This paper reviews major issues and developments in the trade area and outlines the challenges governments face as they seek to liberalize trade in the Uruguay Round of trade negotiations and address new trade issues. In industrial countries, the reorientation of policies was most apparent in steps taken to liberalize financial markets and foreign direct investment, privatize public enterprises, and deregulate services, particularly in the transportation and communication sectors. Among developing countries, a growing number recognized the merits of outward, market-oriented policies and took steps to liberalize their trade regimes and open their economies to international competition. By and large, the increased focus on market principles in industrial countries did not carry over to trade and industrial policies or, most notable, to the agricultural sector. Despite strong growth performance in 1983–1989, little progress was made in rolling back the protective barriers that had risen during the preceding recessionary period; protection persists in agriculture and declining sectors and has spread to newer high-tech areas.


International Monetary Fund, The Rise in Protectionism, IMF Pamphlet Series, No. 24 (Washington, 1978); Trade Policy Developments in Industrial Countries, IMF Occasional Paper, No. 5 (Washington, July 1981); Developments in International Trade Policy, IMF Occasional Paper, No. 16 (Washington, second printing, 1983); Trade Policy Issues and Developments, IMF Occasional Paper, No. 38 (Washington, July 1985). See also Selected References at the end of the paper; and Issues and Developments in International Trade Policy, IMF Occasional Paper, No. 63 (Washington, December 1988).


See M. Kelly and others (1988). In this report, the definition of industrial countries, major industrial countries, and Eastern European countries conforms to that used in International Monetary Fund (1991a).


United Nations Conference on Trade and Development (UNCTAD) (1991), and UNCTAD Data Base on Trade Control Measures.


Sweden unilaterally removed import restrictions on textiles at the end of July, making it the only participating country that did not adhere to the extension of the MFA.


This group comprises Hong Kong, Korea, Malaysia, Singapore, Taiwan Province of China, and Thailand.


The liberalization of steel is proceeding in multilateral negotiations under the U.S. Steel Liberalization Program, which aims to eliminate tariff and nontariff barriers and prohibit most subsidies; an agreement that would phase out subsidies to shipbuilding and repair is being negotiated in the OECD Council Working Party on Shipbuilding. “Gray-area” measures, such as voluntary export restraints, are those taken outside GATT surveillance and whose consistency with GATT rules has not been tested.


For a comprehensive review of structural policies in industrial countries see International Monetary Fund (1989).




U.S. agricultural policy is established in major farm bills, which are usually legislated every five years.


The Cairns Group is a coalition of industrial and developing country agricultural exporters. It comprises Australia, Argentina, Brazil, Canada, Chile, Colombia, Fiji, Hungary, Indonesia, Malaysia, New Zealand, the Philippines, Thailand, and Uruguay. In contrast to other members of the Cairns Group, Canada is also a major importer of agricultural products and has sought to strengthen the GATT provision (Article XI.2.c.) covering supply management programs, which Canada maintains for dairy and poultry products.


See Australian Industry Commission (1990) for an assessment of pro-active industrial policies in Japan, Korea, Taiwan Province of China, Hong Kong, and Singapore.


The recent agreement reached between the EC and Japan to convert national VERs on Japan’s automobile exports to EC countries to an EC-wide monitoring arrangement, which will limit Japan’s exports of automobiles to their current EC-wide level through the year 1999 and which includes projected increases in sales by “transplants,” is a market-sharing arrangement of dubious legality under EC competition laws. While Japan and the EC have agreed to defend this arrangement in GATT, it is similar to other gray-area measures whose consistency with the GATT principle of nondiscrimination is questionable. The Community has announced that it will not request any such cooperative measures after the end of 1999.


GATT (1990a), Vol. I, pp. 40–42 and 55.


Bhagwati (1991) refers to the “spider’s web” phenomenon. Appendix II examines the implications of the globalization of investment and production for existing trade rules. Appendix I on the GATT system discusses recent disputes concerning FDI and local content rules.


Hong Kong, Korea, Singapore, and Taiwan Province of China.


During the 1980s, New Zealand and Australia initiated tariff reform programs that have reduced tariff rates below those indicated in Table 3. Several other countries have reduced tariffs on selected items, including, Canada, Japan, and the United States.


These comprise the EC, EFTA, the U.S.-Canada Free Trade Agreement, and the Australia-New Zealand Closer Economic Relations Trade Agreement.


A trade coverage ratio measures the percent of trade in an item subject to nontariff measures in the total trade in that item. It has important limitations as an indicator of restrictiveness because it is not known what the levels of trade would be in the absence of restrictions. Also, the trade coverage ratio does not capture the effects of a tightening or loosening of restrictions; it only reflects changes in product coverage. Only border measures are covered in the UNCTAD data base.


Restrictive measures affect not only products currently exported, but may also deter investment in areas where developing countries have export potential.


“Core” nontariff measures exclude antidumping and countervailing actions, automatic licensing, import surveillance measures, and para-tariff measures.


In July 1991, the MFA was extended without change through 1992 because of the delay in concluding the Uruguay Round (Appendix I).


See Finger and Nogues (1987), Messerlin (1989), and Nicolaides (1991) for reviews of antidumping and countervailing practices in the United States and the EC. In particular, Messerlin argues that features of the codes governing antidumping and countervailing actions and their application in practice make these two instruments substitutable to the import-competing firm. He finds that cases initiated by the EC and the United States target a narrow range of industries and have a strong sector-specific impact.


According to GATT (1991a, p. 8), these four countries initiated over 1,000 investigations since 1980, 500 of which resulted in protective measures.


Office of the United States Trade Representative (1991).


Circumvention may occur when the location of certain production processes, such as final assembly, changes as a result of antidumping or countervailing duties that have been threatened or applied on a final product.


Office of the United States Trade Representative (1991).


The OECD’s national accounts data define subsidies as all grants on current account by government to private industries and public corporations, plus grants to government enterprises for operating losses owing to government policies to maintain prices below the costs of production.


The surveys exclude important sectors, most notably, energy (except coal), transport (except railways and inland waterways), and housing. Aid granted by the European Community is not covered; about 80 percent of this goes to the agricultural sector.


The EC(10) countries are Belgium, Denmark, Germany, Greece, France, Ireland, Italy, Luxembourg, the Netherlands, and the United Kingdom.


OECD (1990a), p. 111. These figures measure only government-funded research and development in the industrial sector. Research and development spending by governments ranges from about 20 percent of total research and development spending in Japan to about 50 percent in the other major industrial countries.


The previous agreement was concluded in 1973 and revised in 1983.


In the GATT, trade liberalization proceeds by the reciprocal exchange of reductions in trade barriers based on a rough balance of concessions. By contrast, “fair” trade increasingly demands “strict” reciprocity, or equal market access product by product.


Japan agreed to increase public investment in infrastructure, reform land use policies and the distribution system, eliminate exclusionary business practices and make keiretsu relations more transparent, and liberalize the foreign investment regime. The United States agreed to reduce the federal budget deficit and the Government’s exposure to risks from federal credit programs, promote private saving, improve U.S. competitiveness through greater investment in research and development and workforce education and training, and lower the cost of capital.


Recent reviews of the evidence are provided in Ostrom (1991), Bhagwati (1991), and Balassa and Noland (1988). A number of researchers (Bergsten and Cline (1987), Saxonhouse (1983) (1985)) have found that the level and composition of Japan’s imports can be explained by its factor endowments. Others (Balassa (1986), Lawrence (1987), and Lincoln (1990)) find evidence that Japan’s imports are “too low” based on its characteristics or that nontrade barriers have affected the level of imports.


GATT (1990d), p. 200. MITI’s figures do not include exports to the EC of passenger cars, video-cassette recorders (VCRs), and color televisions, which are subject to monitoring measures.


The countries covered in this section include Bulgaria, the Czech and Slovak Federal Republic (Czechoslovakia), Hungary, Poland, Romania, and Yugoslavia. Yugoslavia differs from the other Eastern European countries in several important respects: in the 1950s, it abandoned central economic management; its firms were self-managed by workers; its trade and price regimes were largely liberalized; and it never joined the Council for Mutual Economic Assistance (CMEA). Its trade policies are covered in this section because it shared with other Eastern European countries some of the essential features of the socialist economic system, including public ownership of the means of production, and it has recently accelerated its economic reforms.


The CMEA was founded by Bulgaria, Romania, Czechoslovakia, Poland, Hungary, and the former U.S.S.R. in 1949. The former German Democratic Republic joined in 1950. Mongolia, Cuba, and Viet Nam joined in later years. Albania was a temporary member; Yugoslavia, excluded from membership in 1949, became an associate member in 1965.


Prices were denominated in “transferable rubles” (TR). The Bucharest formula for pricing goods traded among CMEA members was, in principle, a five-year moving average of comparable world market prices. This formula was easily applied to trade in primary products, such as crude oil and natural gas. However, it was more difficult to use for trade in manufactures, since the quality of the goods made in the CMEA region was generally inferior to that of the Western goods used as the basis for the price comparisons. As a result, the five-year average prices served as a basis for negotiations.


For example, in 1989 exports to the former U.S.S.R. as a share of total exports were 65.8 percent for Bulgaria, 30.5 percent for Czechoslovakia, 25.1 percent for Hungary, 20.5 percent for Poland, 21.4 percent for Romania, and 20.9 percent for Yugoslavia (International Monetary Fund (1991a), p. 29, Table 7). Also International Monetary Fund (1990b), p. 65, Table 18.


Socialist countries consisted of CMEA countries, Albania, China, Cambodia, the Democratic People’s Republic of Korea, the Lao People’s Democratic Republic, and Yugoslavia.


Earlier reform efforts are discussed in International Monetary Fund (1991a), Williamson (1991), Gelb and Gray (1991), and, for the Hungarian case, Boote and Somogyi (1987).


See International Monetary Fund (1991a) and Starrels (1991) for a discussion of the main features of economic reforms implemented during the past two years.


The concept of currency convertibility as defined by the International Monetary Fund refers to the unrestricted and nondiscriminatory availability and use of domestic currency to effect payments and transfers for current international transactions.


In Poland, however, households are allowed to acquire foreign exchange in the parallel market at the prevailing rate for transactions for which foreign exchange is not available through the official market.


In Yugoslavia, the state monopoly on foreign trade was removed in 1950.


Licensing was reintroduced in June 1991 for cattle, beef, and butter.


Average tariff rates were about 5 percent in Czechoslovakia (1990), 16 percent in Hungary (1991), about 14 percent in Poland (1991), and about 7 percent in Yugoslavia (1990). In Bulgaria, average MFN tariff rates are 5.9 percent on intermediate goods, 7.4 percent on capital goods, 13.3 percent on consumer goods, and 7.9 percent on agricultural products.


This was applied to all imports except those from countries with which Czechoslovakia has free trade agreements (e.g., Finland). Czechoslovakia notified this surtax to the GATT Balance of Payments Committee because it increased some tariffs above their “bound” levels.


In June 1991, Bulgaria reduced the number of export bans to six items and eliminated all export taxes.


This point was made, for example, by the Hungarian representative to the GATT Council at the time of the review of Hungary’s trade policies under GATT’s Trade Policy Review Mechanism (GATT Newsletter, May/June 1991).


In particular, before the reforms, the prices of energy and raw materials were held artificially low, while quantitative restrictions kept out foreign produced manufactured goods. This resulted in high rates of effective protection for finished manufactured goods and negative rates of protection for energy and raw materials.


McKinnon (1991a). However, sufficient restraint on wages, a significant input for production, would probably address the negative value-added problem.


Increases that would raise tariffs above “bound” levels in GATT would need to be negotiated with other Contracting Parties. Although there is no explicit legal precedent for accepting higher tariffs in exchange for a reduction in quantitative restrictions, Article XXVIII of GATT provides the legal framework for renegotiating tariff bindings.


Some suggest that tariffs are a better option in the short run because (i) tariffs raise revenue and thus improve the country’s fiscal position, whereas devaluation has that effect only if the government owns the export industries; (ii) devaluation pushes up the prices of all goods, including intermediate goods and necessities, whereas tariffs can be concentrated on less essential goods; and (iii) the prospect of future declines in the prices of durable consumer goods as the tariffs are removed will create an intertemporal substitution effect in favor of delaying consumption (see Williamson (1991)).


This issue is discussed in Lane and Dinopoulos (1991), who conclude that some combination of debt reduction and tax reform may facilitate market-oriented reforms in socialist economies.


See Michalopoulos and Tarr (1991) and Kenen (1991) for detailed discussions.


The combined trade of Eastern Europe and the former U.S.S.R. is estimated at less than 10 percent of total world trade (see Collins and Rodrik (1991)).


Center for Economic Policy Research (CEPR) (1990), Collins and Rodrik (1991), and Havrylyshyn and Pritchett (1991). For example, Havrylyshyn and Pritchett show, using a gravity model, that the bilateral trade pattern in Eastern Europe is markedly different from that expected for countries with their income level, location, and size because it was determined less by market forces and more by political and economic relationships binding Eastern Europe and the former U.S.S.R. together in the CMEA. They find that a natural pattern of trade would be almost exactly the reverse of the current pattern where actual flows of trade for Eastern Europe are about 60–80 percent within the East Bloc; and 20–30 percent with Northern Europe. Collins and Rodrik also conclude that the natural trade partners of Eastern Europe are the advanced industrial countries of Europe, rather than each other. Some scope for trade expansion may also exist between Eastern Europe and the dynamic Asian economies. Trade between the two regions has expanded rapidly during the last several years. For example, the value of the Republic of Korea’s exports to Eastern Europe has increased by about five times and imports have increased by about 3.5 times during 1988–90. Major export items by the dynamic Asian economies to Eastern Europe include electronics and other consumer goods, while major export items by Eastern Europe to those economies include machinery, chemical, and steel products. Total amounts of Eastern European trade with dynamic Asian economies, however, remain small relative to trade with industrial markets.


The decline of non-CMEA exports was also significant for these countries, except for Hungary and Poland where non-CMEA exports increased significantly.


In the case of Romania, after the Ceausescu Government renounced MFN privileges in February 1988, the United States suspended the MFN provision under the 1974 Trade Agreement.


Yugoslavia has received GSP treatment from the United States since 1976.


The United States plans to renegotiate bilateral agreements on textile quotas with Hungary, Poland, and Czechoslovakia by the end of 1991. It also plans to increase cheese quotas in the context of the Uruguay Round negotiations (White House, July 12, 1991).


The EC has had a cooperation trade agreement with Yugoslavia since 1980; it provides for financial and technical aid and includes trade preferences for Yugoslavia’s exports.


Bulgaria and Romania have also requested the EC to undertake preliminary discussions on an Association Agreement.


According to one estimate (Center for Economic Policy Research (1990)) a world market price reduction of at least 10 percent for wheat and 9 percent for coarse grains would be necessary to absorb a 20 percent increase in the production of these commodities in Eastern Europe and the former U.S.S.R. from their 1988 levels.


A September 30, 1991 proposal by EC foreign ministers would allow a 10 percent annual increase in beef and mutton imports from the three Eastern European countries for five years. (Current shipments from the three countries are around 28,000 metric tons a year.) Under a “triangular trade” arrangement, the EC would help finance food exports from the three Eastern European countries to the former U.S.S.R.; such exports would be deduced from the increased EC quota (International Trade Reporter (1991), p. 1425).


Steel and coal are covered by the European Coal and Steel Community (ECSC) treaty signed in 1951. The EC has executive authority for the ECSC, but its powers under the ECSC treaty are greater than under the EEC treaty; these powers have been used to control production and prices of steel within the EC.


