances the competitiveness of the Communitys industries involved with raw materials and recycling B.C.R. (Bureau Communautaire de Reference): improves the reliability of chemical analysis and physical measurements to achieve agreement among member states 3. Biotechnology BRIDGE (Biotechnology Research for innovation, Development and Growth in Europe): Promotes cross-border research to speed up the production of biological data, materials and processes needed for the optimal use of natural organisms ECLAIR (European Collaborative Linkage of Agriculture and industry through Research): Promotes application of new developments in life sciences and biotechnology to products for industrial agricultural use FLAIR (Food Linked Agro-industrial Research programmed): Contributes to the competitiveness, safety, and quality of Europes food industry Agriculture (Competitiveness of agriculture and management of agricultural resources): Helps farmers adapt to situations created by overproduction and restrictive price and market policies 4. Energy Radioactive waste (Management and storage of radioactive waste): Perfects and demonstrates a system for managing and sorting radioactive waste Decommissioning (Decommissioning of nuclear installations): Develops a system to manage decommissioned nuclear installations and radioactive wastes TELEMAN (TELEMANipulation dan les environments nucleaires dangereux et perturbes; or Remote handling in hazardous or disordered nuclear environments): Develops advanced remote operated equipment for the nuclear industry, and a scientific and technological basis for remote operating systems Fusion (Controlled nuclear fusion): Establishes physical and technological basis necessary for planning of NET (Next European Torus) JOULE (Non-nuclear energies and rational use of energy): Develops energy technologies that take account of new and renewable energy sources and increases security of supply of energy 5. S.T.D. (Science and Technology for Development): Promotes scientific cooperation between the EC and developing countries 6. Marine Technologies MAST (MArine Science and Technology): improves knowledge of the marine environment and promotes new expiration technologies FAR (Fisheries and Aquaculture Research): Promotes rational and scientific research on resources; develops aquaculture 7. /improvement of European Science/Technology Cooperation SCIENCE (Stimulation des Cooperation internationales et des Echanges Necessaires aux Chercheurs en Europe; Scientific and Technical Cooperation): improves general quality of scientific R&D, promotes training through research; enhances mobility of researchers SPES (Stimulation Programme for Economic Sciences): Establishes cooperation and interchange between European economists Large-scale facilities (Large -scale scientific facilities and installations): Optimizes use of large-scale scientific facilities and installations MONITOR (Strategic analysis, forecasting and evacuation): identifies new directions and priorities for Community research and technological development policy; improves evacuation of R&D program DOSES (Development Of Statistical Expert Systems): Promotes use of advanced statistical techniques; in particular, application of expert systems EUROTRA (Programme EUROpean de TRAduction automatique de conception avancee; machine transition): Develops a machine transition system capable of dealing with all official EC languages VALUE (Valorization and Utilisation for Europe; Dissemination of results): Promotes dissemination and utilization of the results of scientific and technical research, with special consideration of the needs of small and medium sized enterprises SOURCE: Commission of the European Communities, EC Research Funding: A Guide for Ap@cants, op. cit., p. 117.
PAGE 218
214 l Competing Economies: America, Europe, and the Pacific Rim EUREKA is like Framework in putting special emphasis on high-technology areas. Its heaviest funding goes to robotics and production automation, closely followed by information technology and communications. Other fields with strong support are medical and biotechnology, transport, new materials, lasers, energy, and the environment (see tables 5-9 and 5-10). 91 One of the most important EUREKA projects is JESSI, the Joint European Submicron Silicon Initiative. Begun as a l-year feasibility study in December 1986 and going through another 18 months of project definition and participant acceptance, the project started work in mid-1989 toward its goal: to produce static and dynamic random access memories (SRAMS and DRAMs) and logic chips using 0.3 micron feature sizes by 1995. 92 Major cooperators in JESSI include Philips and Siemens, who had a relationship as a result of their Megabit (Mega) project, an earlier effort aimed at producing 4megabit DRAMs (Siemens) and l-megabit SRAMS Table 5-9-Funding of EUREKA Projects, by Area Total cost, Number Project area (million ECU) of projects Energy Technology . . . . . 494 11 Biotechnology and Medical Technology 434 41 Communication Technology . . . 809 15 Information Technology . . . . 820 35 Transport Technology . . . . 559 18 New Materials . . . . . . 136 17 Robotics and Production Automation . 269 11 Environment . . . . . . 523 11 SOURCE: EUREKA Secretariat, EUREKA: Tbgether for the future (Brussels, Belgium: 1989), passim. (Philips). The Mega project, begun in 1983-84, was supported by the German and Dutch Governments to the tune of DM 320 million and DM 160 million, respectively; Siemens paid DM 1.4 billion and Philips DM 1.5 billion for production facilities. In the end, Siemens acquired production technology and know-how to make 4M DRAMs, and had an operation actually making IM DRAMs in Regensberg, West Germany. The Mega project and other development work at Siemens no doubt contributed to this success, but so did Siemens licensing of Toshibas IM miniaturization technology in 1985. 93 However the success of the Mega project maybe judged, European microelectronics companies were still behind the world leaders at its conclusion and were facing escalating competitive pressure, particularly from the Japanese, whose market share in Europe was increasing steadily. Philips and Siemens together initiated the first discussions of JESSI, which other companies later joined. By 1988, a JESSI planning group included 29 companies and institutes from six countries. They put together a multiyear ECU 3.8 billion research effort with four subprograms: technologies for chip production, chipmaking equipment and materials, applications of microelectronics, and basic research (figure 5-5). The overall objective is to secure the availability of world-competitive microelectronics for the European industry. 94 Recently, however, Philips withdrew from its leadership of the SRAM project of JESSI, probably reflecting a severe downturn in Philips financial performance and large cutbacks in operations generally. Philips will probably abandon Table 5-10-EUREKA Projects: Number of Projects by Funding Category Project cost (million ECU) BIO COM EN V ENE INF MAT ROB IA S TRA Total <5 . . . 23 4 2 5 19 11 15 2 9 90 >5<10 . . 3 1 1 2 6 2 11 0 3 29 >10 <20 . 9 3 2 0 4 3 9 4 1 35 >20 <40 . 3 2 3 1 2 0 8 2 1 22 >40 . . . 3 5 3 3 4 1 6 3 4 32 Total . . 413 15 11 17 35 17 49 11 18 208 KEY: BIO = Biotechnology and Medical Technology COM Communication Technology ENE Energy Technology ENV Environment INF Information Technology MAT Materials Technology ROB Robotics and Production Automation LAS. Laser Technology TRA Transport Technology SOURCE: EUREKA Secretariat, EUREKA: Together for the Future (Brussels, Belgium: 1989), p. 15. The total number of projects represented here is current as of June 1989.
