III Antidumping: Solution or Problem in the 1990s?

Peter Uimonen, Arvind Subramanian, Naheed Kirmani, Nur Calika, Michael Leidy, and Richard Harmsen
Published Date:
February 1995
  • ShareShare
Show Summary Details
Michael Leidy1

Antidumping is by far the most frequently used (GATT-legal) instrument of administered or contingent protection2 among industrial countries, and its use has been spreading in recent years with developing countries, and some transition countries, taking an increasing interest in formal antidumping measures. Whether antidumping is a problem or a solution in the multilateral trading system depends on what countries aim to achieve through antidumping policies, whether it is an appropriate instrument to achieve these objectives, and whether the benefits appear to justify the implied social costs.

Antidumping as currently practiced has important anticompetitive effects. While some early antidumping laws, in the United States for example, were viewed as extensions of competition (antitrust) policies and were designed to preserve the conditions of competition, no such standard is applicable to antidumping today (Box 1). Indeed, under current rules and practice, antidumping protects a privileged class of competitor (i.e., import-competing firms) to the detriment of competition.

This paper examines the frequency, incidence, and country composition of antidumping measures, focusing on the four years since July 1989. It reviews certain theoretical aspects of antidumping policy, including a critical evaluation of the most commonly cited rationales for antidumping and a survey of recent work on its trade-distorting effects. The principal changes to antidumping rules under the Uruguay Round are also discussed. The paper concludes with a brief description of reforms that would help to reconcile antidumping policy with the vision of liberal world trade that underlies the multilateral trading system.

Recent Antidumping Activity

The last four years have seen virtually every indicator of aggregate antidumping activity rising, with yearly initiations and new definitive measures (i.e., antidumping duties and price undertakings) recently surpassing the previous peak in activity seen in 1984–85. Several countries used existing antidumping laws for the first time, and others formally instituted new antidumping regimes—thus enhancing the prospect of a further spread of antidumping activity in the future. Traditional users of antidumping also saw an increase in this activity.

Much of the recent rise in activity is attributable to the macroeconomic slowdown as well as the sweeping antidumping initiations against various steel products. Regardless, the overall growth in this activity is an important indicator of the recent stance of international trade policies, and highlights the importance of Uruguay Round efforts to introduce new disciplines in this area.

Antidumping Activity Among Traditional Users

The major users of antidumping over the last decade were Australia, Canada, the European Union (EU), and the United States (Chart 1). Chart 2 shows the stock of antidumping measures in force, initiations, and new measures taken by each of these countries separately since 1989–90. Total initiations among the major users rose from 77 a year in 1989–90 to 202 in 1992–93 (Chart 2). New definitive antidumping duties increased from 44 in 1989–90 to 105 in 1992–93. The stock of antidumping measures in force rose from 470 in 1989–90 to 573 in 1992–93.

Chart 1.Antidumping Activity1

Source: GATT, Basic Instruments and Selected Documents, various issues.

1Before 1991/92, European Union figures exclude initiations and duties on nonsignatories to the GATT Antidumping Code.

Chart 2.Antidumping Activity: Australia, Canada, European Union, and the United States1

Source: GATT, Basic Instruments and Selected Documents, various issues.

1European Union figures prior to 1991/92 exclude initiations and duties on nonsignatories to GATT Antidumping Code.

While the stock of outstanding antidumping measures in Australia had been generally high and increasing since 1980, from 1986 through 1989 there was a sharp decline in outstanding measures—the stock was cut by about 75 percent during this period.3 From 1990–93, however, measures in force more than tripled. In recent years, initiations and newly imposed definitive duties in Australia have returned to the high levels of the early-to-mid-1980s.

Continuing the steady decline since 1988, Canada’s stock of outstanding antidumping measures fell from 103 in 1990 to 71 in 1991, but rose to 80 by June 1993. Canada’s yearly antidumping initiations and definitive duties had also been moving somewhat downward since the early 1980s, but have turned upward since 1990–91.

Box 1.Antidumping: Procedures and Definitions

The antidumping process is usually initiated with a petition from an import-competing industry claiming dumping—alternatively, the authorities may initiate the process. This sets in motion an investigation lasting typically many months to determine whether two fundamental criteria have been met: (1) dumping has occurred; and (2) the import-competing industry is suffering “material” injury, or the threat of material injury, by reason of the dumped imports.

Dumping exists when a product is sold abroad (introduced into the commerce of another country) at less than “normal value,” that is, the export price is less than the comparable price in the home market, in the ordinary course of trade. When sales in the home market are deemed insufficient to permit a proper comparison, the margin of dumping is determined either by looking at a comparable price of a like product exported to an “appropriate” third country, or by estimating the cost of production in the country of origin plus administrative, selling, other costs, and a margin for profit.

Establishing material injury is not a precise exercise and, as such, final determinations are generally regarded as subject to some administrative discretion. The Uruguay Round agreement attempted to clarify some of the rules associated with determining injury but, by the nature of the process, it could not have hoped to succeed in establishing a one-to-one relationship between industry conditions and the outcome of a material injury investigation. Some amount of discretion is inherent. Factors entering into the assessment of injury include, inter alia, industry sales, output, profit, market share, capacity utilization, employment, wages, return on investments, factors affecting domestic prices, the margin of dumping, actual and potential negative effects on cash flow, and productivity. It is also noteworthy that the “material injury” standard is generally thought to be less stringent than the “serious injury” test applicable to a safeguards action under Article XIX of GATT/WTO.1