Steel industry experts in the EC and independent studies suggest that production in the whole of Eastern Europe, including the former U.S.S.R., needs to be cut by 50 percent to achieve the aim of a competitive subsidy-free industry (European Report (1991)).


For example, during 1985–89 these three sectors represented 28 percent of Czechoslovakia’s exports to the EC and 36 percent of its exports to the United States; for Hungary, the corresponding figures were 39 percent and 27 percent; and for Poland, 30 percent and 65 percent.


The United States, however, suspended the application of the General Agreement to Czechoslovakia in 1951. In November 1990, it waived the suspension of MFN treatment (see above).


See Section IV for a separate discussion of Eastern European countries.


Quantitative restrictions are defined to include both trade and trade-related exchange restrictions and tariffs to include all import charges.


Krueger (1984) stresses the relationship between inappropriate macroeconomic policies and the extensive use of restrictions and controls in developing countries in the period up to the early 1980s. Corden (1987) analyzes the Cambridge argument of real wage rigidity to justify the use of controls rather than a devaluation to correct an external imbalance. He notes, however, that there is no particular reason to make this assumption.


During this period, India was one of the few developing countries to maintain relatively restrained macroeconomic policies while using extensive controls to underpin its strongly inward-oriented development strategy. However, the pervasiveness of controls was also reflected in a more appreciated exchange rate than would have otherwise been the case.


See Section VI for a review of agricultural trade policies. In studies reported therein, it is estimated that in the first half of the 1980s direct and indirect government intervention in the agricultural and industrial sectors resulted in substantially negative effective rates of assistance to agricultural export crops and domestic food crops.


Comprehensive trade reform programs are those classified as either major or moderate reform programs.


See also Michaely (1991) for a review of trade reform programs from the 1950s up to 1984.


Among the industrial countries, trade reforms in Australia and New Zealand were also hampered by an inappropriate fiscal-monetary policy mix.


Korea liberalized its trade regime progressively during the 1970s and early 1980s, but its major long-term import liberalization programs began in 1984. By 1984, Korea had largely adjusted to the major oil price shocks of the late 1970s and early 1980s.


The complex linkages between trade liberalization and fiscal policy are discussed in more depth in other papers (see, for example, Farhadian-Lorie and Katz (1989) and Blejer (1990)).


See, for example, Michaely (1991), Krueger (1978), and Cline (1981).


The producer subsidy equivalent of a support policy is the estimated transfer needed to replace a support policy and leave the producer no worse off. It can be expressed as a total value of the transfer or as a percentage of the value of production at internal prices.


Agricultural support policies in the United States are set every five years in the “Farm Bills.” The full title of the 1985 Farm Bill was the Food Security Act of 1985, and the term 1990 Farm Bill covers the Food, Agriculture, Conservation, and Trade Act of 1990, enacted on November 28, 1990, and the Omnibus Budget Reconciliation Act of 1990, enacted on November 5, 1990.


See United States, Department of Agriculture (1991). One reason for this result is the assumed price effects of the Uruguay Round agreement. Even with only partial liberalization it is expected that, for example, wheat prices would rise by 22–28 percent compared with the 1986–88 baseline, with increases of 6–8 percent and 50–55 percent for corn and rice, respectively. These price increases are similar in magnitude to the price increases from complete liberalization discussed below and thus appear to be rather high.


Table A34 shows the preliminary draft budget only, which had predicted a smaller increase.


Austria, Finland, Iceland, Liechtenstein, Norway, Sweden, and Switzerland.


Except as noted, there have not been any recent significant changes in the operation of Japanese agricultural policies; the description of the system in Citrin (1990) remains applicable. The main features are summarized in Table A38.


The tariff rate will be 70 percent for the Japanese fiscal year (JFY) 1991/92 (April 1991-March 1992), 60 percent for JFY 1992/93, and 50 percent for JFY 1993/94. It will subsequently be subject to agreements in the Uruguay Round. There is also a safeguard mechanism utilizing an emergency tariff of 25 percent (Harris (1990)).


Different results for the nominal protection coefficient (NPC) can result from using different measures of world prices and the inclusion of freight. Nevertheless, alternative measures show similar changes. In OECD (1991) the NPC is estimated to have declined from 7.7 in 1987 to 5.2 in 1990.


For a description of the Cairns Group, see footnote 10.


For further details see Sandrey and Reynolds (1990).


A significant exportable surplus has been predicted by a number of authors, for example, Center for Economic Policy Research (1990).


Trade reforms by developing countries are discussed in Section V.


The example of the Australian wool reserve price scheme discussed above is but the latest example of failure.


Hufbauer, Schott, and Elliott (1990), conclude that, in general, embargoes have not been very successful.


An additional justification for intervention in agriculture is to stabilize farm incomes, but a pure income stabilization scheme would result in a zero level of support over time. This is not the case with support policies in most industrial countries.


Anderson (1991) cites evidence on the correlation between chemical fertilizer use and support levels. In 1985, Japan, Korea, and Switzerland used 375–440 kilograms of chemical fertilizers per hectare of arable land under permanent crops. Austria, the EC, and Norway, used 200–300 kilograms; the United States, 94 kilograms; Canada, 50 kilograms; New Zealand, 30 kilograms; and Australia, 24 kilograms.


Studies are reported in Table A41. To a large extent these differences reflect different views about the impact of production controls, especially the U.S. set-aside.


This base period is useful because it is the most recent used by such studies and it is close to the base period being discussed in the Uruguay Round. However, it was a period when support levels were very high.


The baseline scenarios assume that increases in world prices are passed on to consumers and producers, an appropriate response to a permanent change in the price of traded goods.


The model results are shown in Table A41 as A-T (1) and (2) and USDA (1) and (2). The baseline scenarios are designated (1) and the alternative scenarios are designated (2).


For example, the EC, Canada, the Nordic countries, Japan, and Korea. The restricted applicability of this provision arises because the import restrictions have to be on the product subject to the supply controls, or a like product. The problem in Article XI for the protection of the dairy sector is that the supply controls are on milk production but the import restrictions are on processed dairy products.


For a fuller discussion see OECD (1990b).


In the Uruguay Round, whether income supports paid on exported production will be classified as export subsidies is an issue.


The Czech and Slovak Federal Republic and Yugoslavia acceded to GATT as market economies; Poland, Hungary, and Romania acceded to GATT under special provisions; Bulgaria, an observer to GATT since 1967, applied to accede to GATT in 1986. Poland and Hungary have taken steps to renegotiate their Protocols of Accession.


While the immediate benefits of a Uruguay Round agreement could be small, the longer-term beneficial effects are expected to be substantial. Some estimates of the potential gains from a successful Round were reviewed in International Monetary Fund (1990c), pp. 74–76.


These issues are discussed in Collins and Rodrik (1991), Havrylyshyn, and Pritchett (1991), and Junz (1991). The first two studies provide empirical estimates of the potential impact of the restructuring of Eastern European economies and their patterns of trade.


Finger (1989), pp. 17–18. This estimate measures the static gains from the more efficient use of existing resources and changes in the terms of trade, but excludes possible improvements in efficiency stimulated by a more open trading system.


Regional arrangements among developing countries have been less successful in fostering development objectives and economic integration owing, in part, to incomplete internal liberalization and the failure to reduce high levels of protection.


VERs might be classified as restrictive business practices, particularly when they involve industry-to-industry agreements. Their classification above as a trade policy instrument reflects their frequent use as surrogate “safeguard” measures (instead of Article XIX of GATT) and as substitutes for antidumping duties.


To a degree, the implementation of competition laws has been modified in some countries to accommodate the realities of the global marketplace, but the analysis and remodeling of competition laws from the perspective of trade policy remains to be done.


This approach is considered more feasible than a “global competition policy,” which would require prior harmonization of all competition laws and a supranational competition authority.


McCulloch (1991). McCulloch also identifies a third motivation for investment: border-driven investments designed to capture the benefits of being multinational rather than being in a specific location. It should be noted that policies that effectively correct for market failures or externalities would improve resource allocation.


This is the “profit-shifting” model of Brander (1985), pp. 83–100. Its policy conclusions have been shown to be quite sensitive to the assumptions (no retaliation) and parameters, casting doubt on its practical application.


Grossman (1990), pp. 112–14.


For a discussion of these issues see Krugman (1987).


See OECD (1990a), Part 4, for a review of issues and policies related to government assistance to industrial research and development.


Environmental policies that are motivated by protection seek to offset the cost advantages other countries might enjoy because of differences in environmental standards. Trade policies used to achieve national environmental goals usually involve export controls and import bans, with or without complementary domestic measures. At the international level, discriminatory trade measures have been included as enforcement mechanisms in international environmental agreements (e.g., the Montreal Protocol Concerning the Protection of the Ozone Layer and the Basel Convention on the Control of Transfrontier Movements of Hazardous Wastes and their Disposal). Trade measures have also been used to impose one country’s environmental preferences on another country.


See Low and Safadi (1991) for a discussion of the design of environmental policies and the interaction between trade policy and environmental policy. In some cases, trade measures may be needed for effective enforcement, for example, a prohibition on the use of toxic substances.

Appendix I The GATT System

Recent activity in the General Agreement on Tariffs and Trade (GATT), in part, has been influenced by negotiations in the Uruguay Round. Resort to the GATT’s dispute settlement mechanism has been relatively frequent, but implementation of panel findings, in some cases, has been linked to the Round’s results; GATT surveillance was strengthened as the result of the Round’s midterm review; and there was a significant increase in GATT membership, possibly also reflecting, in part, the breadth of the negotiations. GATT membership now numbers 103 countries (Table A43), accounting for nearly 90 percent of world trade. In addition, 28 countries apply the GATT on a de facto basis, and most other major trading countries hold GATT observer status. Further, the Tokyo Round Codes, which clarify, extend, and enhance GATT disciplines in such diverse areas as subsidies and government procurement, have been accepted by virtually all industrial countries and an increasing number of developing countries (Table A44). The code disciplines, excepting government procurement, form part of the Round’s negotiations, with a view that they will apply to the full membership upon conclusion of the Round. Government procurement is the subject of parallel negotiations.

This Appendix deals with recent GATT activities, looking specifically at the Sessions of the CONTRACTING PARTIES,1 the work of the GATT Council of Representatives, GATT surveillance, GATT involvement with instruments of protection, and the Committee on Trade and Development, which is concerned with issues of trade interest to developing countries.

The Contracting Parties

The Session of the CONTRACTING PARTIES, which usually meets once a year, is the GATT’s senior body. As an early result of the Uruguay Round it was agreed in April 1989 that the Session would meet at the ministerial level at least once every two years; this reflects both the increasing importance of trade to international economic policy developments and the perceived need to improve the effectiveness and decision-making capability of GATT as an institution. It was the November 1982 Session, at the ministerial level, that found the multilateral trading system to be seriously endangered by, inter alia, protectionist pressures and an increasing disregard for GATT disciplines. Ministers consequently agreed to make determined efforts to ensure that trade policies and measures would be consistent with GATT rules and principles. This agreement paved the way for the 1985 Session to initiate the basis for launching the Uruguay Round in Punta del Este, Uruguay, in September 1986. At the most recent Session, held within a week after the December 1990 ministerial meeting failed to conclude the Uruguay Round, Contracting Parties reiterated their commitment to a successful conclusion of the Round, and set the stage for resumption of the negotiations by making explicit the general sense of lost opportunity that had permeated the closing days of the Brussels meeting.

The GATT Council of Representatives

In the period between meetings of the Session, the GATT Council of Representatives (from member governments, normally at ambassadorial level) meets about nine times a year to act on ongoing issues. Its recent agendas have been dominated by membership and dispute settlement matters, but it also has considered such diverse issues as the effects of German unification and the Canada-U.S. Free Trade Agreement, the relationship between trade and the environment, and questions regarding the continued appropriateness of a waiver that has allowed the United States since 1955 to use import restrictions on certain agricultural products.


Since August 1990, seven governments—Bolivia, Costa Rica, El Salvador, Guatemala, Macao,2 Tunisia, and Venezuela—have acceded to the GATT. In addition, membership requests from Algeria, Bulgaria, the People’s Republic of China, Honduras, Mongolia, Nepal, Panama, and Paraguay are in various stages of consideration, and Viet Nam is exploring the possibility of joining the GATT, as is Taiwan Province of China. The former U.S.S.R. was granted GATT observer status in May 1990; at the time, it announced that it sees this as a step toward membership, but several Contracting Parties, including Japan and the United States, made clear that acceptance of a Soviet accession request will hinge on progress in its building a market-based price system, consistent with GATT principles. The establishment of the Commonwealth of Independent States and changes in the economic regimes obviously will affect the future relationship with GATT as well.

Further, Poland is in process of renegotiating its Protocol of Accession to reflect the changes toward the greater market orientation of its economy. Accordingly, Poland seeks to replace its present commitment to a specific rate of expansion of its imports from GATT members with acceptance of the full GATT disciplines, including the usual schedule of bound tariffs3 on which GATT accession is normally based. The absence of a bound tariff schedule is making it difficult for Poland to participate in tariff negotiations and could limit its market access opportunities.

Similarly, Hungary has taken steps toward renegotiating its Protocol. It seeks an elimination of the special provision in its Protocol, including its right to use bilateral import measures vis-à-vis other Contracting Parties and vice versa, and Hungary’s waiver from its GATT obligations regarding trade with the members of the (former) Council of Mutual Economic Assistance. The normalization of Hungary’s GATT status can be expected to result in the withdrawal of the U.S. nonapplication (since 1973) of the GATT to Hungary. Although the United States grants Hungary most favored nation (MFN)4 trading status, it does so on a bilateral basis (which can be withdrawn unilaterally).

The new GATT members followed a trend, starting with the Mexican accession in 1986, of lowering and binding most of their tariffs and, in some cases, of taking measures to liberalize nontariff barriers. Thus, Bolivia, Costa Rica, El Salvador, Guatemala, and Venezuela agreed to bind the entirety of their tariffs, joining Chile and Mexico as the only GATT members to have done so; Macao has announced that by early 1992 it would submit a schedule to bind its tariffs at an acceptable level, and Tunisia bound about one fifth of its tariffs.5 In addition, Costa Rica will eliminate import surcharges, surtaxes, and quantitative restrictions by late 1994; Guatemala will eliminate its remaining quantitative restrictions by July 1994; Tunisia is to abolish import licenses on many products; and Venezuela will continue with its present program of removing import prohibitions, restrictive import licensing requirements and other quantitative measures, with the goal to eliminate them completely by 1995.

Dispute Settlement

This subsection deals with dispute settlement under the GATT (normally handled within the context of the Council), as distinct from that under the Tokyo Round agreements (which are discussed below under “Instruments of Protection”).

The working of GATT’s dispute settlement mechanism is normally based on consensus. This has resulted at times in lengthy delays in the establishment of dispute panels and in the adoption of their reports (with a single party able to block consensus). In 1989, the Council decided on certain improvements in the dispute procedures, including the establishment of some firm time limits for the conclusion of various phases of the dispute settlement process.6 In particular, the Council, unless it decides otherwise, is required to establish a panel no later than the Council meeting following that at which the request first appeared on the Council’s regular agenda. However, the changes did not address problems in the adoption of dispute panel findings nor their implementation.

The year 1990 marked a substantial decline in the number of dispute panels established by the Council—one panel in 1990, compared with 5 in 1989 and 13 in 1988 (Table A45). Part of the explanation is that Uruguay Round participants were loath to spoil the negotiating atmosphere by trade disputes with the anticipated conclusion of the Round, in December 1990, near. This is supported by the fact that the failure to conclude the Round was followed by the almost immediate establishment of four dispute panels.

