Abstract

Since 1981, the industrial countries have restored their share of world exports to almost the level prevailing in 1973, before the quadrupling of oil prices (Table A1).20 The counterpart of this increase has been a decline in the developing countries’ share to less than 20 percent of world exports. The trends in the major industrial countries diverge: Japan’s share rose, whereas the U.S. share declined, and the EC share remained roughly stable if intra-EC trade is excluded.

Trade Trends

Since 1981, the industrial countries have restored their share of world exports to almost the level prevailing in 1973, before the quadrupling of oil prices (Table A1).20 The counterpart of this increase has been a decline in the developing countries’ share to less than 20 percent of world exports. The trends in the major industrial countries diverge: Japan’s share rose, whereas the U.S. share declined, and the EC share remained roughly stable if intra-EC trade is excluded.

These trends mainly reflect large terms of trade movements that occurred in the 1980s. In real terms (constant 1980 U.S. dollars), the industrial countries’ share of world exports has remained constant since 1981, as has their share of world exports of manufactures. The divergent trends in the shares of total exports of Japan and the United States are also apparent for manufactured products (Table A2). Although the EC share of world manufactured exports has increased, the share exported to third countries has declined.

A long-term trend that is apparent over the past twenty years is a pronounced increase in intra-industrial country trade as a proportion of total trade. The percentage of intra-industrial country trade in the total trade with the rest of the world of a group of 11 industrial countries21 increased from 46 percent in 1964 to 60 percent in 1985. 22 An important contributing factor to the growth of such trade has been the growth of intra-industry trade—largely in capital-intensive and science-based industries—for which the growth rate of exports has generally been higher than that for total exports. The growing integration of the industrial countries’ economies has allowed them the advantages of the dynamic functions of trade, namely, large-scale production, product differentiation, and specialization. Integration was aided by the significant trade liberalization, particularly in manufactures, that took place in the first three decades of the postwar period. Since then the industrial countries have increased their use of nontariff measures (NTMs), but in general they retain more liberal trading systems than do the developing countries. This factor has probably contributed to the pronounced relative growth of intra-industrial country trade over the past two decades. Sluggish growth in the developing countries since the onset of the debt crisis in 1982 was a contributing factor.

Trade Policies

Tariffs

Successive rounds of multilateral trade negotiations have reduced most-favored-nation (MFN) tariff rates in industrial countries to an average of 5-6 percent on industrial products, although tariffs on agricultural products remain considerably higher (Table A3). Average rates are lower for some products, reflecting tariff reductions beyond those agreed in the Tokyo Round in some countries, as well as preferential trade agreements maintained among industrial countries and between industrial and developing countries. Problems of tariff peaks and escalation, however, remain. Thus, 33 percent of EC tariff lines have MFN rates above 10 percent, while Japan and the United States have such rates on about 17 percent of their tariffs.23 These tariff peaks tend to be concentrated in textiles, clothing, footwear, and some petrochemicals. Tariff escalation is also a feature of most industrial countries’ schedules, particularly for certain foodstuffs, leather, fabrics, and some petrochemicals. For example, gasoline enters most developed countries at low tariffs but polypropylene does so at MFN rates of 12.5 percent in both the EC and the United States and at about 18 percent in Japan. Furthermore, not all tariffs are bound in GATT, particularly on agricultural products (Table A4). Tariff preferences granted under preferential trading arrangements affect a larger proportion of world exports than preferences granted under Generalized System of Preferences (GSP) schemes (Table A5).

Nontariff Measures

The increase in nontariff measures may have largely offset the liberalizing effects of tariff reductions in the postwar period. For example, it is estimated that the economy-wide tariff equivalent of U.S. nontariff barriers on textiles, steel, and automobiles is about 25 percent, bringing protection to its early postwar level.24 Nonfuel imports of industrial countries subject to selected nontariff measures are estimated by UNCTAD to have increased from 19 percent of total imports in 1981 to 23 percent of the total in 1987 (Table 3). The sharp increase in voluntary export restraints (VERs) between September 1987 and May 1988 has probably further raised the total incidence of nontariff measures (Table 2).

Table 3.

Industrial Countries: Imports Affected by Selected Nontariff Measures1

(In percent of total imports)

article image
Source: UNCTAD (1988b).

Includes certain paratariff measures, import deposits and surcharges, variable levies, quantitative restrictions (including prohibitions, quotas, nonautomatic licensing, state monopolies, VERs, and bilateral restraints under the MFA), automatic licensing, and price control measures. In contrast to the Fund staff estimates presented in Table A8, it also includes antidumping and countervailing actions, and import surveillance.