PAGE 219
Chapter 5--EC-92: Trade and Industry Policy .215 Figure 5-5-JESSI Program: Functional Overview and Structure of the European Microelectronic Program Europrojects Application I CAD tools and environment 4 I Process development I DRAM/ V R A M SRAM EPROM Logic Technology I Manufacturing science I 4 Selected strategic areas Equipment and materials k MateLitho rials SOURCE: JESSI Planning Group. NonBack Clean Iitho end room Design methodology Modeling and simulation Advanced processing steps Process integration Basic and long-term research 1 the memory business altogether. While this is not by products before their foreign competitors. This may itself a crippling blow to JESSI, the fact that other participants are also in trouble is not a good omen for JESSIs eventual outcome. JESSI is very much a commercial undertaking. It is principally a EUREKA project but has some funding from ESPRIT as well. ESPRIT funding goes into the basic research subprogram, which accounts for only 14 percent of JESSI cost and 22 percent of the manpower. Nevertheless, the contribution to JESSI accounts for 30 percent of the funding in ESPRITs microelectronics program; some expect the overlap between the ESPRIT microelectronics program and JESSI to be nearly complete within a few years. Some of the impetus for the Mega project, JESSI, and Frameworks ESPRIT programs probably arose from European concern that competing with Japan in electronics would be a different ballgame from competing with U.S. companies. According to one analysis: Because Japanese chip producers were part of larger systems houses, foreign competitors began to suspect that systems divisions of the same Japanese companies were getting access to leading edge have been perfectly natural but it put foreign systems houses at a competitive disadvantage in getting timely access to the new parts. The resurgence of European support for semiconductors in the mid-1980s, in frameworks like the Mega project and the ESPRIT program, in some measure reflected these mounting concerns. Back in the late 1970s, reliance by European systems houses on U.S. semiconductor companies for supplies of advanced chips, though far from welcome, did not pose a strategic problem for European industry. The U.S. merchant chip manufacturers were not, for the most part, vertically integrated into downstream systems. All competitors enjoyed roughly equal access to state-of-the-art components. 95 Emphasis on Cooperation Cooperation across institutions and borders, and among customers, competitors, and suppliers, is valued by Europeans both as a means to improved competitiveness and as a contributor to European political and economic integration. Different kinds of cooperation serve different purposes. Cooperation across borders is a frequently mentioned goal, based on the underlying assumption that Europes competitive problems stemmed in large part from both fragmentation of markets, which deprived compa-
PAGE 220
216 Competing Economies: America, Europe, and the Pacific Rim nies and industries of economies of scale and scope, and fragmented research efforts, which resulted in overlapping but underfunded research and development projects. Claire Shearman points to European market fragmentation as an important cause of Europes weakness in electronics and information technology, and observes that the effect may have been aggravated by the protection of most national markets, which led European companies to focus more on domestic than on international markets. 96 Undoubtedly, the ability to achieve economies of scale is important in some parts of the electronics industry. The cost of investment in new plant and equipment; the difficulties and costs of acquiring engineering, design, and production expertise; and the efficient scale of operation are all quite high compared with most other industries. For example, a new wafer fabrication plant and its equipment can require an investment of half a billion dollars. However, with a few important exceptions, such as electronics products, telecommunications equipment and service, and large jet aircraft, limitations imposed by markets the size of most EC members are probably minor. For example, it is unlikely that national markets in Europe were too small to have allowed producers of textiles, many types of machinery, paper, and some chemicals and pharmaceuticals to produce at an efficient scale. Moreover, while the European market was far from unified and cross-border movement of goods was not without trouble, 60 percent of European trade in the late 1980s was with other European countries, indicating that European companies were already focused on European, rather than strictly national, markets. 97 In telecommunications, the fragmentation was real, but not strictly a function of market size. Telecommunications services in EC nations are provided by governments, through ministries of posts, telegraph, and telephone (PTTs); usually equipment is procured from one or two companies, frequently national champions. Because PITs retained monopolies over provision of services and tight control over procurement of equipment, different countries introduced new technologies (hardware and services) at different rates. The rates were mostly slower than in the United States, where because of deregulation private companies could offer many new services to American consumers. 98 In addition, there was no coordination of the different national services, making equipment and software incompatible from one country to the next. This made international telephone service inconvenient at best. According to one account, To establish a private line between Italy and the Netherlands in the early 1980s, a user had to deal with every PIT in between, frequently making it a nightmare to trace and correct breakdowns. Tariff differentials were striking. A 500-kilometer leased line could cost ECU 840 per month in Britain, and ECU 2,230 per month in France. European businesses reported tremendous difficulties in establishing reliable, modern, efficient telecommunications links across Europe. 99 Fragmentation of research and development efforts, public and private, is also seen as a handicap. According to the CEC, the fact that each of the large members and several smaller ones had mounted their own research efforts in high-technology sectors resulted in many small research efforts, most with inadequate resources. Coordination among them was inadequate, information was diffused poorly, and there was duplication of effort. 100 In response to the problems they see as arising from fragmentation, EC policymakers encourage cooperation of all kinds. This includes cooperation across borders, between customers and suppliers, between industry and academia, between large and small companies, and among competitors. Cooperation Across Borders International collaboration is not strictly a European development. For the past decade or so, multinational firms have expanded their cooperative relationships with firms in other countries through a variety of arrangements: joint ventures, licensing, marketing agreements and exchanges, direct and indirect investments, mergers and acquisitions, and R&D collaborations. 101 In many high-technology sectors, the firm that can best integrate its technical expertise with that of other firms is better equipped than one that relies more on in-house developments. Japanese fins, in particular, have shown how much can be accomplished by integrating technological developments from around the globe with their own strengths in manufacturing and design and close collaborations with customers and suppliers in Japan. The multinational enterprises responsible for the swell of technical collaborative arrangements around
PAGE 221