If a preliminary investigation shows that both injury and dumping are present, preliminary measures are taken (typically, duties, or the posting of a cash deposit or bond not to exceed the amount of the preliminary dumping margin). The Uruguay Round agreement specifies that preliminary measures shall not be taken sooner than 60 days from the date of initiation of the investigation. Sometimes investigations are suspended or terminated without the imposition of provisional measures in response to a voluntary undertaking by the exporter to revise its price upward (not to exceed the dumping margin). Such actions are referred to as “price undertakings.” The Uruguay Round indicates that such price undertakings shall not be sought or accepted before the authorities have made a preliminary determination of dumping and injury. If the preliminary investigation finds dumping and injury, and if no price undertaking is agreed, the investigation proceeds to its conclusion where a final determination on the question of dumping and injury is reached. If both are affirmative, a final antidumping duty generally equal to the estimated dumping margin is assessed. Australia, Canada, and the United States, for example, typically impose duties equal to the estimated dumping margin. The European Union has a “lesser-duty rule”; that is, if a duty less than the margin of dumping would be sufficient to remove the injury, that lesser duty is assessed. But some have argued that the lesser-duty rule has led the European Commission to substitute “injury margins determined through a non-transparent process characterized by a high degree of discretion, for realistic dumping margins based on fair comparisons” (Bellis (1989, p. 5)). The Commission estimates that between January 1991 and the end of June 1992 about half of all cases in which final duties were imposed saw duties less than the full dumping margin (GATT (1993, p. 65)).

1See Jackson (1989), pp. 236-37.

Since 1983, the stock of antidumping measures in force in the United States has increased steadily, from a reported 52 in 1983 to 279 by the end of June 1993. Yearly initiations fluctuated fairly widely over the 1980s but rose steadily from 24 in 1989–90 to 74 in 1992–93. Definitive duties and price undertakings imposed yearly were roughly stable from 1989–92, then rose dramatically during the period June 1992-June 1993. This surge in antidumping measures was due in part to the numerous steel-related antidumping petitions following the end-of-March 1992 expiration of voluntary export restraint agreements with 27 steel-exporting countries.

The European Union began reporting antidumping measures in force against all countries only in 1989. Since that time, aggregate measures in force have remained relatively stable. Prior to 1991–92, the EU was reporting antidumping activity only toward members of the Tokyo Round Antidumping Code. Thus the increase in antidumping initiations from 1990–91 to 1991–92 must be interpreted with care, as it reflects a number of actions against countries not previously included in the reports. Nevertheless, antidumping initiations and definitive duties (or price undertakings) increased in the year ended June 1993.

Much of the recent rise in antidumping investigations in industrial countries was directed toward non-OECD countries (Chart 3). During the period July 1, 1991 to June 30, 1992, the proportion of investigations involving non-OECD exporters was 66 percent, 51 percent, and 83 percent, respectively for the United States, Australia, and the EU.4 From mid-1991 through mid-1993, 47 percent of all U.S. antidumping cases were directed toward industrial countries, 38 percent toward developing countries, and 15 percent toward economies in transition.5 In Australia, outstanding antidumping actions were mostly concentrated against Asian countries, with actions against EU member states the second most prevalent.6 Australian antidumping actions against Central and South American States were the third most frequent, exceeding those against the United States and Canada by a small margin.

Chart 3.Distribution of Antidumping Cases

(Outstanding cases, end of June 1993)

Source: GATT.

While a vast array of products have been subjected to antidumping actions, the bulk tend to be concentrated in a few sectors. In the EU, antidumping duties in effect in June 1993 were mostly concentrated in chemicals, metals, machinery, textiles, and consumer electronics industries. In the United States, various metals, cement, electrical and nonelectrical machinery, and industrial chemicals saw the highest concentration of outstanding antidumping duties. In Australia, antidumping measures were concentrated in chemicals, machinery and parts, and processed agricultural goods.

Antidumping Activity Among Nontraditional Users

Beginning in the second half of 1990, aggregate antidumping initiations and definitive measures over a block of 14 users started a strong upward climb that has continued through 1993 (Chart 4).7 This contrasts with the period 1984–90 when total antidumping initiations, and to a lesser extent definitive measures, had a downward trend. However, 1984–85 was also the previous peak in global antidumping activity, and by 1991–92 that peak had been surpassed.

Chart 4.Aggregate Antidumping Activity, Fourteen Users1

Source: GATT, Basic Instruments and Selected Documents, various issues.

1The 14 users are Australia, Austria, Brazil, Canada, European Union, Finland, India, Japan, Korea, Mexico, New Zealand, Poland, Sweden, and the United States.

Developing countries and other nontraditional users have begun taking an increasing interest in applying formal antidumping measures of their own.8 Mexico, for example, had a total of 62 antidumping initiations from mid-1990 through mid-1993. This was about 10 percent of the total9 and almost three times the number of initiations during the previous three years. Together, India and Korea had 17 initiations (8 and 9, respectively) and 4 definitive measures (all in Korea) from mid-1990 through mid-1993. While this was up from previous years—indeed, India saw its first antidumping initiations in 1991–92—these were rather insignificant against the actions of the major users. Nevertheless, these figures are indicative of the tendency for formal antidumping actions to proliferate among a growing number of countries.

A concentration index of antidumping activity (Chart 5) highlights the spread of antidumping measures beyond the big four since the mid-1980s. During the early to mid-1980s the four traditional users of antidumping undertook virtually every formal antidumping action. From the late 1980s to the present, nontraditional users have consistently held a share of about 20 percent to 30 percent of total yearly antidumping actions (initiations and definitive duties) reported to the GATT.

Chart 5.Concentration of Antidumping Measures in the European Union, the United States, Canada, and Australia, as Percentage of Fourteen User Total1

Source: GATT, Basic Instruments and Selected Documents, various issues.

1The 14 users are Australia, Austria, Brazil, Canada, European Union, Finland, India, Japan, Korea, Mexico, New Zealand, Poland, Sweden, and the United States.