The greater automaticity under the new procedures appears to have made GATT’s dispute settlement more effective. Seven dispute panels (four in 1991) have been established under the revised procedures; in three of these, panels were constituted despite serious reservations on the part of the party complained against (once the EC and twice the United States). Four of the seven panels have submitted their reports, of which two were adopted without difficulty and with no apparent problems regarding their implementation.7 However, adoption of one of these reports, which found a U.S. countervailing duty on imports of Canadian pork GATT-inconsistent, became entangled with the dispute settlement mechanism of the Canada-U.S. Free Trade Agreement (FTA). The United States noted it was the latter mechanism that would determine whether or not U.S. law had been correctly applied. Canada, with wide support, argued that it was not a matter of whether the duty was consistent with U.S. law, but rather that it had been found in contravention of U.S. obligations under GATT. Once the dispute mechanism under the FTA found against the U.S. duty, the United States, in July 1991, agreed to the adoption of the GATT findings. Understandably, this has raised widespread concerns that the GATT mechanism is being undercut.

The improved record is in contrast to that for disputes brought prior to the implementation of the new procedures: Two panels established in 1985 and 1987,8 respectively, still need to start work (due to disagreement on panel members); in 8 out of the 15 panel reports adopted under the old procedures, findings have not been implemented to the satisfaction of some GATT members; and in five of these, the party complained against, in agreeing to adoption of the panel report, tied implementation to the outcome of the Uruguay Round.9

The importance of an effective GATT dispute mechanism is underscored by the fact that even a problem-ridden one has been able to diffuse significant trade differences. Thus, as the result of recent cases, both Japan and Korea have taken measures to liberalize beef imports; Japan has acted to improve market access for eight agricultural products as well as for wines and alcoholic beverages; the United States has introduced measures to bring its customs user fee and its sugar import regime into line with its GATT obligations; and India, Norway, and Sweden have moved to liberalize restrictions on certain agricultural products.

Selected Other Council Matters

In the context of German unification, the EC implemented transitional measures, as of October 1990 until the end of 1991, to govern imports of industrial products from Bulgaria, the Czech and Slovak Federal Republic, Hungary, Poland, Romania, the former U.S.S.R., and Yugoslavia to the extent that they are for use in the former German Democratic Republic (GDR). In connection with these measures, which include tariff quotas, the EC requested a temporary waiver from its GATT MFN obligations. Concurrently, the United States proposed a working party to examine both the possible trade diversion effects measures and the possible trade consequences of the extension of the EC’s Common Agricultural Policy (CAP) to the former GDR. While the United States saw the EC’s waiver request as part of the proposed working party’s agenda, the EC saw it as a pro forma matter. After informal consultations, the waiver was granted by an unusual rollcall vote at the December 1990 Session of the CONTRACTING PARTIES along with a decision to create a working party to examine the effects of the measures and to monitor adherence to the waiver conditions.

Following the entry into force of the U.S. Canada FTA on January 1, 1989, the Council established a working party to examine the compatibility of the FTA with GATT provisions. The working party began its meetings in 1991, and the concerns raised by a number of members indicate that there may be no consensus on the full conformity of the FTA with the GATT. While there is a general view that the FTA follows the GATT closely, particular worries have arisen about its dispute settlement mechanism and its agricultural provisions. On the former, some members question whether parties to the FTA would feel more bound by dispute settlement findings under the FTA than under GATT. On agriculture, certain members felt that the perceived lack of full liberalization of trade in agricultural products with the FTA breached the requirement that a regional arrangement cover substantially all trade between its members.

The 1955 U.S. waiver has given rise to contention in the Council in 1990. The waiver has allowed the United States since 1955 to impose, under its Agricultural Adjustment Act, quantitative restrictions and special duties on certain agricultural products. For some time, a number of GATT members, particularly the EC, have felt that the United States has failed to adjust its agricultural policies sufficiently and that therefore the waiver is no longer justified. These views prevented the GATT working party on the 29th and 30th U.S. Annual Reports on the Agricultural Adjustment Acts from reporting consensus conclusions to the Council in 1990.10 With the United States of the opinion that it was in full compliance with its obligations, the Council’s consideration of the working party report, as well as of the 31st and 32nd U.S. Annual Reports on the Act, has been postponed, probably until after the Uruguay Round. In this regard, the United States’ agricultural offer in the Uruguay Round implied readiness to eliminate the waiver, leading the EC to unblock adoption of a dispute panel report, finding an EC complaint against continued U.S. restrictions on imports of sugar and sugar-containing products unwarranted.

In May 1991, in recognition of the heavy international environmental agenda leading up to the United Nations Conference on Environment and Development (UNCED), the Council held a structured discussion on the relationship between trade and the environment and the role that GATT would need to play. Participants generally were concerned about the widespread misconception that GATT was an obstacle to environmentally sound policies and agreed that this needed to be dispelled. They also concurred that GATT should not involve itself in the design of global environmental policies, but confine itself to its area of competence. Work would be needed, in particular, to ensure that trade measures taken to protect environmental standards were not simply a disguised form of protection.11 Views on how the work should proceed diverged, however. Some, particularly members of the European Free Trade Association, look to the activation of a dormant working group on environmental measures, while others prefer to await the outcome of consultations on the terms of reference for the group.12 In the interim, the Secretariat is to (1) set out the trade aspects of the preparatory work for UNCED and the history of GATT involvement in the area of trade and the environment, and (2) prepare an inventory of trade measures taken for environmental protection reasons as notified to the GATT.

GATT Surveillance

The surveillance functions of the GATT over trade policies and measures have been strengthened significantly since 1989. At that time the Council began its reviews of members’ trade policies and practices under the Trade Policy Review Mechanism (TPRM) and, in that connection, charged GATT’s Director-General to report annually on developments in the international trading system. Surveillance over specific aspects of trade policies and measures of individual members or groups of members is exercised through (1) the Committee on Balance of Payments Restrictions, which is responsible for multilateral surveillance of trade restrictions taken for balance of payments purposes, (2) in monitoring the Tokyo Round codes, and (3) in the Committee on Trade and Development.

Report by the Director-General

The Council, in April 1991, had before it the Director-General’s second Annual Report on the status of the trading system, in which he noted that, although slowing down, the growth in the volume of world trade in 1990 had continued to outpace that of output; however, the weakening of the global economy risked intensifying defensive trade actions. Nevertheless, there had been notable positive developments in trade policy, including far-reaching liberalization by Central and Eastern Europe and by some developing countries; and the growing and strengthened participation in the multilateral trading system by a number of countries as evidenced by the recent stream of accessions and accession requests and the disinvocation of cover for import restrictions maintained under the GATT’s balance of payments provisions. On the negative side was the temporary suspension of the Uruguay Round in December 1990. Although the negotiations had resumed in February 1991, major outstanding issues continued to present difficulties, including agriculture, services, tariffs, and rules. Other identified areas of concern were persistent trade frictions between the EC, Japan, and the United States, and recourse to antidumping and countervailing duty actions with protectionist overtones. Further, there was the need to guard against the risk of an introspective orientation of regional arrangements. The Director-General’s overview met with widespread agreement. It is clear, however, that the Council’s discussion of broad trade policy issues needs to evolve over time and, at this time, is inhibited on a number of points important to a thorough discussion by the ongoing Uruguay Round negotiations.

The Trade Policy Review Mechanism

The Trade Policy Review Mechanism (TPRM), established in April 1989, aims to make the trade policies and practices of each country under review explicit and to gauge their possible effects on both the country itself and the global trading system. The reviews, however, are not to judge whether a member’s trade measures conform with its GATT obligations, that being the task of the dispute settlement mechanism.

All GATT members are subject to the TPRM, with the frequency of reviews varying from two years to six years according to the importance in world trade of the country in question. The reviews are backed by a submission by the country and a paper prepared independently by the GATT staff; the papers are published together with the substance of the discussions in the GATT Council. Since initiation of the reviews in December 1989 through mid-1991, 17 members have been examined: all the traders on the two-year cycle (Canada, the EC, Japan, and the United States), as well as Australia, Chile, Colombia, Hong Kong, Hungary, Indonesia, Morocco, New Zealand, Nigeria, Norway, Sweden, Switzerland, and Thailand.

The first round of reviews already is contributing considerably toward transparency by clearly setting out existing policy measures and reasoning, which in turn provides the basis for informed discussions at a national and an international level. The TPRM also has focused attention on the widespread adoption of outward-looking, market-oriented trade policies as an integral part of structural adjustment that aims to improve the competitiveness of the economy; countries as varied as Australia, Chile, Colombia, Hungary, Indonesia, Morocco, New Zealand, Nigeria, and Thailand have chosen this economic policy path, switching from a previous bias toward protection and import substitution. In each case, a positive external environment was stressed as being necessary for the success of adjustment efforts. The reviews also have served to highlight the responsibility of the major traders in safeguarding and improving a liberal system. The relative openness of their economies was acknowledged, but each was subject to criticism. Thus, in the case of Canada, stress was laid on the lack of transparency in its multilayered tariff system. For the EC, emphasis was placed on the trade distortions stemming from the CAP, the harassment aspect of its use of antidumping actions, its widespread resort to bilateral trade restrictive agreements, and on the nontransparent nature of its national subsidies. With regard to Japan, concern was evident about its inclination to settle trade matters bilaterally, the lack of transparency in the administration of trade policies, the high level of support for agriculture, and the difficulty in penetrating its market. On the United States, the review was marked by unease that it may have weakened its past strong commitment to the GATT and the multilateral trading system; this was reflected in the stress on the threat of unilateralism, the resort to measures, such as VERs, and the unilateral interpretation of multilateral disciplines, particularly in the context of antidumping investigations and duties. As to other large traders, both Australia and New Zealand were urged to bind more tariffs and to lower rates in areas of concern to developing countries, such as textiles and clothing. Norway, Sweden, and Switzerland were commended for their generally low tariffs, but criticized for their high levels of agricultural protection and tariff peaks in sensitive areas like textiles and (in the case of Switzerland) for the presence of cartels. Hong Kong drew praise for having essentially no trade restrictions, but was urged to increase its tariff bindings.

Balance of Payments Provisions

The GATT authorizes the use of trade restrictions for balance of payments reasons under Articles XII, for developed countries, and XVIII:B, for developing countries. On restrictions maintained under Article XII, full consultations must be conducted annually by the GATT Committee on Balance of Payments Restrictions. On measures maintained under Article XVIII:B, consultations are normally held under simplified procedures every other year. However, under the simplified consultations, the Committee must decide whether these suffice or whether an indepth (full) examination of the concerned party’s use of trade measures is required. Full consultations consider the overall macro-economic situation of the consulting country, including its prospects and alternative means of adjustment; look at the member’s restrictive system in some detail; and take into account external factors, including access to markets, that affect the country’s balance of payments situation. Background documentation is supplied by the IMF to the Contracting Parties for all consultations and, in the case of full consultations, the IMF makes a statement assessing the balance of payments situation of the consulting party.

The use of the provisions under Article XII by developed countries has declined steadily (Table A46). Czechoslovakia is the only country recently to invoke the Article—in early 1991—in connection with an import surcharge on consumer goods and foodstuffs.

In contrast, Article XVIII:B has been used widely by developing countries and represents the most broadly applied exception to the GATT’s general prohibition on the application of quantitative restrictions; some 80 percent of all such restrictions notified to the GATT are justified under the Article. However, since 1989, five developing countries have disinvoked their current use of Article XVIII:B. Argentina, Brazil, Ghana, and Peru did so while the external financial situation remained under pressure, as part of a wholesale lifting of restrictions in the context of national economic policies to remove market impediments; Korea revoked its use of the provisions under comfortable, external circumstances, as part of lengthy negotiations in the Committee in 1989, and is now lifting its remaining restrictions over an agreed time schedule.

Since early 1990, full consultations have been conducted with Czechoslovakia and Yugoslavia and consultations under simplified procedures, with Bangladesh, Egypt, Nigeria, the Philippines, Tunisia, and Turkey. The consultation with Czechoslovakia went smoothly, with the Committee in firm support of its economic transformation efforts. Czechoslovakia’s use of the provisions was seen by many as being a model, in that the applied measure was price-based and to be removed within a preset time frame, with the intent to do so more speedily, if possible. The consultation with Yugoslavia also went forward in a sympathetic environment, with the Committee expressing understanding for Yugoslavia’s difficult situation and support for its economic policies, commending the scope and pace of economic and trade liberalization efforts.

In the simplified consultations, it was agreed that Egypt, Nigeria, the Philippines, Tunisia, and Turkey would proceed to full consultations. This halted the previous trend of few departures from simplified procedures for developing countries and partly reflects the view of many industrial countries that procedures need to be tightened, particularly to shrink the periods of application and to reduce use of non-price-based measures. Thus, while not denying the right of members under GATT to use trade measures for balance of payments reasons, developed countries have begun to stress their own right to examine such restrictions fully.

Instruments of Protection

Broadly, GATT recognizes tariffs as the only legitimate instrument of protection, although there are exceptions under which nondiscriminatory, nontariff restrictions may be maintained. GATT also allows countries, under certain conditions, to levy selective duties to protect domestic industries against subsidized or dumped imports. This section looks at various instruments of protection and developments in related GATT committees.


Tariffs in industrial countries have been reduced to minimal average levels in successive rounds of multilateral trade negotiations and a large proportion, often in excess of 90 percent, of their tariffs are bound. However, tariff reductions have been less pronounced in some sectors, particularly agriculture and textiles, and both tariff peaks and escalation (successively higher tariffs along the product chain) remain a feature in most countries. In many developing countries, average tariffs are considerably higher, at times at prohibitive levels, than in developed countries; moreover, although there are significant exceptions, a low percentage of their tariffs are bound.

The GATT Committee on Tariff Concessions, after overseeing the phased implementation of the Tokyo Round tariff cuts, continues as a forum for tariff matters, including the supervision of GATT tariff schedules. In recent years, much of the Committee’s work has been related to the implementation, begun in 1988, of the new Harmonized Commodity Description and Coding System (HS). Replacing a variety of different and often inconsistent classifications, the HS provides customs and trade officials with an agreed definition of the content of tariff lines, facilitating both the conduct of, and negotiations on, trade. The HS has been adopted by 65 GATT members, accounting for some 95 percent of the trade of Contracting Parties.

Nontariff Measures

The use of nontariff measures (NTMs) has grown in importance since the late sixties, both in relative terms, given the decline in tariffs,13 and in absolute terms. Thus, in 1990, selected NTMs14 affected 18.5 percent of the value of nonfuel imports into 22 industrial countries—an increase of some 20 percent over the 1981 coverage ratio.15 Many of these NTMs are in the form of VERs16 and other gray area measures that effectively bypass GATT’s safe-guard provisions.


GATT’s main safeguard provision is Article XIX, which allows temporary restrictions on imports when domestic producers experience, or are threatened with, serious injury; the Article provides for compensation to trading partners affected by the restrictions. In practice, Article XIX has proved relatively ineffectual if recourse is measured against the rise of discriminatory, gray-area measures. In the period 1985–90, GATT members used the Article 15 times, bringing the number to 139 for the period 1950–90, while during the same period, the known number of discriminatory export restraint arrangements grew from 69 to 284 (excluding those under the Multifiber Arrangement (MFA)).17 Since 1980, Article XIX has been invoked by industrial countries only, with one exception (Chile) (Table A47). As of the end of December 1990, 22 Article XIX measures were in force, of which 16 were for longer than three years.