Chapter 5--EC-92: Trade and Industry Policy l 217 the world are mostly Japanese, European, or American. The EC, in promoting international collaboration in technology, encouraged a trend that already had a head of steam, but there is little doubt that the encouragement was real; many of the resulting collaborations would not have happened without the Framework program as an enticement. For example, a senior manager in Plessey Research ventured the opinion that Plessey and Thomson, as competitors in gallium arsenide research, would ordinarily not work together, but they are doing so on an ESPRIT project (Advanced GaInAs-based Devices for High Speed Integrated Circuits) l02 because you have to collaborate in order to get funding. 103 Even where the programs do not stimulate wholly new alliances, they strengthen existing ones. An evaluation of BRITE found that it was an excellent catalyst of transborder collaboration. Although three-quarters of the bilateral relationships formed in the program already existed, BRITE strengthened the relationships, according to the evaluators, and helped turn bilateral relationships into multilateral ones l04 The assessment was not all favorable; evaluators found that BRITE was not particularly effective at helping potential participants in isolated places find partners. Preexisting relationships may be the norm in other parts of the Framework Program besides BRITE. One executive of Siemens, the largest single recipient of ESPRIT funds, states that the company wants to work only with partners it has worked with previously, and does not use the Framework Program as a dating service. Bad experiences with strangers account for this policy. 105 Evidence from several quarters indicates that both Framework and EUREKA have succeeded in engendering cross-border collaboration. European evaluators seem confident that the Framework Program has done so. EUREKA has linked more than 800 organizations in more than 200 projects. The CEC notes that the Framework Program has many more applicant consortia than the available funds can support l06 Part of this success is due to the Framework Programs requirement that funded projects must have participation from different countries. 107 Part is undoubtedly driven by the new optimism for market integration in Europe and the stimulation of European technology and science initiatives generally. The threat from a strong outsider (in this case, Japanese and American technology companies) may also have played a role, especially in industries where the time and money needed for a critical mass of R&D are difficult for companies to bear individually. Customer-Supplier Cooperation The success of Japanese firms is due in part to their development of links upstream to suppliers and downstream to customers. These links allow the joint development of technologies, standards, and quality controls. The days when arms-length relationships with customers and suppliers were regarded as the best business practice are ending, especially when development costs of new products and techniques are high. Development of large jet aircraft is an example. The cost of developing the Boeing 777 is likely to be in the neighborhood of $5 billion, not counting the engine, and Pratt & Whitney estimates the cost of developing a new high-thrust engine at $2 billion (see ch. 8). High development costs are nothing new in this industry; in the late 1970s, the cost of launching the 757 and 767 exceeded the net worth of the Boeing Co., and the development cost of the DC-10 was over three ties the capitalized value of the Douglas Co. Still, aircraft development costs are rising exponentially, with the result that all the major aircraft manufacturers are asking first-tier contractors to shoulder more of the development costs. Cooperation is also becoming more important in electronics. Producers of computers, telecommunications equipment, and consumer electronics can no longer rely on off-the-shelf commodity chips as the primary building blocks of new products. Integrated circuits are now much more often application specific, and even entire systems can be integrated onto a single chip. This means that systems makers must transfer design information, including proprietary design information, to semiconductor producers, who could at least in theory convey that information to the competitors of the systems producers. 108 These technical changes make vertical integrationproduction of components as well as final products within a single fro-an increasingly attractive strategy, and reliance on a vertically integrated competitor for semiconductors an increasingly risky one. 109 All the big Japanese systems companies are vertically integrated. Few American or European systems makers are, and even those that 292-889 0 91 8 QL:3
PAGE 222
218 l Competing Economies: America, Europe, and the Pacific Rim are still have to depend on outside suppliers for key components. During the 1980s, the outsiders in question were more and more often Japanese, for the vertically integrated, well funded, and technologically proficient Japanese firms had seized the competitive high ground in semiconductors and were steadily capturing world market share. Systems companies in America and Europe became uneasy about relying on their powerful competitors for components. This unease spawned efforts to improve coordination between semiconductor equipment producers and semiconductor makers, and between semiconductor producers and systems makers, in both the United States and Europe. Cooperation between customers and suppliers in electronics became a survival tactic. In its evaluation of ESPRIT, the CEC made this point emphatically: The belief that all the industries which are becoming dependent on embedded microelectronics can develop competitively by purchasing standard components from remote and competing nations is fallacious. Close working relationships between major semiconductor users and suppliers are essential. 110 The idea of fostering vertical linkages is not confined to high-technology industries. There are advantages to coordination and cooperation even in less technology intensive industries such as textiles and apparel. For example, increasing coordination all the way along the chain from apparel design through fiber and textile production, sewing, wholesaling, and retailing is the objective of the Quick Response program of leaders in the U.S. fiber, textile, and apparel industries. The sparse information available on the subject suggests that the Framework Program has had some effect in generating links between customers and suppliers. Many of the BRITE projects involved small and medium-sized firms making capital goods (e.g., machinery and equipment) for industrial customers; here, the diffusion of technology that BRITE achieved was in the interest of all participants. lll As the evaluation of the BRITE program stated: Projects which unite customers and suppliers are usually successful because there is a community of interests. 112 JESSI explicitly emphasizes customer/supplier links. Like Sematech, JESSI recognizes the advantage of Japanese vertical integration, and one goal is to help overcome this advantage by forging stronger bonds between customers and suppliers. Its planning report states: What is lacking-above all-in Europe is the spirit of togetherness which is typical for Japan Incorporated. Japanese companies compete strongly amongst each other, but not before having shared in the effort to conquer a new market for Japan. Many semiconductor manufacturers, especially in [the] far east, are integrated into vertically structured system houses. Therefore, careful attention must be paid to defining user/producer interfaces in order to guarantee protection of the users know-how.llq User/supplier links are less prominently stated in RACE documents, but the goal is there nonetheless. Closer relations among industrial suppliers of equipment and telecommunications service providers (the PTTs) is one of the aims of RACE, where such linkage is regarded as necessary to accomplish RACEs other goals of making Integrated Broadband Communications and other innovative services available across Europe and setting common standards for European telecommunications services. 114 There is evidence that such linkages may be happening. RACE mostly aims to provide services, but participation is heavily tilted toward telecommunications equipment manufacturers. With 395 participations in RACE, equipment makers have a larger share of total participations than any other group. It can be inferred from this heavy participation, plus the positive evaluations of RACE by participants, that both suppliers (the equipment manufacturers) and users (the PTTs) find it beneficial to work together in RACE projects. 115 There are a few hints that achieving cooperative relationships between users and suppliers could be improved. The CECs evaluation of ESPRIT acknowledged that more inputs from major software system suppliers would be valuable in the Software Technology part of ESPRIT.. The program has made technical progress in software technology, but is not yet exploited commercially, and the time when all commercial progr ammers use standard techniques, and all universities teach them, is still far off. 116 The ESPRIT evaluators made similar comments on the Computer Integrated Manufacturing (CIM) projects. Computer integrated manufacturing requires an understanding not only of electronic equipment, but also of mechanical equipment; the evaluators noted that the absence of so-called mechatronic projects in ESPRIT, which united mechanical and electronic skills, was regrettable. The recommendation was to
PAGE 223