Several factors have contributed to the recent spread of antidumping policy to nontraditional users. First, the effectiveness of antidumping as a tool of selective protection from foreign competition was demonstrated by the traditional users throughout the 1980s, and is now being emulated by nontraditional users. Second, recent trade liberalization in some of these countries (e.g., Brazil, Mexico, and Korea) has put increased pressure on governments to respond to appeals for protection, and antidumping policies help to vent these protectionist pressures. Third, many developing countries have implemented for years measures such as discretionary import surcharges, reference prices, or minimum import prices, often in response to allegations of dumping. The adoption of formal antidumping regimes effectively enables developing countries to continue the protection afforded by these other measures while using an internationally sanctioned (GATT/WTO-legal) trade remedy.

How Much Protection Is at Stake?

A review of the most recent antidumping measures shows a wide range of duties. At the high end, in the United States, for example, recent definitive duties on ferro-silicon from China, Kazakhstan, Russia, and Ukraine were all in excess of 100 percent; steel wire rope from Mexico was at 112 percent; and professional power tools from Japan faced duties in the neighborhood of 50 percent. Duties of less than 10 percent, however, were also fairly common. Chart 6 gives an indication of the frequency distribution of definitive antidumping duties for the United States, the EU, and Mexico, based on a sample of recent actions.10

Chart 6.How Much Protection? Range and Relative Frequency of Final Antidumping Duties1

Source: GATT, semi-annual reports to Committee on Antidumping Practices, July 1, 1992-June 1993.

1When a range of duties was reported for a single case, a simple average was used and treated as one observation.

As Chart 6 shows, it is not uncommon for antidumping duties to be quite high. Like ordinary tariffs, even a small antidumping duty applied to a relatively homogeneous good can have a significant effect on imports. In addition, it is noteworthy that antidumping duties may have a somewhat stronger trade-inhibiting effect than ordinary tariffs, other things equal, because antidumping duties may be increased retroactively. When a duty is assessed (e.g., in the United States), the importer pays, in effect, an estimate (the announced antidumping duty) of the final duty on every unit imported.11 Subject to a request for an annual administrative review by an interested party, the investigating authorities may reassess applicable duties and apply them to past imports. Importing under an antidumping duty thus exposes importers to an open-ended contingent liability.12 The greater the uncertainty in calculating dumping margins—which can be considerable—and the more risk averse importers are, the greater is the trade-inhibiting impact of a given antidumping duty.

Rationales for Antidumping

Antidumping duties, like other trade taxes, generally imply an efficiency loss for the imposing country and for the world as a whole. In addition, as will be discussed in the next section, the distortions introduced by an antidumping regime can be expected to go well beyond the static efficiency losses underlying outstanding antidumping duties. Whether recent developments in antidumping warrant concern depends on whether countries individually or collectively are better off with such rules in place than without them. Perhaps the costs of antidumping are more than offset by the benefits of antidumping policies. Whether this might be true depends on the rationale for antidumping.

The rationale is often stated in terms of countering an “unfair” trading practice. Indeed, antidumping and countervailing duty laws are commonly referred to as “unfair trade laws,” as distinguished from safeguards policies (GATT Article XIX), which are intended to deal with disruption due to “fairly” traded goods. Examples of alleged unfair practices include pricing below cost, international price discrimination, and predatory pricing. Concrete rationales for antidumping policies seek to identify a trade-off between the economic losses caused by temporary trade restrictions and longer-term economic gains. Three arguments are typically made in support of antidumping policies. The first asserts that dumping may be predatory, and antidumping may be needed to counter the anticompetitive effects of predation. The second points out that dumping may be transitory and that antidumping may help to avoid the costs of temporary structural adjustments induced by unsustainably low prices. The third, and probably most tenable argument for antidumping policies, suggests that antidumping may be needed to help move the political process forward toward greater multilateral trade liberalization, by acting as a safety valve to vent protectionist opposition to liberalization. The merits of these rationales are considered in turn.


The theory of predatory pricing, whereby a firm or cartel intentionally prices below unit cost until it drives rivals from the market and establishes itself as a monopoly, underlies certain prohibitions against price discrimination under competition laws.13 Indeed, the United States Antidumping Act of 1916 was originally viewed as an extension of the Clayton Antitrust Act and explicitly included a requirement that “intent to injure” or “intent to restrain” competition be established as a precondition to granting antidumping duties.14 These clauses were dropped in the Antidumping Act of 1921 and the linkage of antidumping to antitrust ended there. Even so, although evidence of predation or predatory intent is not required in modern antidumping legislation, press accounts and popular defenders of antidumping policies continue to use the rhetoric of predation, and implicitly the rationale of competition policies, to support antidumping actions. Is antidumping necessary to counter potential instances of predatory dumping? Several points can be made.

First, the trading system may require antidumping laws because domestic competition laws, in their current form, cannot adequately address instances of foreign predatory pricing. But if this were the objective, antidumping laws should at least make some effort to differentiate between predatory and nonpredatory dumping. The standard of injury should be based on actual or threatened injury to competition. This would place antidumping more on par with competition policies, where the goal is to protect the conditions of competition rather than the interests of selected competitors (i.e., import-competing producers).

Second, there is considerable doubt about the viability of a strategy of predation. For a strategy of predation to work, a firm must be able to sustain monopoly profits after driving competitors from the market. If a market remains contestable, the predator will be unable to set a monopoly price and, thus, be unable to recover the losses incurred during the interval of predation.15 Further, dumping cases typically target foreign firms from a number of countries. Even if domestic firms were driven from the market, competition among a multicountry set of foreign rivals would be sufficient to maintain competitive prices. For these reasons, “economists have routinely dismissed predatory dumping as so unlikely that it should not be used to justify antidumping duties.”16

Finally, the empirical evidence on predatory pricing suggests that it is at best a rare event.17 Even laboratory experimental work whose designed institutions were intended to be conducive to a strategy of predatory pricing found no evidence of such practices.18