The Subsidies Code, to interpret and apply GATT Articles VI, XVI, and XXIII, was negotiated in the Tokyo Round. The Code prohibits export subsidies on manufactures; for primary products it enjoins members to refrain from export subsidies that result in a “more than equitable share” of world export trade. Domestic subsidies are to be avoided to the extent that they have adverse trade effects. Offsetting countervailing duties can be imposed under the Code in the case of injury to domestic producers from imports of like products that benefit from subsidies (export or domestic). As of the end of 1990, the Code had 25 signatories,18 accounting for about 90 percent of Contracting Parties’ merchandise trade.

It is widely agreed that the Code has shortcomings, including interpretative problems. Views differ on the definition, scope and measurement of subsidies, measurement of injury, and the breadth of application of countervailing duties. The differences include such basic points as whether for a subsidy to be actionable a financial contribution by government is required, how different types of subsidies are to be treated, and the definition of “primary” and “manufactured” products, “domestic industry” and “like products.” There is also disagreement about the application of the injury test in determining the need for countervailing duties: some members (e.g., the EC) apply the test to all countries, whereas others (e.g., the United States) apply it only to signatories of the Code or to those that have entered into bilateral agreements to discipline subsidies.

The Committee on Subsidies and Countervailing Measures is charged with the oversight of the Code. For doing so, it depends on members notifying their relevant use of subsidies. Compliance with the notification obligation is sporadic, partly because of genuine data problems, but partly for defensive reasons. Thus, enforcement of the Code is foundering on nonsupport by the signatories, and trade frictions on subsidies have increased.

Nor has dispute settlement under the Code been very effective. Five of the six panel reports brought before the committee have not yet been adopted, as the consensus practice requires the consent of the party losing the case. The sixth report was adopted in the context of a bilaterally agreed solution to the dispute, with the report essentially limiting itself to an outline of the solution, but not coming to any findings. More recently, in May 1991, a panel was established to consider a U.S. complaint about the exchange rate guarantee scheme operated by Germany in favor of its civil aircraft manufacturing industry. The EC, though not objecting to the establishment of the panel, maintained that as GATT recognized trade in civil aircraft as a special sector, the proper forum for the dispute was under the Civil Aircraft Code. This has raised the more systemic point that action on a report in one forum may be held until parties see that they cannot do better elsewhere.


The Tokyo Round Agreement on Antidumping Practices came into effect in 1980. It interprets the provisions of GATT Article VI, which lays down the conditions under which duties (ADDs) may be imposed as a defense against dumped imports. At present, there are 25 signatories to the agreement, which is administered by the Committee on Antidumping Practices.

In some countries, antidumping actions are the most actively used trade policy instrument (tariffs, of course, are the broadest instrument). Despite a recent decline in the number of antidumping investigations, the broadening scope of such measures has become an important source of trade friction among Contracting Parties. In particular, the Antidumping Code, and its interpretation, has involved controversy over definition and measurement of both dumping and injury, and over the actual application of duties; problem areas also include the application of dumping in third markets,19 the use of price undertakings in antidumping proceedings, and revision and termination of undertakings. Legislation by some countries to widen the coverage of antidumping duties has been a source of concern to trading partners.

An important recent controversy relates to “local content” arguments in antidumping cases: what determines whether a product is derived from local or foreign sources—factory residence or domestic value added? In June 1987, the EC adopted regulations stipulating that ADDs may be applied, under certain conditions, to products assembled in the EC but using imported parts as materials; this was to prevent circumvention of ADDs on finished products. On this basis, the EC Council in April 1989 imposed ADDs on certain electronic products produced (but largely assembled) by five Japanese-related companies in the EC. Exports to the EC of similar products by the non-EC based affiliates of these companies were already subject to ADDs. The EC argued that these firms subsequently set up assembly plants in the EC to evade the original duties. The EC held that the newly imposed duties only enforced compliance with the original ones and, therefore, fell within allowable GATT exceptions to assure, inter alia, that nothing in the GATT shall prevent a member from adopting measures to secure abidance by GATT-consistent laws or regulations (Article XX:D). Japan, in its first dispute settlement case against another GATT member, filed a complaint against the duties. The panel, established by the Council20 in October 1988, found against the EC on the grounds that nothing in the GATT precluded action aimed at the avoidance of duties. The EC levies thus involved a discriminatory internal tax, contravening the requirement of national treatment for imports. The EC agreed to the adoption of the panel report, in May 1990, but noted that it would not consider amending its own legislation until the Uruguay Round had dealt effectively with the circumvention of duties (for which GATT provisions do not now exist).

More recently, there has been a notable increase in dispute activity in the context of the Antidumping Committee. The first dispute panel established by the Committee, in January 1989, found against U.S. ADDs on certain steel imports from Sweden; however, the report, which was submitted to the Committee in September 1990, remains unadopted.21 Late in 1990, the Committee established a dispute panel to examine a Finnish complaint against Australian duties on imports of Finnish power transformers and was informed of Canadian and Mexican requests for dispute settlement consultations regarding possible U.S. anticircumvention measures on duties on imports of color television components and of Mexican consultations with the United States on the latter’s imposition of duties on certain cement imports from Mexico. This increase in dispute activity is not surprising in view of the growing perception that antidumping activity is being used defensively, possibly frustrating rather than fostering fair competition.

Government Procurement

The Agreement on Government Procurement, negotiated in the Tokyo Round, is administered by the Government Procurement Committee, which has 12 members. With the government and its agencies the largest single purchaser in many countries, the agreement seeks to increase transparency in procurement laws, regulations, and procedures, and to ensure that such measures do not serve to protect domestic producers from international competition or discriminate among foreign suppliers or goods.

Much of the recent work of the Committee has been taken up with negotiations to improve and extend the coverage of the agreement. These negotiations, involving signatories only, are conducted apart from the Uruguay Round, but clearly are influenced by the dynamics of the Round. The first phase of the negotiations was concluded in 1988, with a lowering of the threshold value (above which procurement contracts must be open to bidding to the agreement’s signatories) from SDR 150,000 to SDR 130,000, and an extension of the agreement’s coverage to leasing contracts. Since then, negotiators have concentrated on increasing (1) the number of procurement entities subject to the agreement and (2) its scope to include services. By late in 1990, negotiating offers had been put forward by all major participants, with improvements and the formulation of a final package expected to be coming from further bilateral and plurilateral discussions. However, delay in concluding the Uruguay Round halted the momentum in the government procurement negotiations, which have made little headway since Brussels.

In only its second action on dispute settlement, the Committee established a panel in July 1991 to examine an EC complaint that a U.S. government entity had purchased equipment without following the required tendering procedures.22

Textiles and Clothing

Trade in textiles and clothing is subject to special restrictions under the MFA. The MFA was initially negotiated in 1974 (to succeed the Short-Term (1961–62) and Long-Term (1962–73) Arrangements on Cotton Textiles) and has been renewed three times. MFA IV went into effect in August 1986 and was scheduled to expire on July 31, 1991; it was recently extended for 17 months, pending the results of the Uruguay Round, which are expected to include a phase out of the MFA.

Overall management of the MFA is in the hands of the Textiles Committee, made up of representatives of its 41 signatories and chaired by the Director-General of GATT. MFA members account for some 80 percent of world textile and clothing exports (excluding intra-EC trade). While some 21 members maintain few, if any, restrictions, those implemented by others affect up to 50 percent of trade in the sector. The restrictions, in the form of bilaterally agreed export restraints, constitute a major departure from GATT’s central principle of nondiscrimination.

Recent activity in the Textiles Committee centered on the present extension, to the end of December 1992, of MFA IV. Initially, some exporters operating under restraints sought assurances that these would not be tightened during the extension period. Others sought a straightforward extension on the grounds that the appropriate venue for negotiations on the MFA was the Uruguay Round. The latter view held sway, as many MFA signatories thought that detailed negotiations on the extension could complicate the potential Uruguay Round agreement to gradually eliminate, or bring into GATT conformity, all MFA measures.

The Committee on Trade and Development

The Committee on Trade and Development (CTD) deals with issues of trade interest to developing countries. In particular, it examines how Contracting Parties put into practice Part IV of the General Agreement, which provides for differential and more favorable treatment for developing countries on a best efforts basis. The CTD also has primary responsibility for oversight of the Tokyo Round’s Enabling Clause, which is the GATT’s legal basis for MFN derogations for preferences to and among developing countries. The CTD, accordingly, monitors the implementation of the Generalized System of Preferences (GSP)—under which industrial countries acting individually grant preferential treatment, mainly in the form of lower tariffs, to developing countries—and, more recently, that of the Global System of Trade Preferences Among Developing Countries (GSTP).

Preoccupation of virtually all GATT delegations with work related to the Uruguay Round has kept the CTD’s activity low key in the recent past. Its main issues of concern have included graduation of certain countries from GSP schemes, country coverage of the GSTP, and credits and recognition in GATT (and the Uruguay Round) for unilateral trade liberalization by developing countries.

The Enabling Clause establishes the principle of integrating developing countries fully into the multilateral trading system—that is, the graduation of these countries from differential and more favorable treatment; however, the Clause does not lay down criteria for graduation. Therefore, some industrial countries, notably the United States, have incorporated graduation aspects into their GSP schemes. This has resulted in some instances in removal of GSP benefits to a number of newly industrialized economies (NIEs), whose competitiveness was seen as no longer warranting special preferences. Developing country concern has centered on the fact that GSP schemes are to be nondiscriminatory (under a 1971 GATT decision and the Enabling Clause). However, with most developing countries now full participants in the Uruguay Round and ready to offer market access concessions in return for those from developed economies, the graduation issue may be less divisive. Rather, the focus of CTD debate has shifted from preferences under GSP to general market access questions in the context of the success or otherwise of the market-oriented, outward-looking policies adopted by many developing and Central and Eastern European countries. The CTD is unanimous on the need for improved market access, but most feel that the structure and extent of a possible improvement is a matter for negotiation in the Uruguay Round.

A related element in the debate on how to help assure domestic and external balance emerging from economic reforms is “credit and recognition” for unilateral market opening by developing countries. While the logic of credit is acknowledged, the how and when is still not defined. It is, however, tacitly agreed that credit should be considered only—in terms of market opening—if liberalization is made irreversible, by binding it in GATT; in this regard, nonbound liberalization could be “recognized” by temporary reciprocity by industrial countries. The airing of views on this issue in the CTD is recent, but any resolution will stem from actual negotiation in the Uruguay Round.

The CTD’s concern about the GSTP involves the universality of its country coverage. The GSTP went into force on April 19, 1989 and now has 28 GATT members as signatories.23 Because membership is limited to the Group of Seventy-Seven, some industrial countries and some non-Group of 77 developing countries, believe that the GSTP is at odds with the nondiscriminatory requirements of the Enabling Clause.

Appendix II The Interface Between Trade and Competition Policies

This appendix discusses some of the major interactions between trade and competition policies. It aims to bring together some of the current thinking on these issues and is not intended to be prescriptive. A comprehensive analysis of the issues involved would require an in-depth review of the aims of competition policies, the legal principles underlying competition laws, and differences across countries in competition laws and their enforcement; trade theories and trade rules embodied in GATT and national legislation, as well as their implementation; and traditional and new theories of industrial organization. While a flavor of all issues involved is given, the paper focuses on inter-linkages primarily from the perspective of trade policy. Future work could examine these inter-linkages in greater depth from the perspective of competition policy.

Recognition of the fact that the effects of competition policies spill across national boundaries and that trade and competition policies are interrelated is not new; however, these spillover effects and interrelationships are now receiving increased attention. This is reflected in the communiqué of the OECD ministerial meeting of June 1991 and in extensive work programs under way in the OECD and elsewhere on these issues. The current focus on these issues primarily reflects increased tension and new challenges associated with the growing and rapid integration of the world economy through increased trade, but more particularly the globalization of investment and production and the associated increase in intra-firm trade and foreign sales and purchases of firms located abroad.1

The continued integration of the world through trade and investment has not, in many countries, been accompanied by the structural adjustment necessary to cope with increased competition. Thus, a number of industrial countries continue to protect declining industries and promote “strategic” sectors and “national champions” through nontariff and domestic measures. Such measures frequently conflict with competition laws of countries implementing these measures, as well as with competition policies of exporting countries.

The globalization of investment and production raises issues beyond those that have emerged, because of increased integration through trade. First, to the extent corporations choose to compete in foreign markets by establishment in those markets rather than through trade, their access to markets and the location of industry depends on foreign investment policies for new investment (“greenfield”) and takeovers of existing assets (mergers, acquisitions, etc.). Regarding takeovers, the increase in merger activity and takeovers associated with EC 1992 and the extensive powers granted to the EC Commission under its New Merger Control Regulation have brought these issues into sharper focus.2 Second, the emergence of the “global corporation”3 has blurred the distinction between “national” and “nonnational” firms and has raised questions about the relevance of traditional trade policy instruments and rules that assume nonnational firms supply domestic markets primarily through trade and about the relevant criteria for analyzing foreign penetration of domestic markets and domestic penetration of foreign markets.4 From the perspective of competition policy, globalization raises fundamental issues concerning the appropriateness of national competition laws and the role of governments in fostering competition.

To provide perspective on the issues covered in this appendix, a few general observations would seem necessary. Many of the problems at the interface of trade and competition policy are, in a sense, problems of trade policy and could be resolved within the existing multilateral framework for international trade. Indeed, some of these issues are on the Uruguay Round agenda; success in the negotiations could go a long way toward resolving international conflicts and creating more competitive markets. It is recognized, however, that a number of the issues at the interface of trade and competition policies go well beyond the Uruguay Round agenda.5 These include a growing divergence between the “nationalistic” objectives of governments and the “global” strategies of corporations;6 an inconsistency between “nationalistic” concepts underlying domestic trade and competition policies and the global concepts that are necessary in an increasingly interdependent world;7 the need for a better understanding of the access problems created by various business practices; and unresolved issues related to the policy implications of new theories of industrial regulation and “strategic” trade theory. Progress in these areas will likely require substantial analytical and empirical work to get a better handle on the issues involved and their importance.

The rest of the appendix is organized as follows. The first part defines the aims of competition and trade policies, their interaction, the potential for conflict, and the broad implications of these linkages. The second summarizes previous and ongoing efforts to harmonize competition laws and to reduce inconsistencies between trade rules and competition policies and laws. The third discusses the specific nature of the conflicts between these policies with reference to a selected group of trade measures and restrictive practices that currently are a major source of trade friction. The last part outlines recent proposals for resolving conflicts in this area and describes major research under way in the OECD and elsewhere on these issues.

Definitions, Aims, Sources of Conflict, and Implications

Competition policy defines the rules that regulate competition among firms. Although views differ on the appropriate aims of competition policy8 and the role of government in achieving those aims,9 competition policy, in broad terms, seeks to promote the efficient functioning of markets so as to maximize the welfare of consumers. Competition laws cover trusts, monopolies, mergers and acquisitions, and property rights, and they are intended to curtail the abuse of market power through price discrimination, price fixing, collusion, and other restrictive business practices. Such laws basically have been developed with domestic transactions in mind, although the competition laws of some countries (for example, the United States and Germany) and the European Community (EC) do have an extraterritorial reach when the conduct of a foreign enterprise in a foreign country has some “effect” within the domestic territory.10

Trade policy, which includes various multilateral, regional, and bilateral trade agreements, as well as national laws, that give effect to these agreements and other national trade objectives, in a broad sense, regulates competition among firms across national boundaries, although the codes and laws that regulate such competition are less comprehensive than those that regulate domestic competition. Competition and trade policies have similar roots in that both aim to promote competition and to provide remedies for dealing with noncompetitive behavior. Policies that promote freer international trade serve to sharpen domestic competition, and thus they reinforce the broad objectives of competition policy.