Chapter 5--EC-92: Trade and Industry Policy .219 pay more attention to user attitudes and understanding, which, more than a shortage of technology, delay wider adoption of CIM in Europe. 117 Cooperation Between Competitors This is one of the most difficult of all forms of cooperation. Competition and cooperation are not mutually exclusive, but neither are they broadly compatible. Competition, and the benefits that it brings to companies and economies, often falls apart if cooperation extends to collusion. In Japan, industrial interests and bureaucrats usually manage to balance promotion of competition-assuring that no one firm attains too great a market share within Japan-with encouragement of cooperative arrangements in R&D (and in some circumstances, production). 118 But even in Japan, with its long tradition of industrial cooperation and a deeply rooted perception of the need for national cohesion, cooperative research is difficult to manage. It took several years of administrative massaging by the director of Japans VLSI (Very Large Scale Integration) Project to overcome the suspicion and fears of the participants from five companies and MITIs Electrotechnical Laboratories before real research couldbegin.119 In Europe and America, where cooperation without collusion is less well established, the problem is greater. Evaluators of the BRITE program are unequivocal on this point. Projects which bring together competitors are unlikely to succeed. The Commission. hoped that cooperation in research might lead to such links. But these industrialists are often engaged in stern competition and their research forms the core of their business. When they do join together for research, it is usually for research that is on the periphery of their true concerns and therefore of doubtful industrial value; otherwise, disagreement soon arises. European culture is not the same as Japanese culture, and the only way any cooperation of this sort can succeed is if it is imposed at the highest level in the company. In reality, a number of projects in this category have patently failed, and the Commission must stop them. I N The evaluators of ESPRIT came to a somewhat different conclusion. They note that early work plans for ESPRIT, though put together substantially by the Big Twelve European information technology companies, 121 did not address the core business needs of the principal protagonists and concentrated on activities where competitors could actually agree to collaborate in a precompetitive way. 122 Unlike the BRITE evaluators, however, the ESPRIT review board did not regard this as reason for stopping projects; rather, the problem was treated as minor. Cooperation Between Academia and Industry European universities have long been viewed as aloof from the concerns of industry, compared with those in the United States and Japan. 123 One goal of the Framework Program is to reinforce a growing trend toward stronger relationships between academia and industry. ESPRIT and BRITE, for example, have specifically encouraged university and industrial researchers to unite in certain research projects. 124 In ESPRIT 75 percent of the collaborations include firms and academic research units. 125 Although the overall impression is that industryacademic cooperation is good for both and should be encouraged, there are a few caveats. These partnerships will help bring academic research out of the ivory tower and into the real world. But the CEC warns that universities should not let industrys R&D funding lure them into abandoning inquiries of a fundamental or scientific nature and concentrating too heavily on short-term problems of companies. While noting that some shift of emphasis was exactly what was needed, the CEC is concerned that the pendulum could swing too far. University participants in BRITE had the same concern; several academics indicated to the BRITE evaluation review team that the industrial participants emphasis on empiricism was excessive and could undermine the value of the results. 126 Further, exclusive relationships with industry could restrict the availability of research reports from universities, a problem already emerging in the United States. 127 These caveats are relatively minor. Most collaborative projects involving academia are viewed by ESPRIT participants as working either adequately (45 percent of respondents) or well (over 20 percent). 128 In BRITE, three-quarters of the contractors viewed collaboration between industries and universities as enhancing the projects chances of success. 129 However, this kind of cooperation is not sufficient to promote competitiveness in Europe without managerial know-how and skilled people to exploit market opportunities.
PAGE 224
220 l Competing Economies: America, Europe, and the Pacific Rim Precompetitive Projects As noted, the Community is trying to find a way to balance competition and cooperation and obtain the best from both. One way to strike a compromise is to limit projects to precompetitive R&D. According to the CEC, this means limiting projects to a stage prior to industrial development, or not yet market ready. This is to ensure that EC research does not lead to any distortion of competition, which would be in breach of the EC Treaty regulations. 130 This is not a precise definition. There is some understandable puzzlement over what is precompetitive, and some frustration with precompetitiveness as a requirement for acceptance. The BRITE evaluation team said it best: The panel remains puzzled by the incorporation of the concept of pre-competitivity in a research programme whose primary goal is to enhance the competitiveness of European industry. This is a major contradiction that the Commission cannot escape. Some excellent projects were not funded because they were considered too competitive by the selection team. Concern about the idea of pre-competitivity has also had the effect of pushing some projects upstream so that their industrial interest diminished. The panel fully realizes that there were good legal reasons for this criterion to be written into BRITE. But the law should be the servant of the Community, not its master. 131 This concern is echoed by some participants. For example, one member of the Industrial Research and Development Advisory Committee (IRDAC) to the EC expressed frustration that no one knows what precompetitive means, and stated that efforts to keep JESSI precompetitive were making it uncompetitive. 132 A participant in JESSI from Siemens disagrees with the perception of JESSI as precompetitive, and believes that the consortium is successfully developing complex integrated circuits that can be integrated into systems; ESPRIT, according to this researcher, had failed in this respect. 133 Another observer notes a strong conflict between precompetitive research and the process of developing a specific product. 134 The limitations imposed by the requirement that the research be precompetitive tend to show up often in comparisons of ESPRIT and EUREKA, which is not required to be as far from the market as the Framework Program. The EC will probably continue to stress precompetitivity, partly because the law constrains it to, and partly because many people still think that it is an appropriate guiding principle. 135 But certain adjustments can be made. In the second phase of ESPRIT, work continues to be precompetitive but more emphasis is placed on projects of clear benefit to industry. Among the goals of ESPRIT II are providing a sustainable capability to design and produce application-specific intergrated circuits (ASICs), developing technologies needed for next-generation information processing systems, and enhancing the ability of European industry to integrate nextgeneration processing into complete application systems. 136 Small Business The Framework Program emphasizes participation by small business. In Europe, as in Japan and the United States, most R&D is carried out by large organizations, with a few notable exceptions such as Silicon Valley microelectronics startup firms. Small businesses often lack staff and money to keep abreast of technological change and are often left behind as technology advances. To compensate, many developed nations have various types of extension and training services for small businesses; Japans is particularly well funded and extensive. 137 The ECs emphasis on including small and medium-sized enterprises (SMEs) 138 in the Framework Program is intended to improve their technological competence and ability to innovate. Both are important goals, though the former is explicit and the latter most often implied. Some programs are naturals for SMEs, and others are not. ESPRIT, whose designers were the Big Twelve European information technology companies, is not. Much of the R&D for information technology (IT) involves too great expense for most small firms to handle. While there is a natural place for small, specialized firms in supplying the large IT companies with specialty materials and chemicals, production equipment, and other niche products, even these are difficult markets to enter because of high capital requirements and the increasing domination of Japanese fins, many of which are conglomerates and operate on a global scale. Even so, ESPRIT evaluators feel that the program has been successful in attracting and funding projects for SMEs. The Big Twelve participated in 70 percent of the ESPRIT projects and received 50 percent of the funding; SMEs participated in 65 percent of the projects and received 14 percent of the funding
PAGE 225