Whether or not there is a reasonable prospect of predation (and economists do not uniformly reject the possibility),19 if antidumping policies are to guard against predatory behavior, antidumping rules should move toward competition policy criteria.20 This would involve, inter alia, substituting the current injury standard for one based on injury to competition. Alternatively, there is no reason in principle why antidumping should not be eliminated altogether and replaced with suitably amended competition policies. Several recent events are illustrative. First, as part of the Australia New Zealand Closer Economic Relations Agreement (ANZCERTA) both signatories agreed to replace antidumping with competition law provisions in bilateral trade. Second, EU and European Free Trade Association (EFTA) members recently agreed to no longer apply antidumping actions against each other. The EU’s competition rules will now apply to EFTA countries in interregional trade.21 Third, with a view to moving closer to a competition policy framework, the Canada-United States Free Trade Agreement (CUSFTA) contain a provision calling for the Parties to work toward the replacement of antidumping with a substitute system of rules. This intention was recently carried over to the North American Free Trade Area (NAFTA) at the suggestion of Canada.22

Sporadic or Transitory Dumping

If dumping is temporary, say, the consequence of a sudden change in exchange rates or a cyclical decline in the exporter’s domestic demand, then resource-allocation adjustments in response to this relative price change may be wasteful. Once the transitory dumping has passed, exporters would return to a sustainable price, and resources responding to current price signals would return to their predumping allocation. Antidumping policies might be justified as a means to avoid the costs of such circular structural adjustments.

If such a framework were to justify antidumping, antidumping rules should reflect the intent to avoid circular structural adjustment. This suggests that antidumping should not be used, for example, to counter long-term dumping. Yet it is well known that under existing rules antidumping duties sometimes remain in place for ten years or more (see below). Further, “safeguards” provisions are already available to avoid significant temporary shocks from foreign competition. The reason industries petition for antidumping rather than safeguards (under Article XIX of GATT) is that the injury standard applicable to safeguards protection is generally regarded as imposing a higher threshold of injury, antidumping can be discriminatory (i.e., not applied on a most-favored-nation basis),23 the politics of antidumping are substantially more favorable to the interests of import-competing firms,24 and, unlike safeguards, antidumping has neither a “compensation requirement” nor the related possibility of retaliation.25 Because of this, antidumping has become a preferred substitute for safeguards actions.26 If antidumping is really about safeguards, antidumping policies need to be reformulated to reflect that objective (discussed further below).

Antidumping as a Safety Valve

It has also been argued that safeguards, countervailing duties, and antidumping provisions are needed to help defuse protectionist opposition to trade liberalization. In this view, an offer to accommodate appeals for protection through administrative channels may ease opposition to broad-based liberalization efforts.

Still, if one accepts the proposition that a safety valve is needed, the difficulty remains in striking the appropriate balance. When broad-based trade liberalization is purchased by granting liberal access to protection by other means (such as antidumping), the risk is that the price will be too high; that one form of protection will simply be substituted for another. The frequency of antidumping actions among the four major users and the recent spread of antidumping regimes to developing countries suggest that the cost of antidumping is not inconsequential, even when measured against the benefits of the incremental acceleration of multilateral trade liberalization it may have helped to induce.

If antidumping is a tool to defuse protectionist opposition to liberalization, one should question whether, as currently designed and implemented, it is the best tool, and whether there may not be alternative policies to achieve that end at a lower social cost. Indeed, the safeguards or “escape clause” of the GATT was itself intended largely to satisfy the need for a protectionist safety valve.27 But antidumping (and other measures, such as VERs) have become substitutes for the safeguards instrument. Further, if antidumping is primarily a tool to mitigate protectionist pressures, it should not be draped in the fabric of “unfairness,” which may serve to legitimize and deepen protectionist pressures. Part of the attractiveness of antidumping as an instrument of contingent protection is the public relations value of combating “unfair” traders.28 A pure safeguards approach would eliminate any such association.

Incentive Effects of Antidumping Policies

Proponents of antidumping argue that since duties apply only to a relatively small share of a country’s total imports, antidumping policies have inconsequential economy-wide effects. But the economic cost of antidumping policies cannot be evaluated by simply looking at outstanding antidumping duties. Antidumping is not merely a set of duties, it is a system of rules. Like other systems of rules, it generates incentives: deterrent effects, opportunities for manipulation, “out-of-court settlements,” and the like. Judging the full economic costs of antidumping requires close attention to all of this. An emerging literature explores the complex incentive effects of various forms of administered protection. The mere threat or prospect of antidumping may be sufficient to introduce systemic market distortions, including strategic export restraint, significant changes in the nature of competition, and spurious injury.29

Bhagwati (1988), Messerlin (1989), and others have argued for some time that antidumping produces “harassment effects” on exporter behavior. Indeed, Bhagwati suggests that, “the dramatic rise of such unfair-trade cases is itself prima facie evidence of their use for harassment of successful foreign suppliers” (Bhagwati (1988), p. 48). Petitioning for antidumping, or merely threatening to petition, can act as a means of signaling to foreign competitors to restrain their sales. In a related vein, U.S. auto producers announced in January 1993 that they intended to file comprehensive antidumping petitions against vehicle imports. The next month, three Japanese auto producers announced price increases for vehicles sold in the United States, and no antidumping petitions were ultimately filed by U.S. manufacturers.30

Several studies have formalized this harassment effect. In a setting of exchange rate uncertainty, where either intentional or accidental dumping may occur, Leidy and Hoekman (1990) showed that the optimal management of an antidumping threat called for shifting sales from the foreign to the home market. That is, the mere prospect of an antidumping action (a prospect embodied in the antidumping regime) was sufficient to induce an exporter to compete less vigorously than otherwise. Similarly, in a setting without uncertainty and somewhat different market characteristics, Webb (1992) also found that the introduction of an antidumping regime (as opposed to an antidumping duty) was sufficient to induce exporters to cut exports, while import-competing firms would experience higher profits and sales. Staiger and Wolak (1992) examined the effect of antidumping laws on a monopoly exporter that faces cyclical demand for its product at home and a competitive market abroad. They found that the threat of antidumping generally induces the exporter to reduce capacity and thereby exports, for most realizations of domestic demand. These studies show that an antidumping system may induce systemic trade-inhibiting effects in addition to the transparent effects of outstanding antidumping duties. The institutionalization of an antidumping regime may thus create a significant deterrent to vigorous foreign competition.