Nontariff measures can take the form of border or nonborder measures. Voluntary export restraints are a common type of border measure. These gray-area measures are applied on a discriminatory basis and have increased in recent years. Preliminary data indicate that 261 such arrangements existed in May 1988, close to twice as many as in September 1987 (compared with 50 in 1978). About half of them are directed at developing countries, including heavily indebted countries, and four fifths are intended to protect the EC or U.S. markets.

Subsidies are the most important form of nonborder measure, but restrictive government procurement practices and technical standards are also significant. Trade frictions persist over the use of nonborder measures. Bilateral and multilateral discussions are under way to bring more discipline to the use of subsidies, to develop international standards acceptable to all countries further, and to open to international bidding a greater proportion of government procurement (currently less than half is open to international bidding).

Significant differences exist among the major industrial countries in their use of various types of nontariff measures. Protection in the EC and the United States tends to be more selectively targeted at specific exporters than in Japan. The EC accounts for nearly half of all VERs known to exist worldwide as of April 1988. The EC also extensively uses administrative controls, including import licensing to monitor imports and to enforce quantitative import restrictions, as well as variable levies on imports of agricultural products. The United States also relies heavily on VERs. Japan relies on global quotas for some agricultural products; its distribution system and other invisible barriers are seen by some countries as limiting market access.

Antidumping and Countervailing Duties

Industrial countries made increasing use of countervailing and antidumping duties in the 1980s (Tables A6 and A7). Although their number has declined somewhat in recent years, the trade values covered have increased. Disputes on their use have become more frequent as a result of differences in interpretations of definitions, measurement problems, and of the conditions of their legitimate use to combat unfair competition (see Section IV).

Australia, Canada, the EC, and the United States account for nearly all countervailing and antidumping investigations initiated by industrial countries. The United States has made more extensive use of countervailing (as well as antidumping) duties than other countries, reflecting in part the greater ease with which countervailing measures can be applied under U.S. legislation, which predates GATT and does not require ah “injury test” for nonsignatories of the GATT Subsidies Code (or those without bilateral agreements with the United States on subsidies). It also reflects the limited use of subsidies as an instrument of industrial support in the United States, and the lead the United States has taken in trying to reform other countries’ practices in this area. By contrast, EC countries have initiated few countervailing duty actions. They find it easier to invoke antidumping provisions because EC antidumping legislation is broader and sharper than countervailing duty legislation. Japan has almost never initiated antidumping or countervailing duty investigations.

The increased number of antidumping and countervailing duty investigations has resulted in charges from both industrial and developing countries that such investigations are used as a form of “administered protection” rather than to counter “unfair” trade practices. There is indeed some evidence that countervailing and antidumping duties are sometimes used as a substitute for safeguard measures;25 that foreign exporters are subject to disciplines that are not imposed on domestic producers;26 and that investigations are used to harass foreign exporters and force them into export restraint arrangements.27 A number of voluntary export restraints that protect the EC and U.S. markets are the result of antidumping or countervailing investigations. Recent changes in legislation to broaden the scope and intent of national legislation have heightened concerns in this area. These changes include the introduction of separate EC antidumping legislation to cover shipping and products assembled in the Community by foreign firms (Appendix I); and a ruling by the U.S. Court of International Trade that would make it more difficult to lift antidumping duties after they have been imposed (Section IV). Partly in response to these developments, in its latest report on world trade developments the GATT Secretariat notes that antidumping legislation, in particular, has evolved into a major tool of trade policy.28

Trade Policy in Individual Countries

United States

Over the past several years, growing external deficits in the United States and a heightened sensitivity to “unfair” foreign competition have intensified protectionist pressures. In line with its declared policy of free and fair trade, the U.S. Administration has attempted to avoid increasing protection by playing an active role in the Uruguay Round and by pursuing an aggressive policy to increase its access to foreign markets. This strategy has to some extent been successful. The Administration resisted protectionist proposals by Congress in specific sectors (e.g., safeguard measures against footwear imports in 1985 and new efforts to restrict textile imports), which were causing concern to both the Administration and trading partners. It has also undertaken bilateral market-opening discussions with trading partners. Nevertheless, the share of imports of manufactures covered by nontariff measures has increased more in the United States than in any other Group of Five (G-5) country in the first half of the 1980s (Table A8). In part this situation may have resulted from the appreciation of the U.S. dollar, which served to erode the competitiveness of a number of sectors. The United States maintains 62 voluntary export restraints out of a total of 261 known to exist worldwide as of May 1988. These affect mainly imports of automobiles, steel, and textiles and are directed at a broad range of exporting countries.