Chapter 5--EC-92: Trade and Industry Policy .221 (figure 5-6). SMEs share of the program has increased and that of the Big Twelve has fallen. 139 RACE is another project that does not invite easy participation by SMEs. RACEs design, like ESPRITs, relies heavily on large organizations-in this case, the large PTTs that provide telecommunications service in EC countries, and telematics equipment producers. RACE evaluations are nearly mute on the topic of SME participation. The management audit of RACE in 1989 discloses that just over two-fifths of RACE partners were SMEs, but there is no mention of their share of funding (figure 5-7). 140 By size class, the largest number of participants had more than 5,000 employees. BRITE, on the other hand, was built for small enterprises; not entirely, to be sure, but substantial participation of SMEs was intended and desired in BRITE in a way that was never true of ESPRIT and RACE. The exact share of SMEs in BRITE funding is unknown; the evaluation team found that the reported figure of 31 percent is misleading because some of the companies placed in the category of SME are part of larger enterprises. 141 But it is also very likely true that a larger share of BRITEs funding goes to SMEs than is the case for RACE or ESPRIT The CEC seems generally satisfied that it is meeting its goals in attracting and funding proposals horn SMEs, at least in Figure 5-6-ESPRIT I SMEs 14% the big Framework programs. Funding Allocated by Sector The Twelve 50% SOURCE: Commission of the European Communities, Communication from the Commission to the Council and the Parliament Concerning the Pedonnance and Results of the Programme ES,RIT, SEC(89) 1348 final (Brussels, Belgium: Sept. 7, 1989), p. 12. Figure 5-7-Size of Partners in RACE Program Number of companies 120 I 110 100 t 8 0 t k 60 55 4 0 20 n 25 46 45 22 Les s 51-100 101-500 5ol 1,001More than than 50 1,000 5,00 0 5,000 Number of employees SOURCE: Commission of the European Communities, RACE W, (Brussels, Belgium: 1990), p. 93. Whether or how much SME participation is contributing to the overall aims of Framework is less clear. Large organizations do the bulk of R&D for a good reason: R&D is generally an expensive proposition, requiring specialized equipment and people, a longer term outlook, and stable financing; many small enterprises are unable to mount such efforts and most do not need to. There is merit in small organizations having the technological expertise to keep up with the larger organizations that they supply, and in that sense bringing them into R&D consortia is helpful. But the most pressing technological issue for SMEs is usually not how to do more R&D, but how to apply what is known, as the BRITE evaluators said: The challenge facing traditional S.M.E.slike many other firms-is not in fact the need to adopt new technology, but rather to integrate into their production methods technologies that already exist on the market. Such an objective, which is crucial to the future of European industry, does not fall within the scope of BRITE [T]he Commission should discuss the possibility of launching a much larger and more ambitious programme than the current SPRINT 142 programme, with suitable procedures for promoting the dissemination of technologies within the industrial fabric, and in particular for small and medium-sized firms. 143 Another salient finding of the BRITE evaluators is that SMEs benefit from contacts with larger organizations (companies and universities), in part because those larger organizations have easier access to information and can make it available to
PAGE 226
222 l Competing Economies: America, Europe, and the Pacific Rim smaller companies. Whether large firms benefit in equal measure from their participation in consortia with small ones is less clear (see the section on Sleeping Partners). International Participation in the Framework Program The Framework Program was designed to improve the competitiveness of European firms vis-vis American and Japanese firms. Naturally, this makes U.S. and Japanese efforts to join Framework Program consortia a thorny issue. Up to now, American firms that want to participate have been given access, but not always as much as they would like. For example, IBM was allowed to participate in JESSI, which receives some funds from ESPRIT, but is not permitted to take part in working groups or committees that make management decisions about projects. RACE, on the other hand, has allowed IBM to function as a prime contractor, and EUREKA projects let IBM participate as the only computer company in a given consortium. In contrast, ESPRIT projects require IBM to have a European computer company as a watchdog. 144 Although participation by IBM is becoming more accepted, this does not end all the problems, either for IBM or for Europe. IBM would like unconstrained access, especially since it has substantial manufacturing and R&D in Europe (IBM conducts 12 percent of its R&Din Europe, 8 percent in Japan, and the rest in the United States). One reason for IBMs restricted access to JESSI and the Framework Program is that European companies do not have equal access to Sematech, which Europeans view as a comparable program. More to the point, Europeans are still uneasy about access of foreign companies. While a few American companies besides IBM (and many companies from EFTA countries) have been able to participate in Framework and EUREKA projects, the question of who should get in is unsettled and controversial. An official of the Dutch consumer electronics firm, Philips, maintains that while the frost goal of the Framework Program is the integration of Europe, the second goal (especially of the telematics programs) is to break the domination of AT&T and IBM. 145 This statement is no doubt an exaggeration, but if there is any truth in it, uneasiness about even limited participation on the part of IBM and AT&T could be expected. Nevertheless, it could be self-defeating to exclude IBM from projects such as JESSI, which is supposed to foster tighter links between supplier and customer fins. As one IBM official put it: We have to be in JESSI; all the key players--our suppliers and usersare there. IBM is the largest buyer and supplier of chips in Europe. The same goes for IBMs participation in RACE, Prometheus (a EUREKA project that links modern information and automotive technologies to improve transportation efficiency and safety), and other projects: 3 years ago, IBM was involved in only 3 projects; now it is in 20. 146 Still more problematic is the question of Japanese participation, for Japanese firms are a greater competitive threat than most American firms, especially in information technology. Two things concern Europeans facing dependency on Japanese suppliers. For one, the big Japanese firms not only supply components, they also compete directly with systems integrators. Depending on a competitor for key supplies is risky at best. In addition, many fear that the Japanese firms will act nationalistically, or in keeping with long-established intercompany ties, serving the needs of their Japanese customers before foreign customers. Until 1990, the issue of Japanese participation in the Framework Program was avoidable. Then Fujitsu, a longstanding supplier of products (both components and systems) to the British computer firm ICL, bought 80 percent of ICL. ICL was one of the original Big Twelve information technology companies in Europe and a participant in many ESPRIT projects. Although ICL met and still meets the ECs criteria for participation in the Framework Program-an integrated presence in Europe, including R&D, sales, and manufacturing 147 the prospect of one of the Japanese giants having immediate access to and decisionmaking power in ESPRIT projects is forcing a reevaluation. As of spring 1991, ICL had been removed from the working committee that manages JESSI and was invited out of three of five projects. So far, beyond the Fujitsu-ICL imbroglio, Japanese firms have not participated in Framework projects. Whether this is because they have been uninterested, reached a polite understanding that they were unwelcome, or been formally deterred, is unknown. l48 But the participation of Japanese companies is likely to be a bigger issue in the future, for they are beginning to knock on the Commissions door, as one participant put it, to ask for admission to the programs. 149
PAGE 227