Many antidumping petitions are withdrawn before a final determination is reached because they are settled by voluntary export restraint agreements.31 This can be likened to an out-of-court settlement in civil suits.32 Since an antidumping petition can lead to a government-enforced VER, and under certain conditions such a VER may be highly favorable to an exporter, Anderson (1992) and Schuknecht and Stephan (1994) showed that the threat of antidumping combined with the prospect of a VER may induce an exporter to expand exports and possibly to “dump.”33Yano (1989) and Hoekman and Leidy (1990) described similar incentives to expand exports induced by the prospect or threat of VERs without directly linking them to antidumping. This perverse incentive effect suggests that systems of antidumping combined with the practice of granting VERs, particularly in the United States and the EU, may have contributed unintentionally to the widespread use of both VERs and antidumping cases in the 1980s.

Still other studies have shown that the structure of competition can be significantly altered by the existence of an antidumping regime. Staiger and Wolak (1989) and (1994), Fischer (1992), Leidy (1994), and Prusa (1994) have all shown that an antidumping regime can create conditions favorable to collusive outcomes. Staiger and Wolak (1989) and (1994) showed that the threat of antidumping, made credible by occasional petitions (but not necessarily resulting in actual duties), can support more successful collusion and greater market share for domestic firms by providing a means of punishing incipient defections from a tacit collusive agreement. Prusa (1994) and Leidy (1994) investigated the distortions introduced by unfair trade laws (antidumping and countervail) in a duopoly setting. Bertrand competition34 in the presence of antidumping was shown to introduce an effective price floor above the equilibrium price that would prevail, other things equal, in the absence of an antidumping regime.35 Under Cournot competition,36 the incentive to manage the prospect or threat of antidumping or countervail produced industry-wide restraint of trade characteristic of a cartel.37 Moreover, the de facto cartel-like equilibrium was shown to be stable, in the sense that it was free of the usual incentives to defect. This can occur since an import-competing firm has an incentive to enhance the prospect of winning an antidumping action by cutting sales, employment, capacity, and other relevant indicators of injury,38 while the exporting firm is motivated to exercise export restraint as it attempts to mitigate the antidumping threat. The incentive for joint restraint of trade moves the market outcome in the direction of the collusive solution, even though both firms are acting uncooperatively. Fischer (1992) also reports trade-inhibiting results across a broad set of instruments of contingent protection including antidumping.

The above-mentioned literature thus raises concerns about the possible unanticipated systemic consequences of “unfair” trade laws. The following section reviews the changes in the multilateral rules governing antidumping under the Uruguay Round.

Results of the Uruguay Round

The Uruguay Round negotiations on antidumping rules were divisive. Some industrial (e.g., Japan) and developing countries argued for strong disciplines to curtail future use of this instrument. Other industrial countries (e.g., the United States and the EU) resisted weakening of domestic laws, but rather sought stronger disciplines against attempts to circumvent antidumping actions. The final compromises reached helped clarify rules and in some respects strengthen them, but they also codified many existing practices, thereby probably diminishing prospects for any significant curtailment of future use of the antidumping instrument. The following review of the Uruguay Round agreement on antidumping focuses on selected areas of potential significance.39

De Minimis Provisions

There are two de minimis provisions (dumping margin and market share) in Article 5.8 of the Uruguay Round Agreement on antidumping. The first de minimis provision requires the immediate termination of an antidumping investigation when it is determined that the margin of dumping is less than 2 percent. The Tokyo Round Antidumping Code specified that a case should be terminated immediately where “the margin of dumping … is negligible” (Article 5.3).40 The difference is that a “negligible” margin was not defined under the Code. In practice, provisional dumping margins of less than 2 percent have been scarce in most countries. Established practice in the EU and Canada already considered dumping margins of less than 1.5 percent de minimis. Australia, while having no formal de minimis margin rule, has on several occasions treated margins in the neighborhood of 5 percent as too small to impose duties.41

In the United States, however, an estimated dumping margin at or above 0.5 percent42 is treated as sufficient to proceed beyond the preliminary investigation.43 By explicitly specifying a de minimis threshold of 2 percent, the Uruguay Round agreement might help to reduce at the margin the capacity of import-competing firms to use antidumping as an instrument to harass foreign competitors.

The second de minimis provision requires the termination of an antidumping investigation when the volume of dumped imports is “negligible.” This is defined to mean “normally” when the volume of imports from a particular country accounts for less than 3 percent of total imports of the like product, unless collectively such countries account for more than 7 percent of imports. Again, this is distinguished from the requirements of the Tokyo Round Code by the explicit definition. If, for example, imports account for 30 percent of the market, only exporters with less than a 0.9 percent market share will be dropped from the investigation, unless there are, say, three such exporters who collectively hold over 2.1 percent of the domestic market. The United States and the EU have commonly dismissed cases when imports were less than 1 percent of the domestic market, and Australia has dismissed cases with imports as much as 10 percent.44 Thus, while there may be some economic effects of the new de minimis import volume rule, they are unlikely to be appreciable.

Sunset Provision

Article 9 of the Tokyo Round Code on Antidumping required that an antidumping duty remain in place only as long as necessary to counteract dumping. It required the authorities to review extant duties “where warranted” on their own initiative or on the request of an interested party who also submitted information supporting the need for a review. Under current practices (e.g., in the United States), an interested party can initiate a review after one year,45 and an antidumping action will be revoked if a party can persuade the authorities that sufficiently changed circumstances warrant it.46 This approach places the burden of petitioning for revocation on an importer or an exporter, who may be deterred by the nontrivial transactions costs of presenting a case for review. In practice it has not been unusual for antidumping duties to remain in place for a decade or more.