The major aspects of U.S. policy of concern to other countries include uncertainties in the manner of implementation of the recently passed Omnibus Trade and Competitiveness Act of 1988;29 its use of bilateral approaches to settle trade disputes, which third countries sometimes perceive as disadvantageous to them; apprehension that the United States will increase its use of bilateral free trade agreements to the detriment of the multilateral trading system; its competitive subsidization of agricultural exports, which is contributing to depressed commodity prices with adverse effects on efficient agricultural exporters; and a perceived increase in Us use of “administered protection.”

In this context, a number of countries have noted changes in recent years in U.S. trade legislation including its frequent use of legislation on “unfair” trade.30 Over the period 1980-87 the United States has initiated 411 antidumping investigations, 283 countervailing investigations, 60 safeguard investigations, and 60 investigations of “unfair” trade practices abroad—-the latter under Section 301 of the U.S. Trade Act of 1974 (Table A9). Almost half of all antidumping investigations were directed against the EC and japan, while countervailing investigations were mainly against suppliers in Brazil, Mexico, and the EC. It is noteworthy that Canada considers “administered protection” a barrier to trade and cited the desire to reduce its adverse effects among the reasons for entering into a free trade agreement with the United States.31 These effects include the administrative and legal costs imposed on Canadian exporters by the initiation of an investigation and the uncertainty caused by the threat of contingent protection measures, which has by itself an adverse effect on trade flows and investment.

Partly reflecting the above concerns, the EC now publishes an annual list of U.S. trade barriers, to give a “more balanced view” of the U.S. trading environment. The EC document was first published in 1985 in response to the Report on Foreign Trade Barriers compiled by the U.S. Trade Representative’s Office. The most recent document lists more than 30 barriers maintained by the United States, including tariff peaks on products of interest to the EC, agricultural import quotas, import monitoring, “Buy American” policies for machine tools, standards, testing, labeling, and certification requirements, government procurement practices, the Export Enhancement Program, tied-aid credits, and the U.S.-Japan semiconductor agreement. Also of concern to the EC are customs user fees, which the EC believes do not accurately represent user costs insofar as they are linked to the value of imports. The United States has agreed to modify its user fee system in response to a GATT panel ruling.

Major trade disputes and trade policy actions involving the United States have occurred in a number of sectors, including agriculture, electronics, and construction. These disputes center on the initiatives that the United States has taken to counter subsidies or other “unfair” trade practices abroad or to improve its access to foreign markets.

The United States signed a bilateral free trade agreement with Canada in January 1988, subject to ratification in both countries. The agreement provides for the gradual elimination of tariffs and quotas between the two countries over a ten-year transition period starting in January 1989 and for the reduction of barriers to trade in services (including banking services) and investment. In agriculture, the agreement additionally prohibits export subsidies on bilateral trade, including some transportation subsidies on Canadian exports to the United States. It also provides for a reciprocal opening of government procurement, reduction of technical barriers to trade, and introduction of a dispute settlement mechanism in which decisions will be binding. The two countries view various aspects of the agreement as a possible model for multilateral agreements in the Uruguay Round.

Although the impact of the U.S.-Canada Free Trade Agreement has not yet been systematically analyzed, it can be expected to provide substantial benefits to both countries as well as to the rest of the world. Earlier studies have estimated that the gains from specialization, competition, and the achievement of economies of scale could be substantial, particularly for Canada.32 A priori, it can be expected that the amount of trade diverted from third countries as a result of the elimination of tariffs between the United States and Canada would be small in proportion to total trade. The low tariffs in the United States reduce the competitive advantage that Canada would gain through duty-free access to the U.S. market. Although U.S. exporters would gain a more significant advantage on the Canadian market—which is protected by higher tariffs—the considerably lesser importance of the Canadian market in world trade reduces the scope for trade diversion. This scope is further limited by the large volume of trade between the two countries, which presently amounts to more than one third of their total trade. Although probably small in proportion to total trade, the amount of trade diverted could be significant in certain sectors (such as textiles) that are protected by relatively high tariffs in Canada as well as by quotas under the Multifiber Arrangement.33 This effect may nevertheless be mitigated by the intended reduction of tariffs on textiles in Canada. In addition, the real income gains in both countries resulting from the removal of trade barriers can be expected to benefit exporters in third countries. The effects on third countries of the simultaneous reduction in technical barriers as well as barriers to investment and trade in services are more difficult to assess because they are less transparent.

The United States also signed a bilateral framework agreement with Mexico in February 1988. The agreement is limited to establishing a bilateral consultation mechanism governing trade and investment relations without committing either party to trade liberalization measures. It consists of a framework of principles, procedures, and an agenda under which specific sectoral and other agreements are to be concluded at a later date. It also establishes procedures for mediation of bilateral trade and investment disputes and covers a number of sectors including steel, automobiles, textiles, and agriculture, as well as services, intellectual property rights, and trade-related investment measures. Bilateral agreements under the framework agreement reached with Mexico are expected to be negotiated after the U.S. presidential election in November 1988.