Conflicts between competition and trade policies and among nations arise for two main reasons. First, not all countries have well-developed competition laws, and, where they do exist, approach, jurisdiction,11 investigatory powers,12 procedural rules and remedies,13 and enforcement differ. These differences may limit access of foreign firms to domestic markets directly or indirectly via effects on inward direct investment, or they may enable domestic firms to unfairly target foreign markets. Second, both GATT and national trade laws permit trade-limiting actions in certain circumstances, which may weaken competition in both international and domestic markets. In addition, the trade-promoting policies of one country may conflict with the competition laws of other countries. Conflicts may also arise when countries have similar competition laws; for example, restrictive business practices that normally are actionable under competition law escape such action if they are taken under the “cover” of trade policy or government compulsion.14

Many of the issues being discussed under the U.S.-Japan Structural Impediments Initiative (SII) cover trade-related competition issues. These include restrictive business practices and lax enforcement of competition laws, which may restrict competition directly, and restrictions on foreign direct investment, which may restrict competition indirectly. Problems associated with the use of antidumping duties (ADDs) and voluntary export restraints (VERs) are examples of competition-related trade policies.15 These trade policy instruments tend to foster noncompetitive behavior by encouraging the formation of cartels, oligopolistic market structures, and the use of restrictive business practices. The specific nature of these conflicts is discussed below.

Inconsistencies between trade and competition policies, together with the increased potential for these inconsistencies to cause conflict in an interdependent world, call for (i) a re-examination of the GATT and national trade laws in terms of the principles of competition policy, and (ii) an examination of the relevance of competition laws for international trade.

Trade rules in particular need of review include remedies to deal with “unfair” trade practices (for example, antidumping and countervailing laws, and national laws to protect intellectual property rights); remedies to deal with import competition from “fair” trade (for example, Article XIX of GATT, VERs, and other gray-area measures); export promotion policies (for example, direct and indirect subsidies); instruments used to pry open foreign markets (for example, Section 301 of the U.S. 1974 Trade Law); and local content rules and rules of origin.16

In the area of competition policy, issues that need to be re-examined from a trade angle include the appropriate economic and geographic definition of the “market”; the criteria applied in assessing the impact on market structures and competition of mergers and acquisitions (both nationally and across borders); the impact on international trade of collusive behavior among domestic producers vis-à-vis foreign producers; collusive behavior among domestic and foreign firms; competition in production and marketing activities; and the activities of regulated industries.17

Related to the above issues are the rules governing foreign direct investment and the conditions under which foreign firms operate in domestic markets. Domestic policies in these areas affect the access of foreign firms to domestic markets and the conditions under which they compete against domestic firms. Disciplines on trade-distorting investment measures are under negotiation in the Uruguay Round. Other concerns in this area, which have arisen in the context of the rapid increase in outward direct investment during the 1980s, have given rise to proposals to create a more comprehensive rules-based system for investment that would complement the GATT.

Past Efforts to Deal with the Interaction Between Trade and Competition Policies

A number of attempts have been made in the past at the multilateral, regional, and bilateral levels to harmonize competition policies and to reduce conflicts between trade and competition policies. Most progress in this area has been achieved in the context of regional arrangements.18

Multilateral Efforts

The first and most ambitious attempt to deal with interrelated trade and competition issues was contained in the Havana Charter (1945) for an International Trade Organization. Chapter V of the Havana Charter contained policies and principles to assist in preventing business practices that adversely affect international trade, restrain competition, limit access to markets, or foster monopolistic controls.19 After the Havana Charter failed to obtain the necessary votes for ratification, a proposal was made in Economic and Social Council of the United Nations (ECOSOC) in 1953 to revive Chapter V as a separate instrument. The proposal was not accepted, however, because it did not receive the support of the United States or of a number of other countries.20 Similar, but unsuccessful, attempts were made in GATT between 1954 and 1960. Finally, in 1960 the CONTRACTING PARTIES adopted a resolution that recommended consultations whenever a Contracting Party felt it was being harmed by restrictive business practices of firms from other parties.21 Such activities were to be reported to the Secretariat; no such reports have been made to date.

Beginning in the 1970s, the United Nations continued efforts to develop harmonized international competition policies. In 1980, in line with the 1974 recommendations of a UN Group of eminent persons, “The Set of Multilaterally Agreed Equitable Principles and Rules for the Control of Restrictive Business Practices” (UNCTAD Code) was approved by the General Assembly. The Code is a set of voluntary guidelines that promotes the exchange of information and informal cooperation while leaving each country to devise its own laws. The Code deals with restrictive business practices, trade and antitrust issues, and the activities of multinational corporations, which are of particular concern to developing countries. The Code was reviewed at an UNCTAD conference in 1985, where particular attention was given to OECD analysis that pointed to the risks of antidumping legislation being applied anticompetitively.22

Within the OECD, discussions among members on restrictive business practices began in 1953.23 Subsequently, a series of Council Recommendations were adopted that provided the basis for cooperation on competition and trade issues. These include the 1976 Guidelines for Multinational Enterprises (MNEs); the September 1979 Council Recommendation Concerning Cooperation Between Member Countries on Restrictive Business Practices Affecting International Trade; the Revised Recommendations of the Council of May 1986; and the Recommendations of the Council of October 1986 for cooperation between member countries in areas of potential conflict between competition and trade policies.24 The 1976 agreement contains voluntary guidelines dealing with issues related to dumping, subsidies, distribution and licensing, and conflicting requirements that may be imposed on MNEs as a result of different approaches by OECD member countries to law enforcement and jurisdiction, and it provides for consultation between member states before the OECD Committee on International Investment and Multinational Enterprises. The 1979 and subsequent recommendations encourage governments to take account of the impact of their practices on competition, and they contain provisions for notification, exchange of information, consultation, and conciliation (by the Committee of Experts on Restrictive Business Practices—now called the Committee on Competition Law and Policy) in the event consultation does not produce a satisfactory solution. So far, the conciliation provision has not been used. To further enhance the coordination of trade and competition policies, in April 1985 the OECD Council approved the Indicative Checklist for the Assessment of Trade Policy Measures, which had been developed jointly by the Committee of Experts on Restrictive Business Practices and the Committee on Consumer Policy, in consultation with the Trade Committee (Table 18).25 More generally, trade and competition issues are discussed in an interdisciplinary manner within the OECD by a number of committees including the Trade Committee, the Economic Policy Committee, the Committee on International Investment and Multinational Enterprises, the Consumer Policy Committee, and the Committee on Competition Law and Policy. The work of these Committees is to be further integrated as part of the expanded work program that the OECD has in progress on trade and competition issues.

Table 18.

Indicative Checklist for the Assessment of Trade Policy Measures

article image
Source: Organization for Economic Cooperation and Development (1984), p. 24.

Regional Efforts

Attempts to harmonize trade and competition policies have been taken within the context of various regional arrangements. These include the European Community (EC); the European Free Trade Association (EFTA); and the Australia-New Zealand Closer Economic Relations Trade Agreement (ANZCERTA). These issues are also under discussion in the U.S.-Canada Free Trade Agreement (FTA), in the negotiations between the EC and EFTA for the creation of a European Economic Area (EEA), in the negotiations among Canada, Mexico, and the United States to form a North American Free Trade Area (NAFTA), and in the negotiation of Association Agreements between the EC and a number of Eastern European countries.

The Treaty of Rome, which established the EC, contains the most comprehensive set of rules to ensure the effective functioning of markets among different countries and to minimize the scope for conflict between trade and competition policies. Articles 85 to 94 give the European Commission power to control anticompetitive practices affecting the common market.26 In particular, noncompetitive behavior related to price discrimination and subsidies within the EC are covered by these articles and not by the EC’s Antidumping and Countervailing Laws, which apply to nonmember countries. In addition, Article 113 of the Treaty of Rome provides for a common commercial policy, although exceptions have been made (under Article 115) for purely national measures. Although the Single European Act (EC 1992) is silent about trade and competition policy, it is understood that the Treaty of Rome will be applied in full after 1993. In particular, the Commission has stated that it will not permit exceptions under Article 115, and laws related to state aids (subsidies) are being enforced more strictly.27 In addition, the EC Merger Control Regulation (implemented in September, 1990) creates extensive new obligations for the notification of concentrations between large companies that meet certain threshold requirements. The extent of extraterritorial powers in the directive has the potential to create conflict between the Commission and other competition authorities.28

With regard to EFTA, Articles 13 through 17 of the Stockholm Convention contain a set of competition rules covering government aids (subsidies), government purchasing practices, the right of establishment, and dumped or subsidized imports. To prevent dumping by companies of one EFTA member in the markets of other EFTA members, EFTA competition rules include a “boomerang” provision, which provides for the re-export of products to domestic markets of producing firms.29 Unlike the EC, however, EFTA does not yet have a supranational body to enforce competition policies within the free trade area.

With regard to ANZCERTA, Article 4 of the 1988 Protocol on Acceleration of Free Trade in Goods provided that, with the achievement of full, free trade in goods on July 1, 1990, neither country would take antidumping actions against the other; rather, each country would apply its competition laws to anticompetitive conduct affecting trans-Tasman trade in goods.30 To give effect to these commitments both countries amended their competition laws—the Trade Practices Act (Australia) and the Commerce Act (New Zealand)—effective July 1, 1990 to extend prohibitions on the anticompetitive use of market power contained in both acts to cover the abuse of market power within trans-Tasman markets. In addition, the competition authorities in each country were given new investigatory powers to obtain information and documents in the other country and the courts of each country were permitted to sit and take evidence (including by videolink) in each other’s country.31

In the negotiations leading up to the establishment of the United States-Canada FTA, issues related to the interaction between trade and competition policies were discussed. Indeed a major motivation behind Canada’s decision to negotiate this agreement was an attempt to shield its exports from “contingent” protection measures (antidumping and countervailing duties) in the United States. For its part, the United States was particularly interested in more effective discipline on the use of subsidies and measures to deal with subsidies and dumping. In the event, little harmonization of these policies was achieved in the FTA because of unresolved technical issues, uncertainty about the economic impact of greater price competition that would result if the existing antidumping laws of both countries were not applied against each other, and difficult questions on the conduct of domestic policy and the exercise of sovereignty.32 In addition, neither country was willing to dismantle subsidies, particularly domestic agricultural subsidies, on a bilateral basis, given concerns about subsidy practices in other countries. A Working Group has a five-to-seven-year timetable to negotiate rules on subsidies and methods for dealing with unfair pricing and government subsidization.33

In June 1990, the EC and EFTA launched negotiations to form a European Economic Area (EEA). The negotiations aim, inter alia, to harmonize competition laws and rules on subsidies; to agree on mechanisms to ensure that such rules are enforced; and to explore the possibility of using competition policy rather than antidumping and countervailing laws to deal with anticompetitive pricing practices within the EEA.34

Bilateral Efforts

The United States currently has bilateral agreements cover cooperation on antitrust issues with Australia, Japan, Canada, Germany, and the EC. In addition, France and Germany have a bilateral agreement with each other.35 These agreements aim to avoid or minimize conflict in the enforcement of antitrust laws and contain clauses to notify each other of pending proceedings, to cooperate in the development of evidence, and to consult about proceedings. Some provisions of these treaties specifically reflect the close relationship between trade policies and competition law enforcement.36

The agreement between the United States and the EC (September 1991) is the most recent bilateral agreement to have been signed. It includes provisions whereby each party will notify the other when its own antitrust enforcement activities raise the possibility of a jurisdictional conflict. It also includes “lead role” provisions whereby one party may take the major enforcement role in a case of common concern. The lead role provisions reflect the principle of “jurisdictional rule of reason.” This principle was developed in U.S. Courts to limit jurisdictional claims resulting from the prior reliance on the “effects doctrine,” which allows a state to claim jurisdiction over conduct that has an effect in its own jurisdiction, even if the conduct originates elsewhere. The parties have agreed to allocate jurisdiction on the basis of “functional advantage” (i.e., a case would be given to the authority with the greatest enforcement powers), the principle of “comity” (the maintenance of international respect and courtesy), an “agency clause” (which allows one party to request the other to take action against anticompetitive conduct in its own territory), and information sharing. Although the agreement does not contain a binding mechanism for resolving disputes, it is expected to provide one avenue for resolving antitrust disputes, including those that may arise under the EC Merger Control Regulation.

Bilateral cooperation between the United States and Japan on competition issues has been strengthened in the context of the U.S.-Japan Structural Impediments Initiative (SII), which was launched in September, 1989. The first Annual Report of the Working Group (1990) on SII contained commitments to implement measures to solve structural problems that, inter alia, were intended to remove impediments to more efficient, open, and competitive markets. These commitments cover a wide range of structural issues including a number that deal with the application and the enforcement of competition laws. In response to U.S. concerns about structural impediments to the Japanese market, Japan agreed to reform its distributive system, to make the keiretsu system more open and transparent, to eliminate exclusionary business practices, to liberalize the foreign direct investment regime, and to implement measures to adjust price differentials between Japan and the United States.37 For its part, the United States agreed to reform its antitrust policy so as to improve the legal climate for joint production ventures and to reduce uncertainty about the treatment of such ventures under antitrust laws. It also agreed to reform its product liability system so as to encourage innovation and enhance international competition, to maintain an open and nondiscriminatory policy toward foreign direct investment, and to eliminate VERs on machine tools (in the context of the Uruguay Round negotiations) and steel (in the context of the negotiation of a multilateral steel consensus).

Actions to fulfill these commitments were reviewed in the 1991 Annual Report of the Working Group. Significant measures implemented by Japan included a liberalization of foreign direct investment (consistent with the OECD Code of Liberalization of Capital Movements) and revised guidelines to ensure transparency and improved enforcement of the Antimonopoly Act. The guideline document, “The Anti-Monopoly Law Guidelines Concerning Distribution Systems and Business Practices,” was issued on July 11, 1991 by the Japanese Fair Trade Commission (FTC). It covers 28 major types of business practices in distribution and other business spheres and gives the FTC’s interpretation on enforcement policy under the Anti-Monopoly Law.38 The document was made available to other OECD member countries for review prior to its implementation.

Examples of Conflicts Between Trade and Competition Policies

This section examines selected trade policies that are currently a major source of friction among countries and their relationship to competition policies. These include trade measures to deal with “unfair” trade, including business practices that may limit access to markets, measures to provide temporary relief against a surge of imports associated with “fair” trade, and the use of subsidies to promote “strategic” industries.

Unfair Trade Practices and Remedies

Dumping and Antidumping Actions

In principle, antidumping and competition laws have similar objectives with respect to dumping (international price discrimination) and domestic price discrimination, respectively.39 Both seek to ensure that suppliers do not abuse market power so as to reduce or eliminate competition. In practice, however, the objectives of these laws tend to diverge. Whereas domestic competition laws are primarily concerned with protecting “competition,” dumping remedies often seem to be more concerned with protecting “competitors.”40 This, and other differences between antidumping and competition laws, results in different standards being applied to import-pricing practices in the enforcement of antidumping laws than would be applied if such practices were examined under competition statutes.41 For example, the standards for measuring price discrimination, assessing the adverse impact of price discrimination on producers, and choosing entities to look at when considering this adverse impact differ between antidumping policy and competition policy.42

To understand the source of conflict between antidumping and competition laws, it is useful to consider if dumping is harmful and, if it is, why can domestic competition laws not be used to contain its harmful effects? As summarized in Table 19, dumping may do economic harm to an importing country if it involves predatory pricing that is intended to drive domestic firms out of business or if it is done on an intermittent basis. Other forms of dumping typically involve pricing behavior that is considered normal market behavior if it occurs in domestic markets. As noted in Table 19, systemic arguments based on the concept of fairness, which are used to defend the use of ADDs to counter nonpredatory dumping, do not stand up to close scrutiny.

Table 19.