Chapter MC-92: Trade and Industry Policy .223 Sleeping Partners The Framework Program is required to contribute to the nations that fund it, though not in exact proportion to the funding. The contributions from the large economies--Germany, France, Great Britain, and Italy-are much greater than those of the smaller countries, particularly ones like Portugal and Greece that lack well-developed high-technology sectors. While every project need not scrupulously include participation from every member nation, and participation does not have to match exactly relative funding levels, there is some pressure to include participants from tie smaller and weaker economies. Sometimes this is a minor problem. A few representatives from strong system companies mention that a drawback of Framework projects is that they were pressed to include unwanted organizations that contribute little to the project; these are variously termed sleeping partners or alibi partners. In response to questionnaires, participants in BRITE, RACE, and ESPRIT have all indicated that they are generally satisfied with the alliances formed within the programs, but not universally. Some of the disenchantment is undoubtedly due to the necessary coupling of strong and weak enterprises; this factor is also sometimes viewed as a weakness in the American high-technology consortium, Microelectronics and Computer Consortium. However, the inconvenience may be balanced, in part, by the benefits gained by weaker enterprises and nations. 150 Performance of Framework and EUREKA Programs Europeans are excited about the Framework Program and EUREKA. Most are happy with the significant increase in transborder cooperation, new linkages forged between companies and academic institutions, and alliances between small and large enterprises. The programs have also contributed to standard-setting in Europe. They claim to be eliminating wasteful duplication within national programs, as well as fragmented and unconnected national programs themselves. And they have produced some commercially exploitable results. What the technology programs have not yet done is make much of a dent in the basic problem: Europes slipping competitiveness in high technology. It may be too soon for the programs to have made such a difference; the oldest of the programs have been in operation for less than a decade. Moreover, Framework and EUREKA are only part of Europes new economy, and many of the more sweeping changes planned for the single market have not yet been implemented. AccomplishmentsIn some ways, the programs have much to show for the short time they have been in operation, most notably an increase in cooperation. International cooperative agreements were on the upswing before Framework and EUREKA got into gear, but observers inside and outside Europe agree that the programs have increased European firms and research organizations knowledge of each other, helped to arrange collaborations that would have been difficult before, and set the stage for future alliances. They have helped producers and users to hook up, and sometimes have become indispensable for that purpose. There is still room for improvement; RACE and ESPRIT have been criticized for not including enough user input, and concentrating too much on the agenda of the Big Twelve. But for the most part, policymakers, participants, and Europe in general seem pleased with the progress in linking up users and suppliers. An often-cited contribution of the technology programs is in standard-setting. The programs help in two ways: greater communication among companies working in the same business helps to build consensus on harmonizing varying standards; and the cooperation of the 12 nations helps to make Europe more of a force in world standard-setting. All EUREKA communications projects at least touch on aspects of standardization, and five aim directly at developing standards, including, for example, a project to develop high-definition television production and broadcasting standards and equipment for producing and receiving next generation television broadcasts. 151 RACE is particularly interested in developing standards for integrated broadband communications (IBC) across the EC by 1995, which in turn will require common standards for telecommunications equipment and service. These and other such projects may lead to EC-wide standards and possibly to worldwide standards. That is both a significant accomplishment of European technology programs and a source of international friction. Foreign companies fear that by being excluded or accorded second-class status (e.g., being left out of the critical process of making decisions) in technology programs they will be at a
PAGE 228
224 Competing Economies: America, Europe, and the Pacific Rim competitive disadvantage with regard to standards, for it is unlikely that their own product standards would be considered for the European Community 152 The Framework Program has also helped to solve one of Europes particular problems: knitting together 12 separate, sovereign nations. The EC does not intend to become one unified economy and political entity-a United States of Europe-but one major aim of unification is to reduce the fragmentation and division that characterize the 12 members now. The Framework program, along with the 279 directives of the Single Market Act, will help to lower technical barriers among these nations, and to reduce the fragmentation caused by the traditions of different European institutions. If the evaluations of Framework programs are any indication, 153 Euro peans are confident that the Framework programs have, in the words of the ESPRIT evaluators, influenced several thousand of scientists and engineers in information technology fields to think European and to do so in a positive way. Another contribution of the programs is helping to avoid duplication of effort. This is a benefit claimed for most R&D consortia, but in Europes case may be well-founded because of prior national support for duplicative efforts. Several participants mentioned the efficiencies of learning from others. Officials of Siemens, speaking of the companys participation in the Eurolaser project, indicated that one of the benefits was the ability to go to the labs of another company and possibly do investigations there. This, according to Siemens, helps a company to do its technical work more economically. 155 Duplication, or redundancy, is another issue that cannot be neatly resolved. The benefits of avoiding wasteful duplication are obvious, but there is also a danger of eliminating all overlap. Von Hipple points out that the best climate for innovation is one in which many organizations are involved in independent efforts to solve the same problem. There is perhaps a danger that too much cooperation could focus innovation efforts too narrowly. However, that danger probably is minuscule for Europe now, especially in view of the fact that some EUREKA projects are similar to but independent of EC programs. 156 Competitiveness-The Framework program and EUREKA have not yet made a big difference in the competitiveness of European industry. Whether they are on track to do so is an open question. EC evaluations tend to avoid the issue of competitiveness. This is understandable; it is difficult to measure. More importantly, the EC seems determined to give the programs plenty of time to worka necessary part of competitiveness programs, particularly in programs designed to catch up with the leaders. Still, some attempts have been made to assess the programs in terms of market outcomes. A few ESPRIT projects have already yielded commercial results, or are on the verge of doing so. For example, ASM Lithography B.V., a Dutch company, has the lead role in an ESPRIT project to develop a deep ultraviolet wafer stepper for semiconductor production. It recently announced that its deep UV stepper was being tested in several locations throughout the world. The stepper is capable of etching lines less than 0.3 microns wide, and if the tests go as expected the company stands to gain significant market share. The Japanese companies Canon and Nikon already have deep W wafer steppers, but at the end of 1990 they were not yet available outside Japan. 157 Another ESPRIT project that helped produce a commercial product was the Supernode project. 158 It ran 3 years, and produced a modular, reconfigurable computer than can link a few to many processors together. Since the end of Supemode in 1988, both Telemat of France and Parsys of the United Kingdom have made and marketed Supernode computers, and by April of 1990 they had sold more than 200 of the machines. 159 Stories of commercial exploitation of products resulting from Framework or EUREKA projects are beginning to appear, but are not yet plentiful. Participants in the first phase of ESPRIT reported that 143 of the 227 projects resulted in industrially significant results (table 5-11), but only about 12 percent of the projects of ESPRIT I resulted in a commercial product by the time of the evaluation. Concern is surfacing over the issue; a recent news item reported that several influential officials of the CEC criticized ESPRIT for producing little commercially useful technology, and were concerned that European electronics companies were still overly dependent on EC subsidies and trade protection. 160
PAGE 229