Japan

Since 1985, Japan has implemented a series of marketopening measures. A three-year Action Program was launched in July 1985, guided by the principle of “freedom in principle, restrictions only as exceptions.” The pace of implementation of the program was faster than planned, with most measures in place before the July 1988 deadline. Under the program, tariffs on a broad range of industrial and agricultural products were reduced by 20 percent on average and a number of measures were enacted to improve market access. Technical standards, testing, and certification requirements were eased, and government procurement practices were modified to make competitive tendering more extensive. Quantitative restrictions on imported leather and leather footwear were abolished in April 1986. Japan submitted a proposal in the Uruguay Round to abolish all tariffs on imports of almost all industrial products in industrial countries.

In 1988, Japan agreed to reduce trade barriers as a result of GATT panel investigations and bilateral negotiations outside GATT. In the agricultural sector, which has traditionally been heavily protected, Japan agreed to phase out quotas on eight out often products found to be inconsistent with GATT rules. Japan is implementing changes necessary after conclusion in November 1987 of a GATT panel investigation (initiated by the EC) that the different tax rates applied to liquor of different quality and grade favored local products. Separate bilateral discussions with the United States and Australia also resulted in Japan’s agreement to phase out its quotas on imports of beef and replace them with tariffs; and bilateral negotiations with the United States resulted in its agreement to phase out quotas on citrus fruits. Bilateral discussions have also been undertaken with the EC to ease barriers arising from standards, testing, and certification procedures in Japan, particularly on automobiles and pharmaceuticals.

Certain market-opening measures have also been taken as a result of bilateral discussions with the United States on sector-specific liberalization. These discussions covered electronics, telecommunications, pharmaceuticals and medical equipment, forestry products, and auto parts. Although conducted bilaterally with the United States, the talks are perceived as having improved access on Japan’s market for all exporters. However, the U.S.-Japan semiconductor agreement that was concluded in September 1986 outside this framework has given rise to concerns of discrimination against third countries (Section VI). Similar concerns have been expressed about the U.S.-Japan bilateral discussions on government procurement.

While the measures that Japan has undertaken in recent years to stimulate domestic demand and improve access to its market have helped to some extent to reduce protectionist pressures directed against Japan, foreign exporters nevertheless continue to question the openness of its market. “Visible” barriers to trade in industrial products in Japan, as measured by tariff rates and common types of nontariff measures, are among the lowest in industrial countries (Tables 2 and A8). Japan maintained 13 known voluntary export restraints as of May 1988 affecting mainly imports of textile products from China, Korea, and Pakistan. Allegations of “invisible” barriers are often based on attitudes or traditions rather than on legal or institutional barriers. Aside from standards, testing, and certification requirements that have been eased to some extent under the Action Program or as a result of bilateral discussions, Japan’s licensing system for some businesses and its distribution system are also considered by some countries to operate as barriers. Control over Japan’s distribution system by Japanese producers in some industries is perceived to be exercised through loyalty to long-established business relations and exclusive distribution arrangements. The incomplete pass-through to import prices of the yen appreciation since 1985 has been interpreted by some countries as evidence of price-fixing through lack of competition in the distribution system or through administrative guidance to importers. The EC has also complained about Japan’s indirect tax system, which applies higher taxes on large cars. Japan is presently reviewing its indirect tax system.34

With regard to Japan’s access to major industrial country markets, about two fifths of its exports to the United States and the EC are subject to some degree of restraint. These include voluntary export restraints on automobiles, electronic products, machine tools, and steel. Several were introduced in the early 1980s and were meant to be temporary but have been rolled over beyond their expiration date, in spite of the sharp deterioration in Japan’s competitiveness as a result of the appreciation of the yen after 1985.35

Voluntary export restraints and “administered protection,” including antidumping duties, may have influenced the pattern of foreign direct investment by Japan as exporters have sought to circumvent them by setting up operations in protected markets (Table A10). This is most evident in Japan’s direct investment in the machine tool, electronics, and automobile sectors in North America and Europe, to which an increasing proportion of Japan’s rapidly expanding direct investment abroad is directed (Table 4).

Table 4.

Japan: Shares in Stock of Foreign Direct Investment in Manufacturing

article image
Source: Based on data from Table A10 and from GATT (1988a).

Number of voluntary export restraints directed against Japan as of May 1988.