The Uruguay Round agreement will now require the termination of antidumping duties not later than five years from imposition, unless a review is requested and determines that the expiration of the duty would likely lead to continuation or recurrence of dumping and injury. Thus the multilateral rule governing the duration of antidumping duties has gone from “do nothing and a duty continues,” to “do nothing and a duty expires.” The new sunset provision may help to ensure that outstanding antidumping duties are not viewed as virtual long-term entitlements by import-competing industries.

Among the principal users of antidumping, Canada, the EU, and Australia already have sunset clauses that conform to the Uruguay Round provision, under which antidumping duties expire automatically after five years.47 As stipulated in the new Uruguay Round agreement, each of these countries also provides for a review of measures upon request by an interested party. The recent experience in Australia might appear to suggest a rather definitive effect of such sunset provisions on the longevity of antidumping measures. As of June 30, 1993, Australia had no outstanding antidumping measures with definitive duties imposed prior to October 1990.48 In the EU, however, only about three fourths of all antidumping measures have expired on schedule since 1985;49 of the 16 reviews completed between January 1990 and June 1992, 12 resulted in new measures. In Canada, among the antidumping duties in force on June 30, 1993, 11 of 74 cases had original dates of finding earlier than 1984. In the United States, 61 of 268 duties in force at mid-1993 had original order dates prior to 1984, with a number of these dating back to the 1960s.

The new sunset clause means that longstanding actions may now be contested at the multilateral level, rather than simply at the national level. Sunset clauses can be expected to result in the automatic lapse of those measures deemed of insufficient value to the beneficiary industry to warrant petitioning for a review. On the other hand, when an antidumping action has a relatively significant protective effect—and thus is highly distortionary—bearing the transactions costs of a review will tend to be in the interest of the import-competing industry, and such measures may not be allowed to lapse automatically.

Transparency Provisions

When compared with the Uruguay Round agreement, the Tokyo Round Antidumping Code included rather casual transparency provisions.50 Even so, a fairly high degree of transparency exists in the systems of the four major users. The Uruguay Round agreement calls formally for public notice at every stage of an antidumping investigation51 and specifies the parameters of such notice, thus potentially alerting interested parties (including consumer groups and user industries) in a timely fashion of antidumping proceedings as they progress. This may help to enhance transparency in the antidumping process and potentially bring a better balance among different interests. It may also serve to enhance the flow of information in support of consultations and multilateral dispute settlement. But the Uruguay Round agreement falls short of requiring a strong public interest clause in antidumping law, which could have mandated national authorities to give balanced consideration to the broader national interest before imposing an antidumping duty. In the United States, the new transparency provision of the Uruguay Round could play some role in bringing somewhat greater balance to the injury investigation, particularly as there appears to be a measurable noneconomic dimension to the final decision.52 It may also affect, over time, the development of new antidumping rules and procedures that might eventually give voice to downstream industries and consumer groups hurt by antidumping actions. In the EU, the community interest provision53 has rarely been used to override a finding of dumping and injury.54 Complaints of a lack of transparency have occasionally surfaced,55 and EU injury decisions appear also to have been influenced by non-economic motives.56 To the extent that transparency is enhanced by the requirements of the Uruguay Round agreement, this may help improve the functioning of the EU’s community interest provision.

Cumulation and the Determination of Injury

In the determination of injury by reason of dumped imports, there has been some disagreement as to whether a small exporter should be caught up in the net of an antidumping action, given that its exports alone could not be responsible for causing material injury. GATT Article VI (antidumping and countervailing duties) and the Tokyo Round Antidumping Code were both silent on this issue. The Uruguay Round agreement clarified this by explicitly allowing the authorities to cumulatively assess injurious effects over all countries and all exporters found to be dumping (with margins in excess of de minimis), as long as a targeted exporter accounts for a nonnegligible share of imports (defined above).57

In practice, Australia, Canada, the EU, the United States, Mexico, and other users of antidumping have all routinely applied cumulation rules to assess injury.58 A number of countries argued for explicitly prohibiting the practice of cumulation in order to curtail the reach of antidumping. The conditions under which cumulation is now explicitly allowed are not very restrictive and are not likely to appreciably alter existing practice.

New Rules for Antidumping Panels

Regarding the settlement of antidumping disputes, the Uruguay Round agreement introduced a new qualification that requires dispute settlement panels to accept the judgment of the local investigating authorities, when that judgment is consistent with an unbiased and objective examination of the facts, “even though the panel might have reached a different conclusion.”59 In effect, this means that the panel cannot insist on its own interpretation of the facts if both interpretations are permissible.

The implications of the new provision will be clear only after experience with implementation of the Uruguay Round agreement. While the overall dispute settlement process has been strengthened due to the new “negative consensus” rule,60 the new provision for antidumping appears to reduce the authority of antidumping panels to assert a liberal interpretation of the facts.

Below-Cost Pricing and the Determination of Normal Value

GATT Article VI and the Tokyo Round Antidumping Code left a degree of ambiguity in the interpretation of procedures for calculating dumping margins. One source of the problem came from the phrase “in the ordinary course of trade.” In the calculation of a dumping margin, the export price is compared with the comparable price (normal value) “in the ordinary course of trade” in the exporter’s home market. Interpretations of what constitutes the ordinary course of trade led most countries to disregard those prices found to be below average total costs (fixed plus variable) in the exporter’s home market.61 Disregarding price observations below average total costs increases the comparator price in the exporter’s home market and thus imparts an upward bias to the calculation of dumping margins. However, it is well known that profit-maximizing firms will find it optimal to sell at prices below average total costs during periods of slack demand, as long as they are able to cover their average variable costs.62 In economic terms, therefore, pricing below average total cost, but above average variable cost, is very much “in the ordinary course of trade.”