Dumping and Antidumping

article image

Although predatory pricing may be harmful to an importing country, the conditions required for successful predation are quite stringent (Table 19) and, in practice, this does not seem to be the predominant form of dumping.43 Intermittent dumping may occur more frequently, but it is difficult to deal with because harm may be done before ADDs could be applied. Thus, the possibility of successful predation would seem to be the main reason antidumping laws are needed.

In the above connection, the trade system needs antidumping laws to deal with predatory dumping because domestic competition laws cannot prevent foreign companies from becoming monopolies.44 A major defect in GATT’s antidumping provisions, however, is that they make no distinction between predatory and other kinds of dumping. Nor has such a distinction been made in national trade legislation.45 In practice, ADDs tend to be used primarily to counter dumping that is unlikely to be harmful, rather than to deal with predation. In many cases, ADDs are thought to present greater problems than the dumping they are intended to remedy.46 Indeed, existing antidumping regulations are considered to be anticompetitive in their objectives, methods, and instruments.47

The major problems in the implementation of antidumping laws are related to the determination of injury, the calculation of dumping margins, insufficient recognition of the principle of national treatment,48 and the use of antidumping provisions to sustain collusion in domestic markets (Table 20). The latter reflects the frequent use of price undertakings (EC) and VERs (U.S.) as remedies against dumping and an apparent lack of concern as to whether injury is done to a competitive or an oligopolistic industry (Messerlin (1990)).49 In this connection, recent reviews of antidumping cases in the EC are instructive.50

Table 20.

Examples of Problems with Antidumping Laws and Procedures

article image
article image

Based on Messerlin (1991).

Based on Rowat (December 1990).

Prices that do not cover full costs are typical in a cyclical industry (Barton (July 1990)).

This is considered by Messerlin and others to be the most important bias in the antidumping procedure in that it provides quasi-certainty of finding dumping.

For examples of like products that cause bias, see Messerlin (1990).

The EC review (based on Commission documents and data) covered all cases in the chemical industry (1980–89) involving both anticartel action under Article 85 (i) or Article 86 of the Treaty of Rome51 and antidumping action.52 It was found that 52 percent of antitrust cases under Article 85(i) had “twin” antidumping cases; thus EC firms being fined for price fixing were simultaneously asking for legal action against foreign price cutting. In addition, EC antidumping procedures were found to provide a high level of protection, to facilitate cartelization of the industry, and to reduce incentives for the industry to adjust. Also, the large net benefits obtained by firms from ADDs, despite anticartel fines imposed later, would likely result in future requests for more antidumping action. The Canadian study covered all antidumping cases during 1980–88. It concluded that it was extremely doubtful that any of the cases reviewed that were actionable under Canadian antidumping laws would have resulted in action under domestic competition law. In most cases, the use of ADDs reduced the pace of restructuring by domestic producers, which were found to be relatively inefficient in global terms.53

The experience with antidumping duties suggests that greater weight should be given to competition policy criteria, particularly those related to consumer interests, in the determination of antidumping cases. This might be achieved by applying the standards of competition policy to antidumping cases or by replacing antidumping laws with competition laws. In addition, the principle of “national treatment” needs to be recognized in measures dealing with the circumvention of antidumping duties, including through assembly operations that use dumped inputs. With the increased globalization of production, the latter will likely give rise to more frequent international disputes if not addressed.54

Subsidies and Countervailing Duties

Subsidies are provided for a variety of reasons, including those related to social and regional objectives, and are not trade distorting in all cases.55 Thus, GATT and rules embodied in the Subsidies Code permit the use of subsidies on exports of primary products; signatories are to avoid export subsidies on primary products that result in a “more than equitable share” of world trade and domestic subsidies that distort trade. However, the Code lacks precision in that it does not define what a subsidy is, nor does it provide much guidance on what subsidies are to be avoided.56 The GATT also permits countervailing duties (CVDs) to counter the trade-distorting effects of subsidies provided by trading partners. While there is an intuitive complementarity between subsidy rules and rules for CVDs, the rules in fact lack symmetry in that it may be legitimate in GATT for one country to provide a subsidy, and equally legitimate for another country to countervail that subsidy.57 The lack of symmetry and precision in subsidy and CVD rules has permitted these instruments to be used in a manner that reduces competition in domestic and international markets and undermines trade concessions made to trading partners.58

Of particular importance for trade and competition policy interlinkages is the use of subsidies to promote “strategic” industries in an attempt to gain a competitive advantage. In some cases this is defended by strategic trade theory, which suggests that, in certain circumstances, the use of industrial policies (of which subsidies are a key component) are justified to promote strategic sectors and to prevent global monopolization or oligopolization of such sectors. (Subsidies to Airbus have been defended on this basis.) But the “empirical” importance of strategic sectors is not known, and countries disagree on the optimal policy approach to foster such sectors.

The globalization of investment and production and the increased importance of competition based on technological superiority call for strengthened multilateral rules on the use of subsidies and countervailing duties and on access to government-funded research and development.59 In the latter context the treatment of “nonnational” firms, which frequently are not accorded national treatment, is of particular concern.

Protection of Intellectual Property Rights

Intellectual property rights (IPRs) are protected by a number of laws, including patent law (the most important industrial property law), copyright laws, and “unfair competition laws” that protect trademarks, trade secrets, the copying of designs, and the misappropriation of other commercial values.60 The protection of IPRs is considered necessary for the functioning of a competitive market-based economy; in particular, given the externalities inherent in IPRs, patent rights and other laws to protect IPRs are required as “second-best” solutions to allow inventors to recover the research and development costs (plus “normal” profits) of their inventions.61 Views differ, however, on the extent and form of the protection that is required; “There are problems of both ‘too much protection’ and ‘too little protection’.”62

Intrinsically, intellectual property laws, which confer rights on the owners of IPRs to exploit those rights monopolistically, and competition laws, which seek to prevent monopolies, would seem to be in conflict. At the national level, this potential conflict is resolved in some laws by conferring only a limited monopoly right (one that is limited in time, type, and horizontal coverage) to the holder of an IPR.63 These potential conflicts are further reduced in two basic ways. First, competition policy applies to clauses included in licensing agreements; second, competition policies provide the basis for remedies to deal with the abuse of IPRs. However, conflicts arise among countries because not all countries have laws to protect IPRs. Where laws do exist, differences exist in underlying principles,64 norms (coverage, rights conferred,65 duration of protection), and enforcement. In addition, competition laws differ in their treatment of practices and conditions attached to the transfer of technology (e.g., grant-back provisions, exclusive dealing, price fixing, etc.).

Although Chapter V of the Havana Charter included provisions relating to rights under patents, trademarks and copyrights, GATT does not contain similar provisions. Article XX(d) of GATT exempts IPRs from its disciplines; this exemption permits GATT members to take measures to protect IPRs provided such measures are applied in a nondiscriminatory manner. Article IX of GATT provides for cooperation between Contracting Parties to prevent the use of trade names to misrepresent the origin of a product, but no cases have been examined under this provision. During the Tokyo Round, the United States and the EC formulated rules to cover counterfeit goods, but these were not adopted due to the opposition of developing countries.66

In recent years, conflicts over the protection of IPRs have increased as technological progress has become an important determinant of international competitiveness. This conflict is most evident in U.S. trade relations with its trading partners;67 countries targeted by “Special 301” and other countries object to the unilateral character of this provision and to the possibility that concessions made to the United States might not be available to other countries on an MFN basis. Conflicts also exist, primarily among industrial and developing countries, with regard to the conditions included in licensing agreements.

At the national level and within the OECD, these developments have lead to a re-examination of the interaction between competition policies and laws to protect IPRs.68 At the multilateral level, it led to a decision to include trade-related aspects of intellectual property rights (TRIPs) among the “new issues” of the Uruguay Round.69 The decision to bring TRIPs into the Uruguay Negotiations reflected dissatisfaction (by the demanders for increased protection for IPRs) with the provisions of the World Intellectual Property Organization (WIPO), which provides disciplines on the manner (national treatment) but not the extent of intellectual property protection, and with the lack of enforcement mechanisms in WIPO.70

An agreement in the Uruguay Round on TRIPs would likely eliminate some differences in intellectual property laws among countries. Whether or not conflicts would remain in the area of licensing agreements remains to be seen. Many developing countries would prefer that a TRIPs agreement list abusive or anticompetitive practices that are to be excluded from licensing agreements; whereas most industrial countries would prefer to decide these on a case-by-case basis under their domestic competition laws.

The preference of industrial countries for the case-by-case approach reflects recent thinking on the economic and legal impact of particular business practices. In this connection, an OECD report on patents and licensing challenged the notion that competition policy and intellectual property rights are in basic conflict and recommended that competition authorities in member countries adopt a case-by-case approach to licensing agreements.71 The report argued that, in practice, the distinction between horizontal agreements (considered to be more restrictive) and vertical agreements (considered less restrictive) is difficult to make; that the presumption that the holder of an intellectual property right automatically has “market power” is incorrect; and that where market power does not exist, restrictions in licensing agreements may foster competition, provided there are no horizontal restraints; and that even when a technology with market power is licensed, vertical restraints in the agreement may be pro competitive in the absence of certain conditions related to market structure and entry. The conclusions of the 1989 report, which differ considerably from the earlier Council Recommendations on Restrictive Business Practices in Patent Licensing (1974), were accepted by the OECD Council; the Council withdrew the 1974 guidelines and recommended that governments take into account, to the extent possible, the conclusions of the 1989 report.72

Measures to Deal with “Unfair” Trade Practices Not Covered by GATT

Unilateral and bilateral approaches have been used with increasing frequency to pry open foreign markets and to deal with unfair trade practices not covered by GATT, or, in the view of some Contracting Parties, not adequately covered by GATT. Super 301 (now lapsed) and Special 301 (Section IV.e) of the U.S. Trade and Competitiveness Act of 1988, and discussions under SII (Section III) are prominent examples of this approach. Bilateral and unilateral attempts to pry markets open are of particular interest from the perspective of trade and competition policies because of their inherent inconsistency with either the principles of competition policy or, in some cases, with each other. In some cases (for example, the U.S.-Japan Semiconductor-Trade Agreements of 1986 and 1991), they tend to encourage anticompetitive practices in markets that are targeted; such anticompetitive practices are discouraged by the United States within the context of SII discussions.73

The approach used in SII raises more fundamental questions. Ostry (1990) questions the legitimacy of the objective implicit in SII—a desire to have Japan behave more like the United States. A number of other observers have suggested it would be preferable to address such issues in a multilateral forum. In the latter context, a number of considerations are relevant. First, similar conflicts exist among other countries and will likely increase with the completion of EC 1992 and the full implementation of the U.S.-Canada FTA. Second, a greater understanding of the access problems created by collusion and the lack of competition in distribution systems, as well as their impact on international trade, is required before appropriate multilateral solutions can be devised.74 Third, the inconclusive state of the academic debate on the optimal competition policy and the optimal degree of government intervention to influence market structures suggests that solutions in this area require a broader participation of countries if they are to reduce the potential for conflict.75

Measures to Deal with Fair Trade: Voluntary Export Restraints

Article XIX of GATT permits the use of safeguard measures to deal with temporary import surges associated with fair trade. Because the provisions of Article XIX are considered too restrictive by some GATT Contracting Parties,76 VERs and other gray-area measures are frequently used as substitutes.77 While the use of VERs might be challenged under Article XI of GATT, which prohibits quantitative restrictions,78 the parties that negotiate such agreements and have “standing” in GATT (i.e., governments in importing and exporting countries that negotiate VERs) have no incentive to do so. Consumers and users of products subject to VERs who bear the costs of these agreements have “no standing” in GATT.79

The prevalence of VERs raises policy issues that straddle both competition and trade policies. These include conflicts between competition laws and trade policy conducted outside domestic statutory frameworks; inconsistencies between elements of existing trade laws and the principles of competition policy;80 and differences in the enforcement of competition laws among countries.

The implementation of VERs normally requires firms in exporting countries to form cartel-type arrangements (with or without government endorsement) or engage in other restrictive business practices. Enterprises are thus caught between the necessity to solve disputes with trading partners and the requirements of antitrust laws not to engage in restrictive practices.81 In practice, however, such conflicts have not been very frequent because VERs are not subject to antitrust legislation in the importing country if they are covered by domestic trade laws.82 In cases where domestic trade policy has been conducted outside the domestic statutory framework, foreign government compulsion has been used to eliminate the risk of antitrust action in the importing country.83 Most VERs in the EC and the United States have been implemented under the cover of foreign government compulsion.84 But, in any case, the elimination of the risk of antitrust action does not reduce the anticompetitive characteristics of these arrangements.

Such arrangements are possible, not because competition laws and trade laws are inconsistent, but because elements of existing trade laws are not consistent with the principles of competition policy.85 Indeed, as already noted, trade policy overtly encourages anticompetitive behavior involving the formation of cartels and the use of restrictive business practices.86 In addition to these concerns, VERs create conceptual problems for antitrust market share analysis.87

In the above context, the recent agreement between the EC and Japan on imports of Japanese vehicles into the EC is important because it is viewed as a test case of the EC “post-1992” approach to imports in trade-sensitive sectors and because of concerns it has raised both inside and outside the EC.88 Aside from the continued protection of the EC vehicle market for a long transition period, the agreement raises other concerns involving trade and competition policies: the requirement for Japan to monitor exports will likely involve business practices normally considered to be restrictive; to the extent the agreement contains an implicit limit on the output of “transplants,” it raises questions about its consistency with national treatment and EC competition law; the requirement for Japan to monitor sales from transplants within the EC is only possible because of the block exemption of car distribution from EC competition laws, but this exemption is being questioned by consumer groups within the EC.89 The agreement implicitly recognizes its dubious legality under EC competition rules and the GATT in that it commits the EC to a declaration that EC competition law will not obstruct the operation of the agreement, and both the EC and Japan have agreed to defend this agreement in GATT.

Proposals to Reduce Policy Inconsistencies

Much of the basic rethinking that will be necessary to adapt trade and competition laws to the realities of the global market place is only in the early stages. Work is proceeding on two fronts. Within the Uruguay Round, negotiations to revise trade rules to provide more predictable access to markets have a bearing on a number of the issues covered in this appendix. Outside the Round, competition laws are being scrutinized to identify the extent to which a greater harmonization of these laws is required and how this might be achieved. In addition, thought is being given to the possibility of using competition laws rather than trade rules to regulate competition across national boundaries.

The Uruguay Round aims, inter alia, to refashion the rules on the use of subsidies, countervailing and antidumping duties, and safeguards, and to agree on rules of origin. In addition, new rules covering trade-related aspects of investment and intellectual property rights are being devised, and an agreement appears possible to completely phase out the use of VERs over a specified time frame. A successful conclusion to the Uruguay Round would go a considerable way toward reducing many of the conflicts that arise between competition and trade policies because of inadequate trade rules. An assessment of the extent to which the Round is successful in this regard will need to be made at the conclusion of the Round.

Proposals to refashion competition laws fall into four broad categories. These include the development of a global competition policy (this is considered a nonstarter but is described here for completeness and to contrast it with the EC 1992 model); the achievement of greater harmonization of competition laws through policy convergence (EC 1992 model); continued efforts within the OECD to harmonize competition policies (this would stop short of either of the first two approaches) and to encourage OECD member countries to give greater weight to competition policy considerations in implementing trade policy; and specific proposals to use competition laws rather than trade rules to regulate competition across borders. The latter include proposals related to dumping and safeguards.