European Community

The European Community’s large agricultural surpluses and their effects on world prices have come under increasing criticism in recent years. Reforms introduced in the Community’s Common Agricultural Policy (CAP) since 1984 were motivated primarily by domestic budgetary considerations and are not perceived by some countries as having adequately addressed the underlying problems, in particular those of access to the EC market and subsidization of exports. A number of trade disputes in agriculture revolve around the operations of the CAP; the most recent relates to the imposition of import restrictions on apple imports (see Section IV and Appendix I). One third to one half of all agricultural imports in the major EC countries are covered by quotas and monitoring arrangements (Table A8).

The EC accounts for about half of the voluntary export restraints applied by industrial countries. The number applied on either an EC-wide basis or nationally doubled to 137 between September 1987 and April 1988.36 These restraints are increasingly directed against the imports of developing countries and cover mainly agriculture and food products, textiles and clothing (outside quotas under the MFA), steel, electronics, automobiles, and footwear. Japan is particularly affected by voluntary export restraints on automobiles and electronic products. Among the major EC countries, the share of industrial imports covered by voluntary export restraints and other nontariff measures in 1986 had risen to 15.4 percent in France, 17.9 percent in the Federal Republic of Germany, and 12.8 percent in the United Kingdom (Table A8). Within these totals, restricted imports of automobiles and electronics represented 3-4 percent in each country and were directed mainly against Japan.

In 1984 the EC adopted the New Commercial Policy Instrument intended to counter “unfair” trade practices abroad. Since then, the EC has introduced important changes in its legislation on antidumping that has broadened and sharpened the scope of existing rules and given rise to trade disputes with Japan and Korea (see Section IV). In addition, recent antidumping and countervailing duty actions by the EC have been directed against Brazil, Mexico, and the Eastern European countries.

The EC is implementing a broad-ranging program to reduce regulatory barriers and liberalize trade and factor movements within the Community. The program is expected to improve the EC’s international competitiveness (see Appendix I). With the abolition of internal borders in 1992, national restrictions on imports from third countries will no longer be enforceable. Pressures exist within the EC for adoption of the most restrictive national trade regimes in some sectors on a Communitywide basis after 1992, which are resisted by members holding more liberal views. The EC has noted the possibility of linking access to its integrated market to reciprocal concessions granted by trading partners either on a bilateral or a multilateral basis in the Uruguay Round.

Canada

Canada relies mainly on border measures to protect selected industries, including textiles and clothing, footwear, automobiles, and shipbuilding. Tariffs on textiles and clothing are at least twice the average for all industrial products, and the bilateral restraint arrangements reached under MFA IV are generally more restrictive than those under previous MFAs. Imports of certain categories of footwear are subject to global quotas, and export restraint arrangements covering categories not covered by the quota have been negotiated with Korea and Taiwan Province of China. Until recently, Canada maintained voluntary restraint arrangements with Japan and Korea limiting their exports of automobiles to the Canadian market. The voluntary export restraint with Japan was negotiated to prevent the diversion of Japanese exports to Canada following the U.S.-Japan voluntary export restraint on automobiles. Although these arrangements were not formally renewed after they expired, both Japan and Korea agreed to monitor their automobile exports to Canada in order to avoid disrupting the Canadian market. A few months after the expiration of the VER with Korea, Canada imposed provisional antidumping duties averaging 35 percent on imports of Korean automobiles. Although sectorspecific assistance to industry through nonborder measures has been de-emphasized in recent years, the Government continues to provide considerable support to the shipbuilding sector through subsidies and government procurement practices.

Among the barriers to trade that Canada faces abroad, protection to agriculture by foreign producers is the most important. Canada is a member of the Cairns Group37 and together with it attaches great importance to the liberalization of trade in agriculture in the Uruguay Round. Canada has also frequently been the target of antidumping and countervailing investigations initiated by the United States (Table A9) and has cited this as a barrier to trade.

Industrial Policies

Industrial policy is broadly defined as the deliberate attempt by a government to influence the composition of a nation’s industrial output. The definition encompasses all government actions to foster activity in specific sectors, either to shift resources to higher-productivity activities in support of adjustment objectives or to maintain resources in existing activities for security, political, or other reasons. A number of domestic and trade measures are used by governments to implement industrial policies. An indicative list of selected domestic measures would include subsidies (cash transfers, research and development funding, tax concessions, loan guarantees and insurance, subsidized credits, capital grants, regional aids); government procurement; national product standards; and commodity-specific indirect taxation. An indicative list of selected trade measures would include tariffs (peaks,38 escalation39), and nontariff measures (import quotas, voluntary export restraints (VERs),40 tariff quotas,41 discretionary and nondiscretionary import licensing, countervailing and antidumping investigations and duties,42 health standards, and export subsidies).