Chapter 5--EC-92: Trade and Industry Policy .225 Table 5-1 lindustrially Significant ESPRIT I Results Contributed directly to products and services available on the market . . . . . . . . . 27 Contributed directly to products and services developed for the market but not yet commercially available .......44 Used outside the ESPRIT project, either within the company concerned or in another company . . . 44 Contributed to standardization, either adopted as an international standard or being elaborated by an international standards organization . . . . 28 SOURCE: J.F. Blackburn, London Representative of the Commerce Department for Industrial Assessment in Computer Sciences and Telecommunications, Overview of European Strategic Technology in Computers, Telecommunications, and Electronics, ESN Information Bulletin, ESNIB 90-40, p. 12. Results of BRITE are more promising. Fifty-nine percent of BRITE participants reported that they expected commercial benefits from their projects within 5 years. The evaluation team regarded this expectation as somewhat optimistic, but found it encouraging. 161 JESSI got off to a slow start-according to one participant, work only began in 1990 after an 18-month startup phase 162 --and Philips recently announced it was pulling out of the JESSI SRAM development project. Another indication of trouble is that funding from the national governments has not materialized, and the EC is far behind schedule in its funding. However, several of the projects participants, as well as the ESPRIT evaluation team, indicated that its strategic direction is sound. Another comment on its significance is the news that JESSI and Sematech just initiated their first cooperative venture (others are expected). 163 Overall, the assessment of commercial output from the Framework Program and EUREKA projects is mixed. Even if they come up with commercial products or services, that is not the same as competitive success, which means raising market share while maintaining or increasing standards of living. That, in turn, means helping to eliminate Europes trade deficit in high-technology products. On these points, the evidence is clearer. So far, the projects have not made much difference. For example, in 1975, the EC had a trade surplus in information technology products, but by the beginning of ESPRIT I, it was becoming steadily more dependent on imports. The trade deficit in IT products reached $5 billion by 1980, and increased to nearly $22 billion by 1987 (figure 5-8). The largest deficits were in such products as integrated circuits, consumer electronics, and computers. That year, the EC accounted for 28 percent of world consumption of computer systems but supplied only 13 percent, and consumed 17 percent of the worlds microelectronics while producing 11 percent. The Commission praised European computer companies for giving good service and keeping upmarket share, but also noted that some European vendors have kept their turnover up by adopting the role of system integrator or product distributor which results in a reduction in value added. l @ For example, Siemens supercomputers are made for it by Fujitsu and its laptops are made by Matsushita. Another problem is getting off the dole. It is hard to call a company or an industry truly successful if it needs continuing subsidies or protection to survive. One reason why Japanese electronics companies are judged successful is that they continue to flourish while formal and informal trade protection is disappearing (albeit slowly and sometimes grudgingly). Few European companies have managed this. The Airbus consortium, for example, is widely cited in Europe as a successful venture, yet it has received over $5 billion in success-dependent loans, which have yet to be repaid, and continues to receive substantial development loans from the governments of Germany, France, England, and Spain. Another example is a company called ES2 (European Silicon Structures). In 1985, the company was formed as a joint venture of major electronics firms from nine European countries to produce ASICs. In Figure 5-8Europes Trade Deficit in Electronics in 1987 (21.9bn) Professional electronics measurements Instrumentation Medical electronics Telecommunications Passive components Software Office systems Industrial automation Active components Consumer electronics Computers -1412-10-8 -6 -4 -2 0 2 4 $ billions SOURCE: Information from the Commission of the European Communities, Communication From the Commission to the Council and fhe Parliament Concerning the Performance and Results of the Programrne ESPR/T, SEC(89) 1348 final (Brussels, Belgium: Sept. 7, 1969), p. 6, who cite EIC as their source.
PAGE 230
226 l Competing Economies: America, Europe, and the Pacific Rim 1989, it posted revenues of $19.3 million, 45 percent higher than in the previous year. No other company of the same size participated in so many European technology programs; it was a player in ESPRIT and JESSI. A success? The company had not yet managed to break even, much less turn a profit, by the end of 1989, and its revenues were considerably below the prediction of its founders, who announced in 1985 that its sales would break $100 million in 5 years. 165 Even the evaluations, which are generally positive in tone and outlook, sound a few warnin gs. For example, ESPRIT evaluators noted that too much of the technological advance was in niche areas with limited market potential; that work in the software technology portion of the program had produced valuable inhouse developments but had yet to result in commercial exploitation; and the work in the office systems portion of the program were disappointing overall. The major success of the program was the increase in trans-European cooperation made possible by EC funding, but even that was judged possibly not robust enough to survive substantial funding reduction. 166 European policymakers and many experts point out that it is too soon to bring the jury back on Framework and EUREKA, and they are right. The programs have achieved some notable results, and it is laudable that those in charge have resisted the temptation to make snap judgments. But at this point they have little to teach others about new ways to improve competitiveness. EC-92: COMPETITION POLICY AND EFFICIENT FIRMS EC-92s removal of trade, financial, and technology barriers within Europe will create big enough markets to satisfy the needs of any European industry that may have been handicapped by a national market too small to allow economies of scale. The Commission wants to encourage the formation of firms large enough to take advantage of the bigger market, but at the same time to foster competition. EC competition policy has two distinct aims. One, similar to the U.S. antitrust tradition, is oriented toward the regulation of restrictive practices and cartels in the private sector. The other is aimed at removing member countries favoritism toward their national companies in a wide range of industries and sectors, including telecommunications, passenger air transportation, energy, autos, chemicals, and an array of service industries. The goal of these policies is to help create stronger, more competitive, possibly pan-European firms. European Antitrust Policies Competition policy governs mergers, acquisitions, and formal linkages such as joint ventures (merger policy) and corporate practices that limit competition in a market (restrictive practices). Until recently, antitrust policies in the EC have been run at the national level. In general, European countries are less sympathetic to antitrust actions than the United States. Some, like France and Germany, evaluate mergers and restrictive practices case by case, and allow firms a great deal of market power when they believe that consumers or the national interest gain thereby. France and Italy have also used antitrust law to favor national champions involved in mergers and takeovers. Britain, while in principle opposing restrictive practices by companies, is more relaxed in practice than the United States. Britain, France, and Germany all use balance of benefits arguments to determine whether business arrangements are appropriate, and their responses to unacceptable practices rarely involve structural change in the market. In 1990, the EC took control of very large mergers in Europe. EC policy will probably remove the controls imposed by national governments, keep a close eye on firms restrictive practices that tight distort competition among European countries, and at the same time encourage the creation of larger firms, including cross-border mergers. 167 Many EC officials believe that European firms in many industries are simply too small to compete effectively, and that a new single market can offer economies of scale that larger companies can exploit. However, some voices in the EC have warned against a simple view that bigger is better. In particular, Sir Leon Brittan, the EC Commissioner for competition policy, has argued that she alone does not create world beaters, and that the EC needs more competition in home markets. 168 As table 5-12 shows, European manufacturing firms are somewhat smaller than U.S. firms. They are closer in size to Japanese firms, but this maybe deceptive since most leading Japanese manufacturing companies belong to industry groups (keiretsu) that in effect expand their size.