A Global Competition Policy

A global competition policy would need to be based on agreed objectives of competition policy and on the role of governments in achieving those objectives; global economic welfare (rather than net national welfare) would be a leading contender among the possible objectives.90 These would provide the foundation for global competition laws, which would be enforced by a supranational competition authority. Global competition laws would involve rules for traditional areas covered by national competition laws such as antitrust, and abuse of dominant positions, as well as agreed rules on the use of subsidies and the protection of intellectual property. National competition laws would need to be made consistent with these rules and could be used to deal with the harmful effects of dumping and subsidies rather than existing national antidumping and countervailing laws. Goldberg and Ordover (1991) suggest that competition laws might also be used instead of GATT’s safeguard provisions to provide temporary relief for industries against import surges (see Section V.4.b). Such an approach would need to explicitly recognize that trade policies can undermine the objectives of competition policy and that current differences in competition laws and their enforcement among countries can undermine trade liberalization agreements. Moreover, it would be based on principles (for example, the geographic and economic definition of the market) and practices (for example, assessment of the effects of mergers and acquisitions on market structures) that allow for the increased integration of the world economy through trade, investment, and production.91

A Global Competition Policy with associated laws and enforcement procedures is not considered within reach in the foreseeable future. This primarily reflects a view that most governments would be reluctant to concede the sovereignty over domestic policies implied by this approach as well as a lack of agreement on what constitutes an optimal competition policy and on the appropriate role of government in fostering competition. Beyond these reasons, such a task would be difficult in that it would likely require a major revision of legal systems in many countries as well as harmonization of thousands of rules and regulations. Indeed, such an approach would go well beyond what has already been achieved in regional arrangements such as the EC, ANZCERTA and the U.S.-Canada Free Trade Agreement and also beyond what is being implemented in the context of EC 1992. As indicated below, however, EC 1992 suggests that a Global Competition Policy may not be necessary to achieve a greater harmonization of competition policies among countries.

Harmonization of Competition Laws Through Policy Convergence

A somewhat less demanding approach, but one which involves the most advanced model of policy convergence, is that being employed by the EC to implement the Single Market Program. This involves the prior harmonization of a minimum number of essential regulations; “mutual recognition” by member countries of regulations that are not harmonized; and factor mobility. As factors of production (capital, entrepreneurship, and labor) will likely move to locations with preferred regulatory systems, this will put pressure on national authorities to adopt these preferred systems and thus result in a convergence of such policies among EC member states. The advantages of this approach are (i) it does not require the prior harmonization of all laws, rules, and regulations; and (ii) it accepts different tastes and preferences among consumers in different countries, differing performance of individual enterprises, and acknowledgement that government policies affect the performance of market participants. Policy convergence under this model aims to optimize factor mobility and competition, not to harmonize tastes, culture, and corporate behavior.92 Despite its appeal, this approach is not considered feasible on a global basis, or even among all OECD members, because the necessary conditions (factor mobility and mutual recognition) are not all likely to be met. In particular, the requirement of labor mobility is considered a major obstacle to this approach.

The Role of the OECD

Given the difficulties inherent in achieving complete harmonization of competition policies, more pragmatic approaches are generally preferred to the far-reaching proposals outlined above. These involve stepped-up efforts within the OECD to achieve greater harmonization of competition laws.93 The OECD has in fact already expanded its work on competition policy and the linkages between competition, trade, and innovation policies. This program involves the trade and competition directorates and was agreed by the June 1991 OECD ministerial meeting (Table 21). In this connection, a study on “Nontariff Barriers to Trade and Competition: Theory and Evidence”94 has already been completed and other studies on antidumping and distribution-related barriers to international competition are in progress.95 The preparation of these studies is intended, in part, to demonstrate the use of the “checklist” approach approved by the Council in 1985 and thereby encourage OECD members to apply this approach at the national level.

Table 21.

Excerpts from OECD Communiqué, June 1991

article image

Authorities in a number of OECD countries have suggested that a greater harmonization of competition and trade policies would need to be achieved in three stages. First, competition laws would need to be established in all countries, including non-OECD countries, based on agreed principles. Second, steps would need to be taken to achieve greater uniformity in the enforcement of competition laws across countries. Third, at a later date, steps might be taken to ensure greater convergence of these laws across countries.

The Use of Competition Policy to Regulate Trade Across National Boundaries

Antidumping Laws

Most economists who have analyzed the effects of antidumping actions suggest that competition principles need to be given greater weight in the implementation of antidumping laws.96 One way to achieve this would be to replace antidumping regulations with the principle of national treatment under domestic competition laws.97 Thus, the same definition of undesirable pricing behavior that is applied to domestic firms would be applied to foreign firms exporting to domestic markets; this is similar to the approach used under ANZCERTA.98 Ostry (1990) and others suggest that this approach is unlikely to be adopted for two reasons. First, dumping is seen as a policy that achieves strategic trade objectives; it must be controlled to ensure a domestic base in strategic sectors because of “rents” or “externalities.” Second, because competition policy differs significantly among countries (in this case with respect to the treatment of pricing behavior), the application of national treatment would result in a substantial divergence in international pricing practices, which would likely be unacceptable to many firms and countries. Third, free trade among countries is considered essential if competition laws are to be applied to dumping to ensure that “parallel imports” arbitrage the difference between prices in exporting and importing markets.99

Other observers suggest that the above approach may deserve further attention. First, as already noted, dumping complaints normally are not associated with predation. In fact, evidence suggests that antidumping laws are more frequently used to shelter “sensitive” sectors and promote strategic sectors. Second, the argument that antidumping duties are necessary where arbitrage is not feasible is not persuasive; if the barrier to arbitrage is a formal trade barrier, it would be preferable to lower the barrier through a GATT proceeding or an administrative law action than to counter it through an antidumping duty;100 and where the barrier is a purchasers’ refusal to deal, as is often alleged to be the case in Japan, antitrust action is plausible.101 Antitrust action would not require a formal supranational enforcement procedure, although Barton suggests a GATT-type dispute settlement process might be useful to persuade nations to improve their domestic laws and to assure effective access to domestic courts for foreign parties.

Competition policy considerations could, however, be given more prominence in antidumping cases while stopping short of replacing antidumping laws with the principle of national treatment. This approach, which is supported by competition authorities in a number of OECD countries, would assign competition authorities a much greater role in examining dumping complaints than is currently the case; in fact, some argue that there is no reason why competition authorities should not be responsible for the implementation of antidumping laws.102 In this connection, the Uruguay Round negotiations are concerned more with streamlining and strengthening legal procedures and rules than with considering ways to give greater weight to efficiency considerations and consumer interests in antidumping laws.103

Voluntary Export Restraints and Article XIX of GATT

Both trade and competition authorities agree on the need to limit or ban the use of VERs because of the costs they impose on consumers and their adverse effects on competition. Within the Uruguay Round it is proposed to phase out the use of VERs over a transitional period and, in the meantime, subject the use of these measures to GATT surveillance; this is linked to new safeguard rules, where the main issue is selectivity. Although transition periods to remove VERs may be overly long, an agreement to totally phase out VERs could be a major step forward. To eliminate the possibility of government-to-government VERs being replaced by industry-to-industry VERs, it has been suggested that the exemption of pure export cartels from most competition laws (which is based on a “national” welfare criteria) should be repealed and that antitrust authorities in exporting countries should challenge such arrangements in the interest of “global” welfare.104 Given the strong built-in incentives for these authorities to consider only the national interest, this proposal may only be workable if it can be enforced by a supranational body.

The negotiations within the Uruguay Round on Article XIX are based on the implicit premise that “safeguard” actions are a necessary policy tool to sustain the multilateral commitment to free trade. While not disagreeing with this assumption, Goldberg and Ordover (1991) suggest that import relief might be based on competition policy, rather than on trade policy. This would involve a more relaxed antitrust policy to permit industries to restructure. In their view, competition policy is more transparent and nondiscriminatory than trade policy, and it is more likely to reflect efficiency considerations as well as consumer, national, and even global welfare concerns.

Ongoing Research on Trade and Competition Policies

In addition to the extensive OECD work program on trade and competition issues, major research programs are under way in thinktanks and elsewhere. A brief description of two programs involving The Institute of International Economics (Washington, D.C.) and The Royal Institute of International Affairs (London) is provided below. The projects will involve input and advice from trade policy practitioners, civil servants, business practitioners, and staff of international institutions.

The Institute for International Economics project, Global Competition Policies, is far-reaching and aims to survey national competition policies and the potential tensions they create in a globally integrated economy, and to evaluate proposals for international harmonization and strategic unilateral alteration of these policies. The project will include several country studies covering the philosophy of competition policy, statutes, practice, and institutional issues, including the relationships between relevant institutions for administering trade and competition policies and the extraterritorial reach of competition policies that cause conflict and inefficiency; thematic studies on issues, such as dumping and antidumping policy, VERs, subsidies, intellectual property protection and technology transfer, policies toward non-national firms, extraterritorial aspects of antitrust laws and policies toward mergers and acquisitions, the keiretsu structure of Japanese industry, and selective exemptions from antitrust policy; a survey of the history of efforts to achieve an internationally or regionally consistent competition policy, including the EC’s experience in the implementation of EC 1992; and a survey of recent academic literature on “new” trade policies, and “new” thinking on industrial regulation. Based on these studies, and taking into account the results of the Uruguay Round, the authors aim to develop a future agenda of competition policy issues that need to be resolved and to suggest a priority to items on this agenda and possible approaches and institutional arrangements that may be needed to address these issues.105

The Royal Institute of International Affairs project, Anti-Competitive Effects of Trade Policy, aims to investigate whether trade policies in the EC, Japan, and the United States have effects that contradict the objectives of domestic competition policies. Particular attention will be devoted to antidumping. The project will also consider whether trade policy should be reformed to make it more competition oriented and whether the scope of competition policy should be widened to take into account factors that are relevant to competition in international markets.

Appendix III Measures and Costs of Protection

With the decline in tariff rates to relatively low levels that are bound under GATT rules, the use of protective nontariff barriers (NTBs) by industrial countries has become widespread. Countries have resorted to orderly marketing arrangements, VERs, import quotas and levies, and various other forms of discriminatory intervention. In order to assess the significance of these developments, researchers have attempted to quantify the prevalence of NTBs and to estimate their effects. This appendix describes some of the frequently used measures of the incidence of nontariff barriers and the methodological problems involved in their construction and interpretation. It then reviews some of the empirical studies that attempt to quantify the effects of NTBs and the potential gains from trade liberalization. Stoeckel, Pearce, and Banks (1990) have estimated that the potential gains from further tariff cuts are relatively small compared with the eventual benefits that would flow from dismantling the nontariff barriers that have been built up during the past three decades.

Measuring Nontariff Barriers

Various methods have been used to measure the incidence and restrictive effects of NTBs; this section describes two methods that are frequently found in the literature. The first is used primarily to provide descriptive statistics on the kinds, pattern, and frequency of NTBs. The second involves the estimation of the ad valorem tariff equivalent of an NTB, or the nominal level of protection that an NTB provides.1

Descriptive Measures of Nontariff Barriers

The frequency index measures the percentage of tariff lines covered by NTBs. This index is often used to describe the restrictiveness of a trade regime, and it is one of the criteria used in Section V to classify countries’ trade regimes. This measure, however, does not capture the relative importance of the restricted items in either imports or domestic output, nor does it indicate the effectiveness of the restrictions in place.

The trade coverage ratio measures the percentage of import value covered by NTBs. It suffers from the problem of any “own” trade-weighted index in that products facing very restrictive NTBs will enter the calculation with zero or low weights. At the extreme, an NTB that precludes all imports of an item will receive a weight of zero. The index therefore underweighs the most restrictive measures. Another problem is that the coverage ratio includes all items covered by NTBs even though some of these restrictions may not be binding. Thus, as restrictions are relaxed the trade coverage ratio would tend to increase, rather than decline. This suggests that trade coverage measures such as those shown in Table 21 and in Tables A13 and A14 need to be interpreted with caution.

While these descriptive measures have serious drawbacks, their chief advantage is that they can be constructed with information that is relatively easy to obtain, and they give some idea of the prevalence of restrictive measures if not their overall effectiveness.

Price Impact Measures

Price impact measures attempt to quantify the effects of NTBs by computing their impact on prices. This is measured by the difference between the domestic price and the foreign price of a restricted item. The difference between the two prices, adjusted for known factors, is attributed to NTBs and is expressed as a ratio or as a nominal tariff rate.2

There are several problems with these measures. First, it is not always clear which prices to compare. Imported goods are often not identical to those produced domestically so there may be price differences because of differences in quality and other factors. Ideally, the price of the closest domestic substitute should be used, while the foreign price should be the c.i.f. import price inclusive of tariffs and other taxes.

Second, such price comparisons often do not account for transport costs inside the country. Sampling prices at different locations in the same country could significantly bias price impact measures. Third, the domestic-foreign price ratio for any particular year may be distorted by transitory factors, such as poor weather. Fourth, the comparison uses world prices that prevail under protection, which may be significantly lower than prices that would prevail under free trade. This is especially true for agricultural products. Finally, this approach can be biased by the degree of substitutability between domestic and foreign goods. If domestic and foreign goods are close substitutes, then a very restrictive NTB may produce only a small price differential.

Roningen and Yeats (1976) provide an example of this method. They compiled price statistics for 90 products that face NTBs in 15 developed countries. The analysis took the world price to be the lowest observed price in any country. Price differences were adjusted to account for domestic sales and excise taxes, tariffs, and transport costs. The remaining price difference is assumed to reflect the effects of NTBs. Table 22 summarizes some of their results.

Table 22.

Estimated NTB Tariff Equivalents in Japan and the United States, 1973

(In percent)

article image
Source: Laird and Yeats (1990), p. 49.

They found, for example, that in Japan restrictions on meat products were associated with domestic meat product prices that were 161 percent higher than world market prices. In the United States, the price impact of restrictions on meat products was estimated to be much less even though a higher percentage of the products studied was restricted; this could be due to greater substitution possibilities in the U.S. market.

Measuring the Effects of Nontariff Barriers

Significant difficulties arise in quantifying the effects of NTBs. The approach adopted in most studies involves converting NTBs to their ad valorem tariff equivalents, using methods similar to those discussed above. These NTB tariff equivalents are then treated as an additional protection beyond that of the tariff in place.

There are two major problems with this approach. Typically when a quota and tariff are both operating simultaneously, only one of them provides the binding restriction. For example, if a quota is the binding restriction, then the tariff has no impact on the level of trade flows. On the other hand, a sufficiently high tariff may make a quota nonbinding.

The second problem involves the estimation of NTB tariff equivalents. For the reasons discussed above, researchers have found that estimates of tariff equivalents are unstable and vary widely from year to year. For example, domestic and international price changes in agricultural goods have led to large annual swings in such estimates of 100 percent or more. Thus, the estimated effects of NTBs can differ substantially depending on the choice of tariff equivalents and other model parameters, such as supply and demand elasticities in product and factor markets.

Sectoral Studies

Reviews of sector studies of the cost of protection can be found in Kelly and others (1988), Goldberg and Ordover (1991), and Richardson (1989). These surveys provide ample evidence that the gains from protection in terms of jobs saved are outweighed by the welfare loss to the economy. Thus, it would be less costly to provide adjustment assistance to facilitate the exit of resources from sectors where comparative advantage has shifted abroad. Most sector studies reach this conclusion based on partial equilibrium models, but those based on general equilibrium models, which take into account the pull of resources from other sectors and the higher downstream costs to domestic producers, demonstrate more forcefully that the gains from protection are quite small compared with the costs. Reviewing industries characterized by imperfect competition and economies of scale, Richardson finds that the benefits of trade liberalization are even greater than under perfect competition. Selected studies in those sectors most affected by VERs and quota restraints are discussed below. It should be noted that in some sectors conditions have changed since these studies were done. For example, in recent years, the VER on Japan’s automobile exports to the United States has not been a binding constraint for most Japanese companies, and the VER on steel exports to the United States became nonbinding after 1986 for some major exporting countries.