Rationale for Industrial Policies

Government assistance to industry has often been motivated by the desire to shape the industrial structure to serve developmental and political goals and/or to ease the burden of adjustment. Shaping the industrial structure has normally involved the promotion of sectors considered “strategic” (of special value) to the economy because they generate external economies or linkages to the rest of the economy (such as semiconductors), or because they are perceived as necessary to achieve market presence, or because they are for defense. Support to ease the burden of adjustment has tended to be concentrated in traditional industries that have been affected by an erosion of price competitiveness and/or to cushion the economic and social impact of external shocks. Throughout the 1970s and early 1980s, industrial countries made heavy use of subsidies to ease the effects of the oil shocks and to assist mature industries. Assistance to industry is at times conditional on specific commitments by the recipients. Following the oil shocks, subsidized loans and tax incentives were offered to companies that invested in energy-saving equipment or shifted to alternative sources of energy. Similarly, subsidies to mature industries have been linked to either job maintenance or to restructuring efforts involving cost-reducing investment, modernization, and capacity reductions to promote international competitiveness. Industrial policies may also be instituted as a defensive reaction to the adverse effects of the policies of others, for example, subsidies to counter the effects of foreign subsidies.

Approaches to Industrial Policies

National approaches to industrial policies differ widely across industrial countries. Industrial policies can take the form of informal administrative guidance to producers, as in Japan, which additionally provides tax preferences and credit subsidies on a limited scale. In the United States, which has no industrial policy as such, the role of government is generally limited to maintaining a stable macroeconomic environment and enforcing regulations aimed at promoting competition and innovation while protecting consumers. Assistance to industry is provided indirectly through tax deductions and support for defense-related research and development expenditures that may have technological spin-offs for industry. The European countries, by contrast, have made heavy use of assistance to industry through subsidies. These can be fiscal and financial incentives, including interest rate subsidies, tax preferences, and contributions to pension funds to promote investment, research and development, and regional objectives, or to assist small and medium-sized enterprises and ailing companies. In countries in which the government is involved directly in industrial production, as in France, assistance has occasionally been provided to cover the operating deficits of state-owned enterprises or to write off their debts. Australia and New Zealand have traditionally relied on trade protection rather than on subsidies to industry. Canada has recently moved away from sector-specific assistance to more general forms of assistance, including research and development support.

Trends in Subsidies

Assistance to industry provided through subsidies can have an impact on trade flows by distorting relative prices and resource allocation. It may thus constitute a substitute for border protection even though it is not always designed to do so. Information based on a broad definition of subsidies is not available. (A broad definition would include tax preferences, government procurement practices, national standards, and all other government actions that favor domestic over foreign producers.) This situation reflects the lack of consensus over the definition of subsidies with a trade-distorting effect, as well as difficulties in collecting the relevant information from state and local governments in countries with a decentralized administration, such as the United States and the Federal Republic of Germany. Recent efforts by the EC and the OECD to compile information on subsidies have additionally met with resistance by their members to provide sector-specific information that may trigger countervailing duty investigations.

The lack of information on the sectoral distribution of subsidies to industry makes it difficult to ascertain how much industrial policy is used as a substitute for trade protection.43 A few broad trends are nevertheless apparent in most industrial countries. A number of industrial countries have moved away from sector-specific government assistance toward research and development and regional support. Government subsidization increased in all Group of Seven countries in 1972-83 and in 1985 government assistance was higher than in 1972 in all countries except the United States (Table A11). In most countries, government aid to the manufacturing sector rose sharply and became more selective, largely in support of traditional sectors. Since 1982-83, however, some scaling down—or, at least stabilization—of sector-specific government aid does seem to have occurred in a number of OECD countries.44

In Germany the percentage of government subsidies to the manufacturing sector rose eightfold during 1974-84, and those to shipbuilding and steel, from 23 percent of total industrial subsidies in 1977 to 50 percent in 1983; a greater focus on the need to reduce these budgetary costs now exists. In the United Kingdom aid to industry doubled in 1976-81; steel, shipbuilding, and mining together received one fourth of total aid to industry in 1982-83, against 7.5 percent six years earlier. Again, the shift to a tighter budgetary policy and moves toward privatization are containing this trend. Recently, sector-specific subsidies have been reduced in some countries. In Canada, subsidies to the textile and clothing sectors are being phased out; in the EC, limits have been placed on subsidies to steel and shipyards.

However, the above trends may have been partly offset by an increase in less transparent subsidies that also assist specific sectors. In practice, regional assistance can be sector-specific where particular industries are heavily concentrated in one region, as with the coal industry in France, the Federal Republic of Germany, and the United Kingdom and the steel industry in France. Sector-specific assistance has also been increased through the use of border measures, particularly voluntary export restraints.