The Uruguay Round clarified the rules in this area by explicitly allowing the investigating authorities calculating dumping margins to drop price observations in the exporter’s home market found to be below average total cost, but subject to certain conditions. The agreement authorizes authorities to reject below-unit-cost sales in their estimate of “normal value” only if “… such sales are made within an extended period of time (normally one year but in no case less than six months) in substantial quantities (not less than 20 percent of the volume sold in those transactions under consideration for the calculation of normal value) and are at prices that do not provide for the recovery of all costs within a reasonable period of time.”63

Whether the new clarifications in the Uruguay Round agreement will make a significant difference can only be known from experience with implementation. Certainly, the new provisions should reduce the automaticity with which below-average-cost observations are excluded. On the other hand, the cyclical downturns that often lead firms to price below unit cost typically last for extended periods, that is, more than six months, and often more than one year.


The Uruguay Round has improved antidumping disciplines in several respects. The changes are not sufficiently fundamental, however, to generate the expectation of a significant curtailment of resort to antidumping in the post-Uruguay Round period compared with the recent past. As argued above, economic theory does not support the existence of antidumping policy as currently practiced. Economic efficiency considerations suggest the need for users of the antidumping instrument—both traditional and nontraditional—to reappraise the overall economic costs of the proliferation of this instrument and to consider ways of circumscribing its use. This is not necessarily infeasible, as suggested by current practices in some countries that have occasionally implemented stricter provisions in their national antidumping legislation (e.g., the use of sunset clauses in Australia, Canada, and the EU, and the lesser-duty rule in the EU) than required by the multilateral rules.

Ideally, antidumping should be placed into a competition policy framework,64 or replaced by competition laws suitably amended to reach alleged foreign predators. Short of this, the above review points toward a number of disciplines that might be introduced within the current structure of existing antidumping regimes—or regimes soon to be adopted in developing countries and economies in transition. These might include, for example, tightening access to antidumping duties through more stringent de minimis rules for dumping margins and market shares, tighter sunset provisions, an unqualified rule against excluding below-cost observations in the calculation of dumping margins, and stronger limitations on the practice of cumulation in determining injury. This list of possible technical improvements is far from exhaustive.65

Beyond these technical measures, other, more far-reaching reforms have been suggested: (1) require penalties for antidumping petitions whose complaints are adjudicated to have been frivolous;66 (2) introduce a greater say for consumer groups and downstream industries likely to be hurt by antidumping actions;67 (3) require a cost-benefit analysis to accompany new antidumping actions;68 (4) remove the chapeau of “unfairness” that currently accompanies antidumping actions;69 (5) short of moving to an antitrust framework, competition or antitrust authorities might be given a more active role in evaluating the competitive implications of a proposed antidumping action and given a role in the drafting of legal reforms;70 and (6) limit anti/dumping measures (and indeed contingent protection more generally) to pure trade-adjustment assistance without granting any protective duties.71 All of these reforms are intended to reorient antidumping policy from its current sectoral perspective toward a national perspective.


    AndersonJames E.“Domino Dumping I: Competitive Exporters,”American Economic ReviewVol. 82No. 1 (1992) pp. 6583.

    ApplebaumHarvey M.“The Coexistence of Antitrust Law and Trade Law with Antitrust Policy,”Cardozo Law ReviewVol. 9No. 100 (1988) pp. 116973.

    • Search Google Scholar
    • Export Citation

    BaldwinRobert E.“Assessing the Fair Trade and Safeguards Laws in Terms of Modern Trade and Political Economy Analysis,”The World EconomyVol. 15No. 2 (1992) pp. 185202.

    • Search Google Scholar
    • Export Citation

    BaldwinRobert E.“Quid Pro Quo Restraint and Spurious Injury: Comment,” in Analytical and Negotiating Issues in the Global Trading Systemed. by AlanDeardorff and RobertStern (Ann Arbor: University of Michigan Press1994).

    • Search Google Scholar
    • Export Citation

    BanksGary“The Antidumping Experience of a GATT-Fearing Country,” in Antidumping: How It Works and Who Gets Hurted. by J. MichaelFinger (Ann Arbor: University of Michigan Press1993).

    • Search Google Scholar
    • Export Citation

    BellisJean-Francois“The EEC Antidumping System,” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    BhagwatiJagdishProtectionism (Cambridge, Massachusetts: MIT Press1988).

    BoltuckRichardJoseph F.Francois and SethKaplan“The Economic Implications of the Administration of the U.S. Unfair Trade Laws,” in Down in the Dumps: Administration of the Unfair Trade Lawsed. by RichardBoltuck and RobertLitan (Washington: The Brookings Institution1991).

    • Search Google Scholar
    • Export Citation

    BoltuckRichard and Robert E.LitanDown in the Dumps: Administration of the Unfair Trade Laws (Washington: The Brookings Institution1991).

    • Search Google Scholar
    • Export Citation

    DavidowJoel“The Relationship Between Anti-Trust Laws and Trade Laws in the United States,”The World EconomyVol. 14No. 1 (1991) pp. 3752.

    • Search Google Scholar
    • Export Citation

    DeardorffAlan“Economic Perspectives on Antidumping Law,” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    EymannAngelika and LudgerSchuknecht“Antidumping Enforcement in the European Community,” in Antidumping: How It Works and Who Gets Hurted. by J. MichaelFinger (Ann Arbor: The University of Michigan Press1993).

    • Search Google Scholar
    • Export Citation

    FingerJ. Michael“Should Developing Countries Introduce Antidumping? Never,”Outreach: Policy Views from the Country Economics Department (Washington: The World BankMay1992).

    • Search Google Scholar
    • Export Citation

    FingerJ. Michael (1993a) ed. Antidumping: How It Works and Who Gets Hurt (Ann Arbor: University of Michigan Press1993).