Several effects of automobile VERs on prices have been documented. First, prices of imported automobiles subject to the restraint tend to increase as exporters raise prices to clear markets and earn quota rents. Second, to the extent domestic producers behave oligopolistically and domestic and imported automobiles are imperfect substitutes, domestic prices will also increase. Such price increases will be mitigated to the extent that outside competitors gain entry into the protected market. Finally, there will be some quality upgrading under the VER as producers shift toward production of cars with more optional equipment. It has been argued that VERs have led to substantial quality increases so that price increases as a result of them overstate the costs of the restraint to the consumer. However, such quality changes may also have some negative effects if they result in more limited consumer choice.

A number of studies have attempted to estimate these price effects, as well as the costs to consumers and the employment effects of such protection. For example, it was estimated that the VER between the United States and Japan resulted in U.S. import prices being 15 to 25 percent higher in 1985 than they otherwise would have been, while prices of domestic cars were about 6 percent higher. In another study of the effects of removing the VER in the mid-1980s, Collyns and Dunaway (1987) estimated that import prices would have been 23 percent lower, while domestic prices would decline by up to 12 percent. Feenstra (1985) estimated import prices to have been 16 percent higher with the restraint and domestic prices to have been 9 percent higher. The studies by Feenstra and Collyns and Dunaway attributed a significant part of these price increases to quality improvements, with the latter noting that while transactions prices of automobiles rose 23 percent in 1981–84, the automobile component of the consumer price index increased by only 6 percent.

Estimates of the costs of the U.S. VER to consumers have ranged from 17 to 20 percent of the value of total U.S. imports of automobiles (6 to 7 percent of apparent consumption), while net welfare costs, which subtract the gains to domestic producers from the quota, have been estimated at between 4 to 8 percent of imports (1½ to 3 percent of apparent consumption).3

Further, most studies have found the employment gains from the U.S.-Japan VER to have been modest. The estimates of employment gains vary from 20,000 to 35,000 in OECD (1987e) to 55,000 in Hufbauer, Berliner, and Elliott (1986). The costs to consumers per job saved have been estimated to be substantially higher than average annual wages and salaries.

A recent study by Smith (1989) estimated relatively large gains (on the order of 0.2 percent of GDP) from the elimination of the EC’s automobile VERs. This analysis broadly confirmed the results of other studies in concluding that eliminating VERs in the EC would increase Japanese automobile sales and reduce prices, with the reduction in prices of domestic cars being more modest. This study further concluded that a decline in profits of some European producers could result in consolidation of the industry through mergers and acquisitions.

Textiles and Apparel

The restrictive effects of the MFA have been examined in terms of the tariff equivalents of quotas, or the difference between the price under the quota and that under a free trade regime. For the United States, the tariff equivalent of the MFA quotas on textiles and apparel has been estimated at about 15 and 25 percent, respectively.4 A recent USITC (1989) study found that, in most cases, quotas were binding in 1987 and that the tariff equivalents on textiles and apparel were 22 percent and 28 percent, respectively. The tariff equivalent of German quotas on men’s and boy’s shirts has been estimated at 54 percent. For Canada, estimates indicate that without quotas, landed prices would have been lower by 17 percent for outerwear and 25 percent for shirts. These results are based on the hypothesis of perfectly elastic supply and may therefore be excessive. Estimates for the United Kingdom, which relax this assumption, suggest that abolishing the MFA would reduce the U.K. landed price of imports by only 5 to 10 percent.

These quotas have resulted in significant costs to consumers in the protected markets. Cline (1990) estimated the decline in U.S. net welfare from protection of the textile and apparel sectors to be $0.8 billion and $7.3 billion, respectively, in 1986. This amounts to about 36 percent of the total value of U.S. textile and apparel imports and 6½ percent of apparent consumption. Another estimate by the USITC (1989) based on 1987 data ranged between $4.1 and $4.3 billion or about 16 percent of the value of imports.

The Cline and USITC studies are similar in their estimates of the employment reduction from elimination of the MFA. Cline estimates this at about 234,000, while the USITC study estimates a range between 233,000 and 291,000. The consumer cost per job saved ranges from $14,000 to $19,000 in USITC (1989) to $35,000 in Cline.

In a study of the removal of global quotas and tariffs on textiles and apparel by the developed countries, Trela and Whalley (1990) estimated a global net welfare gain of approximately $23 billion in 1986. As expected, the simulation results indicated that the removal of such protection would result in a reallocation of production from the developed to the developing countries. The developing countries would also achieve a net welfare gain of around $8 billion.


A number of studies have estimated the costs of protection for the U.S. steel industry and the economy. The cost to consumers of U.S. VERs on steel has been estimated at about 3 percent of apparent consumption by Tarr and Morkre (1984) and USITC (1989) (or about 17 percent of the total value of steel imports for Tarr and Morkre and 12 percent for USITC). The number of jobs saved by the quotas has been rather small owing to the capital intensity of the production processes, ranging from 9,000 to 10,000 by USITC and Tarr and Morkre, respectively. A study by Tarr (1989) accounting for interdependence among various industries estimates jobs saved at 20,000. The consumer costs per job saved are generally well above $50,000 a year.

Multilateral Trade Liberalization

There is a broad consensus of support for a successful conclusion to the Uruguay Round of trade negotiations based on the benefits that would accrue from a more liberal trading system. A successful outcome of the Round is generally understood as one that would include substantial liberalization of agricultural trade, the phasing out of the MFA and other NTBs, and extended GATT disciplines in new areas of services, investment, and intellectual property.

While a consensus exists around the desirability of reaching agreement in the Uruguay Round based on a qualitative understanding of its benefits, there is less of a consensus regarding the quantitative effects of Uruguay Round-type trade liberalization. Some studies that attempt to quantify global and regional effects of multilateral trade liberalization do exist; however, the results obtained differ substantially for a number of reasons. Among these are differences in methodology and model structure, the measurement of NTB tariff equivalents, and the type of trade policy change simulated. Such differences, as well as similarities, may be illustrated by reviewing three multilateral trade models: Whalley (1985); Stoeckel, Pearce, and Banks (1990); and Deardorff and Stern (1990). It must be noted, however, that none of these models are capable of estimating some of the complex effects of negotiations in services, investment, and intellectual property.

Model Structure

The differences in the structure of the three models explain some of the differences with respect to simulation results. One common feature of these models, however, is that none of them take account of the dynamic effects associated with increasing returns to scale and imperfectly competitive markets. The simulated effects of trade liberalization measure the one-time static effects that would result after product and factor markets fully adjust to the policy change.

The Whalley model is a computable general equilibrium model and is most closely related to textbook trade theory based on comparative advantages, that abstracts from the macroeconomic effects of changes in aggregate demand and nominal exchange rates. Trade patterns are determined by the interaction of supply and demand in the context of differing endowments of capital and labor between regions. Full employment and balanced trade are assumed for all regions at all times. The basic method is to use given base-year data on production, consumption, and trade, together with certain substitution elasticity parameters for supply and demand relationships in order to calculate equilibrium relative product and factor prices; this step is referred to as calibration. Once the characteristics of the base-year equilibrium are determined, policy simulations may be generated by solving the model for different sets of tariff and NTBs, that is, counterfactual simulation. This method of analysis abstracts from disequilibrium adjustment of output and factor markets, emphasizing long-run real comparative static effects of trade policy changes. These effects are the changes in consumer welfare and producer costs of trade liberalization, terms of trade effects, and decreased tariff revenue and quota rents accruing to the importing country after liberalization.

The Stoeckel model is also a general equilibrium model that focuses on the long-run comparative static effects of trade policy changes. It combines an input-output structure of interindustry flows of intermediate goods with the three primary factors of production: land, labor, and capital. While product, land, and capital markets are assumed to clear, the labor market assumption only requires that there be no unmet labor demand. The primary difference between this model and the Whalley model is that trade may be unbalanced in the former. The Stoeckel model also permits changes in aggregate demand and output.

The Michigan model used by Deardorff and Stern is somewhat different in allowing some additional Keynesian short-run disequilibrium features associated with nominal rigidities and unemployment. It permits nominal exchange rates to have effects on the trade balances and employment. However, the simulations of multilateral trade liberalization incorporate the assumption that aggregate expenditure changes endogenously to maintain a given level of employment, with gross employment changes indicating the required adjustment in labor markets.

The Whalley model consists of 7 regions and 33 commodity groups. Assumptions with respect to consumer preferences allow for varying income elasticities of demand and production technology assumptions permit a variety of elasticities of substitution between inputs. Further, the Whalley model assumes that domestic products and imports from each source are not perfect substitutes.

The Stoeckel model consists of nine regions and seven commodity groups. This model generally assumes a greater degree of substitutability between primary factors of production (land, labor, and capital) and between imported and domestic goods than the Whalley model. However, the Stoeckel model is similar in assuming imperfect substitutability between products from different regions.

The Michigan model is the most disaggregated and includes 34 countries, 22 traded goods, and 7 nontraded goods. Elasticities of substitution between capital and labor, as well as between domestic and imported commodities, are closer to those in the Whalley model. However, while the other two models assume that imported goods from different sources are imperfect substitutes, as are imported and domestic goods, the Michigan model treats imports from all sources as identical while permitting incomplete substitution between domestic goods and imports aggregated from all sources.

Tariff and Nontariff Barriers

The Whalley model simulates the effects of eliminating all tariff and NTBs. NTBs are estimated in terms of their ad valorem tariff equivalents using a residual method developed by UNCTAD. This method entails allocating a part of the difference between domestic and world prices to transportation costs and tariffs, with the remainder expressed as a nontariff barrier term. The tariff equivalents of NTBs are set exogenously. In computing the base-year equilibrium position, this model allocates the quota rents from NTBs to the importing country. The tariff equivalent data for the United States, EC, and Japan are from UNCTAD and an arithmetic average of these was used for the other developed country region. These equivalents are assumed equal to zero for the Organization of Petroleum Exporting Countries (OPEC) and, given a paucity of data for the less developed countries, the Whalley model applies tariff equivalent figures from a study of Bangladesh to the developing country region. Finally, NTB tariff equivalents for the newly industrializing economies (NIEs) are assumed to equal the equivalents for the developing country region multiplied by the ratio of NIE tariffs to developing country tariffs.

The Stoeckel model simulates the effects of a global cut in tariff and NTBs by 50 percent. The NTBs are also expressed in their tariff equivalent form and determined exogenously in the model.

The Michigan model simulates the effects of a complete elimination of tariff and NTBs in the developed countries. In simulating these effects, estimates of ad valorem tariff equivalents were made using available data for major industrial countries and a subset of developing countries. High and low estimates of these data were made with the latter adjusted for the percentage of trade covered by NTBs. The Michigan model is unique relative to the others in that the tariff equivalents of NTBs vary endogenously to reflect the difference between the estimated demand for a product and the quota restriction. Thus, a decline in import demand owing to changes in other variables would reduce the estimated NTB tariff equivalent.

As may be seen from Table 23, NTB tariff equivalent estimates vary substantially from study to study because of the differing underlying data sources used in the method of calculation, the period for which data were collected, and the extent of trade coverage. Generally, it appears that, at least for manufacturing industries, the estimates of NTBs in the Stoeckel model are somewhat higher than for the other two models.

Table 23.

A Comparison of NTB Tariff Equivalents

(In percent)

article image
Sources: Deardorff and Stern (1990), Table A10; Whalley (1985), p. 286; and Stoeckel, Pearce, and Banks (1990), pp. 40 and 50.
Simulation Results

The seven-region Whalley model was used to simulate the comparative static effects of multilateral elimination of tariffs and NTBs. The results were presented in terms of the regional long-run net welfare changes which, when scaled by GDP, yield a 2.2 percent gain for the EC, a 0.5 percent increase for the United States, a 2.8 percent gain for Japan, and an overall global increase in welfare of 0.4 percent of world GDP. The small overall gain is due primarily to terms of trade losses in developing countries. These arise in this experiment because trade barriers in developing countries are very high compared with other regions. In the context of the Uruguay Round, a more appropriate experiment would be an equiproportionate cut in trade barriers.

The Stoeckel model was used to simulate the effects of a 50 percent reduction in tariffs and NTBs. Since trade was permitted to be unbalanced in this model, welfare calculations of the type made in the Whalley model were not, strictly speaking, permitted. However, as mentioned, GDP changes were reported for this simulation and might roughly be compared with the results from the Whalley model. The EC was estimated to gain 5.2 percent of GDP from this liberalization, while North America gains 3.9 percent and Japan, 9 percent. The gain in GDP for the countries reported (and which does not include a residual country category) was estimated at about 5 percent.

Deardorff and Stern used the Michigan model to simulate the effects of a complete trade liberalization by developed countries (including elimination of agricultural subsidies). They do not present welfare calculations for the same reason that the Stoeckel model does not; that is, since trade is permitted to remain unbalanced, welfare calculations would be complicated by issues regarding accumulations and decumulations of reserves. However, a comparison of the effects of this simulation on changes in exports and imports with Stoeckel’s simulation of a 50 percent global liberalization is presented in Table 24. Generally, the Stoeckel model estimates a much greater increase in trade than the Deardorff and Stern results.

Table 24.

A Comparison of the Stoeckel Model and the Deardorff and Stern Model

article image
Sources: Deardorff and Stern (1990), pp. 256–57; and Stoeckel, Pearce, and Banks (1990), p. 70.

Overall, the results of the three models suggest that the Michigan model lies between the more extreme results of the Whalley and Stoeckel models, to the extent that comparisons can be made between these models given their different structures and simulation experiments. Whalley’s results appear to be driven by rather strong terms of trade effects; while Stoeckel’s results appear to be driven by somewhat higher degrees of substitution between products and factors of production. The Stoeckel model also assumes higher NTB tariff equivalents than the other two models; this together with higher assumed elasticities of substitution explains the relatively large gains estimated from trade liberalization.

Statistical Appendix

Table A1.

Major Industrial Countries: Real GNP, Domestic Demand, and Trade Volumes, 1973–90

(Annual changes, in percent)

article image
Source: International Monetary Fund (1991b).

Compound annual rates of change.

GDP at market prices.

Average of expenditure, income, and output estimates of GDP at market prices.

Data through June 30, 1990 covers West Germany.

Table A2.

Shares in World Exports of Manufactures, 1963–891

(In percent)

article image
Sources: General Agreement on Tariffs and Trade (GATT); International Monetary Fund (IMF); United Nations; United Nations Conference on Trade and Development; and IMF staff estimates. Based on GATT classifications.

Standard International Trade Classification (SITC) categories 5 through 8 minus 68.

EC-10: Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, and the United Kingdom.

The newly industrializing economies (NIEs) of Hong Kong, Korea, Singapore, and Taiwan Province of China.

Table A3.

Developing Countries’ Exports by Product, 1980–88

article image
Source: General Agreement on Tariffs and Trade, International Trade (Geneva), various issues.
Table A4.

Shares of Exports of Manufactures by Area and Sector, 1973–88

(In percent)

article image
Source: General Agreement on Tariffs and Trade, International Trade (Geneva), various issues.
Table A5.

The Level of Post-Tokyo MFN, Applied, and GSP Tariffs in Selected Developed Countries

article image
article image
Sources: Laird and Yeats (1990); calculations based on the General Agreement on Tariffs and Trade Tariff Study and the United Nations Conference on Trade and Development’s Series D Trade Tapes.Note: EEC = European Economic Community; GSP = Generalized System of Preferences; MFN = most favored nation.