With regard to other domestic measures, tax preferences have been important in a number of countries. In Germany, tax concessions to enterprises by both federal and regional governments total some 40-50 percent of total subsidies (including such items as housing and consumer subsidies), while in the United States federal tax concessions to industry averaged 1.5 percent of GDP per annum in 1975-87. Recently, however, the tax system in the context of overall tax reform has become more neutral with respect to industry; this example is being followed by other countries, including the United Kingdom. Tax support in the United States has been provided mainly to encourage plant modernization and research and development expenditures. Support for the latter has been high in most developed countries (Table A12). Data on other forms of domestic support is difficult to quantify, or is not available. Considerable subsidies may be involved, however, in government procurement practices: total EC public procurement is estimated to contain a subsidy element of some 10 percent.45 Interest subsidies may also be involved in national product standards, which can be geared to the interest of domestic firms. Further, it is sometimes thought that companies in industrial countries create vested interests by transmitting their own national standards to developing countries through technical assistance.

Substantial progress has also been made in reducing the competitive subsidization of exports through officially supported export credits (Table A13). Nevertheless, the subsidy element in these credits remains high: for the OECD countries as a group it rose from 14.2 percent in 1979 to 27.5 percent in 1981 before declining to 12 percent in 1985. Information on other forms of export assistance is not readily available for a wide range of countries. While subsidized export credits have declined under the OECD Consensus Arrangement, subsidization through the mixing of aid with export credits may have become more widespread.46

The move away from subsidies is part of a broader trend away from government intervention as evidenced by deregulation, privatization, and tax reform in a number of industrial countries. The reduction in sector-specific subsidies reflects, in particular, three major considerations: first, they are increasingly subject to countervailing duties; second, their budgetary cost has placed limits on their use; and third, governments are increasingly recognizing that they may delay adjustment in mature industries. In EC countries this trend has been additionally motivated by the more strict enforcement of EC competition rules (see Appendix I). The first two considerations may have contributed to a substitution of voluntary export restraints for sector-specific subsidies: voluntary export restraints are not countervailable and their cost is borne by consumers rather than by the budget.

Causes and Costs of Protection

Protection reflects a government’s unwillingness or inability to undertake necessary structural adjustment or to withstand pressures for protection from vested interests. The arguments for protection, which have been refuted in a number of studies,47 tend to ignore its costs.

Traditional arguments for protection include the need to preserve or encourage mature industries (such as steel and shipbuilding), strategic sectors with linkages to the rest of the economy (such as high-technology industries), as well as sectors important for security and defense (such as coal in the Federal Republic of Germany and agriculture in Japan), and the need to accommodate the special characteristics of sectors such as farming. These arguments are advanced in terms of promoting the national interest, although protection often promotes sectoral interests at the expense of the rest of the economy. Industrial countries’ arguments for temporary assistance to mature industries to return them to competitiveness are akin to developing countries’ arguments for protection of infant industries. In practice, such assistance has often proved to be self-perpetuating and to spread to other areas through the rent-seeking behavior of interest groups that want similar treatment. Stockpiling of products that are important for defense would obviate the need for a high level of self-sufficiency produced at high cost, while income support delinked from production would accommodate the special characteristics of the farm sector.

As a considerable degree of adjustment of mature industries has occurred in the 1980s in industrial countries,48 other arguments have come to the fore. Protection is frequently motivated by the perceived lack of a “level playing field” (i.e., competition without government assistance and subject to the same rules), particularly against centrally planned economies (China and Eastern Europe), against Japan’s distribution system and other so-called invisible barriers, and against the newly industrializing economies because of the lack of reciprocity and perceived undervaluation of their exchange rates.

Persistent large external imbalances have given rise to the use of trade measures to counter macroeconomic disequilibria. Some market-opening discussions have taken the form of attempts to achieve a better balance in bilateral/sectoral trade. Additionally, the EC automobile industry has argued that access of Japan’s automobile exports to the integrated EC market should depend on the achievement of a specified EC share of Japan’s market. Attempts to balance sectoral/bilateral trade or, more generally, the use of trade measures to improve the current account ignore its fundamental determinants. Protection will not improve the current account unless it affects the savings-investment balance of the economy, as could happen with revenue-generating forms of protection such as tariffs or import licenses that are auctioned. The improved fiscal position might then improve the current account. However, the prevalent forms of protection in industrial countries either transfer the windfall gains to foreign exporters (voluntary export restraints and minimum price undertakings) or entail a budgetary cost (subsidies).