PAGE 231
Chapter 5--EC-92: Trade and Industry Policy .227 Table 5-12-Shares of Manufacturing Industry Held by Large Firms in Japan, the United States, and the EC, 1986 Share of largest firms in total sales of industry Largest 5 10 20 40 (percent) EC . . . . . . . 6.8 10.8 16.9 23.0 United States . . . . . 13.5 19.6 26.5 34.6 Japan . . . . . . . 8.0 12.6 17.5 23.7 SOURCE: EC Commission, Horizontal Mergers and Competition Policy in the EC, European Economy, No. 40, May 1989, table 5-1. The ECs mergers directive (put in effect Sept. 21, 1990) layers an EC merger-control office on top of the national offices, to deal with mega-mergers with a substantial European dimension. The EC office now has the power to veto all mergers, and those joint ventures in which one company is established in the EC, with worldwide revenues of at least ECU 5 billion and EC revenues of at least ECU250 million. 169 (However, the directive allows national merger authorities to intervene when they believe that a legitimate interest is at stake. 170 ) The Commission can compel firms to disclose intentions, as well as block mergers and fine violators. 171 It also has extraterritorial powers over mergers between two foreign-based firms doing business in the EC: competition rules govern all business arrangements that operate witiin the EC, even firms that have no registered home office there. In principle, the EC will approve business arrangements (including both mergers and interfirm links) when they do not impede competition within the Community. The directive offers a safe haven for mergers that will not result in one firm controlling 25 percent or more of a relevant market; there are also block exemptions, covering certain classes of business activity; and firms can seek a letter of comfort from the EC giving tentative approval to a proposed arrangement in advance. All restrictive practices are illegal unless they in some way benefit the interests of the Community. The directive allows the Commission to take into account social and economic factors when considering mergers. 172 It is not yet clear exactly how the Commi ssion will interpret this mandate, but an indication comes from the Commissions decision not to intervene in the merger between Siemens and Nixdorf, on the grounds that competition from other EC and non-EC firms will provide effective competition despite the increase in concentration. 173 A n approach consistent with this decision was earlier outlined in a paper by three senior officials in DG-11. 174 The authors analyzed mergers against two kinds of measures: 1. 2. the danger of reduced competition, in which the key variables are the degree of market power concentration, the speed with which demand is growing (making entry more or less attractive), and the extent of import penetration; and the potential for efficiency gains, focusing on the technology content of the industry and the existence of economies of scale. These criteria suggest how the Commission may try to mesh the aims of encouraging mergers into larger firms with traditional antitrust policy. Such an EC policy would bless and possibly accelerate a process already under way. Merger activity in the EC has taken off as firms jockey for position in the new single market (figure 5-9). The number of significant deals more than tripled between 1982 and 1987, 175 and by 1989 cross-border mergers rose to 1,267 deals with an estimated value of ECU60 billion. 176 Some national policies have also changed. French restrictions on foreign ownership appeared to be declining, as France was the second largest target country in late 1989 and the largest in the first quarter of 1990. 177 The Commission has specifically encouraged much of this activity, as the assertion of market forces over government regulation. 178 Figure 5-9-Total Merger Activity, All National, Community, and International Mergers by Combined Takeover Number of mergers 500 [ 1 400 } 1987 1988 1989 = >1,000 MECU = >2,000 MECU m >5,000 MECU = >10,000 MECU SOURCE: EC Competition Reports [July].
PAGE 232
228 l Competing Economies: America, Europe, and the Pacific Rim Though it seems to be accepted wisdom among some in the EC that bigger firms mean better competitive performance in general, that belief is open to question. In some industries, to be sure, large size seems a distinct advantage. For example, the competitiveness of Japans giant electronics firms is often given as a leading reason for their great success in the 1980s. But in other industries, the advantages of very large size are not so overwhelming. A famous misstep of Japans illustrious MITI was its abortive attempt in the 1960s to reduce the auto industry to three firms, under the conviction that the industry could not support any more. The industry resisted, MITI abandoned the attempt, and there are still 11 Japanese auto producers, most of them doing very well. Likewise, MITI tried to rationalize the machine tool industry, and got no further. Today, Japanese machine tool builders, which are still very numerous, lead the world. Obviously, many factors contribute to success in manufacturing and in some industries size is one of them, but even in those, it may be quite far down the list in importance. Subsidies and Nationalized Companies Peter Sutherland, the former Commissioner in charge of competition, put the Commissions position on favoritism to national companies in these terms: If you remove national trade barriers, and if you also leave the capacity of governments to interfere with trade by supporting unfairly one industry against another, to allow companies to carve up markets and fix distribution, you would have the same effect of destroying 1992 as if the barriers were left untouched. 179 The question is not only one of fairness and removal of intra-EC barriers. Protection and pampering can also sap firms competitiveness. Subsidies for both national champions and lame ducks are a prominent feature of the economic landscape in Europe. Moreover, nationalized industries have always had special relations with governments and their treasuries. Both are under challenge from the EC. Subsidies--Subsidies have long been a central component of European industrial policies. 180 E C member states provide much greater industrial subsidies (known as state aids) than either Japan or the United States. In the EC, state aids average around 3 percent of GDP, compared to 1 percent in Japan and 0.5 percent in the United States. 181 Table 5-13 shows the major subsidies provided by the largest EC member states. There are many different, sometimes subtle, ways in which governments can help out industries. Simple cash grants are one possibility. Tax relief can be specially tailored. Debt can be written off. Different countries tend to provide subsidies in different ways: Germany uses tax concessions (58 percent of total aid); France provides 26 percent through equity and less than 20 percent through tax concessions; France and the Netherlands provide subsidized energy. Most national subsidie |