Certain forms of protection tend to insulate the protected sector from exchange rate movements, thereby slowing the macroeconomic adjustment process (as happens with trade “managed” through voluntary export restraints and quotas). Examples include Japanese import quotas on agricultural products, bilateral import restraints by the EC and the United States on automobiles, steel, textiles, and machine tools, by the EC and Canada on footwear and clothing, and by several industrial countries on agricultural products. A similar effect arises from subsidies and import duties that are designed to compensate for the difference between domestic and world prices. Examples include variable import levies and export subsidies under the EC’s Common Agricultural Policy and variable subsidies to the German coal industry. Indirectly, the same considerations apply to the shipbuilding sector in the EC, where subsidies depend, to some extent, on the difference between domestic costs and those of the most competitive world supplier. Along the same lines, the “dollar clause” proposed by the EC in the aircraft financing agreement under negotiation with the United States would partly insulate the Airbus consortium from exchange rate movements (Section VI). Some modifications being proposed to the producer subsidy equivalent (PSE) concept, which might be used in multilateral negotiations on agriculture, are intended to neutralize the effects of exchange rate changes, at least over certain periods. Moreover, the use of countervailing and antidumping duties as a safeguard measure in cases where exchange rate appreciation affects the outcome of the “injury” test has a similar effect. Sectors that are insulated to some degree account for 24.6 percent of agricultural imports and 12.5 percent of industrial imports of the G-5 countries. 49 These forms of protection are viewed partly as a response to exchange rate instability. Greater exchange rate stability among the major currencies is therefore viewed as promoting more open markets. However, by insulating these sectors from exchange rate movements, protection shifts the burden of adjustment to other sectors and may contribute to larger exchange rate fluctuations than might otherwise occur.

Within the context of the Uruguay Round, industrial countries argue that protection cannot be reduced unless all countries agree to liberalize together. This argument applies particularly to trade in agriculture but has been advanced in connection with all trade, including steel and services. Indicative of this reasoning are the discussions pursued by the EC to obtain reciprocal concessions from trading partners in exchange for access to its integrated internal market (Appendix I).

These arguments ignore the costs of protection and the benefits of unilateral liberalization. The costs of protection have been extensively analyzed in the economic literature.50It is widely recognized that protection imposes costs both on the country initiating it and on its trading partners. Any measure that restricts imports also restricts exports by shifting resources to the import-competing sector. Similarly, subsidies and other non-border measures targeted at specific industries necessarily divert resources from other industries, thereby “taxing” the rest of the economy. Protection can also involve direct budgetary costs or indirect costs through forgone tariff revenue. Protection entails costs owing to forgone specialization according to comparative advantage, as well as losses in terms of scale economies, product differentiation, and research and development efficiency. Additional costs are incurred because scarce resources are directed toward rent-seeking activities, and the enforcement of restrictions imposes administrative costs. By releasing resources for efficient industries, unilateral liberalization can increase potential growth and ease the external constraint.51

Protection provided through nontariff measures tends to be highly selective, favoring a few domestic industries. Nontariff measures compound relative price distortions arising from the dispersion in tariff rates (tariff peaks). A study of the dispersion of protection in German industry found that the coefficient of variation52 of nominal protection increases from 0.4 for tariff protection to 1.0 for total protection including NTMs (Table A14). The coefficient of variation of total effective protection, which includes the effects of the escalation of tariff and nontariff protection on products at higher stages of processing, is calculated at 2.0.

The cost of voluntary export restraints is high for the markets they are intended to protect because of “quota rents” (normally captured by the exporting country) and distortion costs. Distortion costs arise because producers produce goods at marginal costs that do not reflect their opportunity costs and because consumers purchase goods at prices that do not reflect their scarcity values. By contrast, production subsidies—the first best instrument for raising sectoral output—do not involve the costs arising from consumption distortions. The recent trend toward voluntary export restraints, despite their higher cost to the economy per unit of the protected item, reflects the two advantages they have over sector-specific subsidies that were noted above, that is, they are not countervailable and their cost is less transparent because it is borne by consumers rather than by the budget.

The cost of U.S. voluntary export restraints on imports of automobiles, steel, and textiles has been estimated at $21 billion, of which the quota rent amounts to $14 billion and the distortion costs of the quotas, $7 billion.53 The cost of “preserving” a job is estimated at eight times the average annual wage in the textile sector and at three times in the steel sector. Starting from the existing tariff dispersion, the economy-wide tariff equivalent of the quotas on these three sectors is estimated at 25 percent, bringing protection to its level of the early postwar years. Similar studies of the costs of the Multifiber Arrangement have estimated the quota rent transferred to the Asian newly industrializing economies by OECD countries at $2 billion.54 The costs of voluntary export restraints on automobiles maintained by the EC and Canada have similarly been found to be very high (Section VI).