    FingerJ. Michael (1993b) “Overview,” in Antidumping: How It Works and Who Gets Hurted. by J. MichaelFinger (Ann Arbor: The University of Michigan Press1993) pp. 379.

    • Search Google Scholar
    • Export Citation

    FingerJ. Michael and SumanaDhar“Do Rules Control Power? GATT Articles and Arrangements in the Uruguay Round,” in Analytical and Negotiating Issues in the Global Trading Systemed. by AlanDeardorff and RobertStern (Ann Arbor: University of Michigan Press1994).

    • Search Google Scholar
    • Export Citation

    FingerJ.M. with H.K.Hall and D.R.Nelson“The Political Economy of Administered Protection,”American Economic ReviewVol. 72 (June1982) pp. 45266.

    • Search Google Scholar
    • Export Citation

    FingerJ. Michael and TracyMurray“Antidumping and Countervailing Duty Enforcement in the United States,” in Antidumping: How It Works and Who Gets Hurted. by J. MichaelFinger (Ann Arbor: University of Michigan Press1993).

    • Search Google Scholar
    • Export Citation

    FischerRonald D.“Endogenous Probability of Protection and Firm Behavior,”Journal of International EconomicsVol. 32 (February1992) pp. 14963.

    • Search Google Scholar
    • Export Citation

    FudenbergD. and J.Tirole“A Signal-Jamming Theory of Predation,”Rand Journal of EconomicsVol. 17 (1986) pp. 36676.

    GATTThe Tokyo Round Agreements (GenevaAugust1986).

    GATTTrade Policy Review Mechanism: Canada Report by the Secretariat (Geneva: GATT Secretariat1992).

    GATTTrade Policy Review Mechanism: European Communities Report by the Secretariat (Geneva: GATT Secretariat1993).

    GATT (1994a) Trade Policy Review Mechanism: Australia Report by the Secretariat (Geneva: GATT Secretariat1994).

    GATT (1994b) Trade Policy Review Mechanism: United States Report by the Secretariat (Geneva: GATT Secretariat1994).

    GATT (1994c) Final Act Embodying the Results of the Uruguay Round of Multilateral Trade Negotiations (Geneva: GATT Trade Negotiations Committee1994).

    • Search Google Scholar
    • Export Citation

    HoekmanBernard M. and Michael P.Leidy“Policy Responses to Shifting Comparative Advantage: Designing a System of Emergency Protection,”KyklosVol. 43No. 1 (1990) pp. 2551.

    • Search Google Scholar
    • Export Citation

    HorlickGary N.“The United States Antidumping System,” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    HorlickGary N.“How the GATT Became Protectionist: An Analysis of the Uruguay Round Draft Final Antidumping Code,”Journal of World TradeVol. 27 (October1993) pp. 517.

    • Search Google Scholar
    • Export Citation

    IsaacR. Mark and Vernon L.Smith“In Search of Predatory Pricing,”Journal of Political EconomyVol. 93No. 2 (1985) pp. 32045.

    • Search Google Scholar
    • Export Citation

    JacksonJohn H.The World Trading System: Law and Policy of International Economic Relations (Cambridge, Massachusetts: MIT Press1989).

    • Search Google Scholar
    • Export Citation

    JacksonJohn H. and EdwinVermulsteds.Antidumping Law and Practice: A Comparative Study (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    KellyMargaretAnneMcGuirk and othersIssues and Developments in International Trade Policy World Economic and Financial Surveys (Washington: International Monetary Fund1992).

    • Search Google Scholar
    • Export Citation

    LeidyMichael P.“Quid Pro Quo Restraint and Spurious Injury: Subsidies and the Prospect of CVDs,” in Analytical and Negotiating Issues in the Global Trading Systemed. by AlanDeardorff and RobertStern (Ann Arbor: University of Michigan Press1994).

    • Search Google Scholar
    • Export Citation

    LeidyMichael P. and Bernard M.Hoekman“Production Effects of Price- and Cost-Based Antidumping Laws Under Flexible Exchange Rates,”Canadian Journal of EconomicsVol. 23 (November1990) pp. 87395.

    • Search Google Scholar
    • Export Citation

    LeidyMichael P. and Bernard M.Hoekman“Spurious Injury as Indirect Rent-Seeking: Free Trade Under the Prospect of Protection,”Economics and PoliticsVol. 3No. 2 (1991) pp. 11137.

    • Search Google Scholar
    • Export Citation

    LowPatrick and AlexanderYeats“Nontariff Measures and Developing Countries: Has the Uruguay Round Leveled the Playing Field?” (draft Washington: World Bank1994).

    • Search Google Scholar
    • Export Citation

    MagnusPeter A.“The Canadian Antidumping System,” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    MesserlinPatrick“The EC Antidumping Regulations: A First Economic Appraisal, 1980–85,”Weltwirtschaftliches ArchivVol. 125 (1989) pp. 56387.

    • Search Google Scholar
    • Export Citation

    MurrayTracy“The Administration of the Antidumping Law by the Department of Commerce,” in Down in the Dumps: Administration of the Unfair Trade Lawsed. by RichardBoltuck and RobertLitan (WashingtonThe Brookings Institution1991).

    • Search Google Scholar
    • Export Citation

    National Consumer CouncilInternational Trade: The Consumer Agenda (United Kingdom: National Consumer Council1993).

    Organization for Economic Cooperation and DevelopmentPredatory Pricing (Paris: OECD1989).

    PrusaThomas J.“Why Are So Many Antidumping Petitions Withdrawn?”Journal of International EconomicsVol. 33 (August1992) pp. 120.

    • Search Google Scholar
    • Export Citation

    PrusaThomas J.“Antidumping Law and Firm Behavior,”Journal of Economic Integration (June1994) pp. 25988.

    SchuknechtLudger and JoergStephan“EC Trade Protection: Produmping or Antidumping,”Public ChoiceVol. 80 (1994) pp. 143156.

    StaigerRobert W. and Frank A.Wolak“Strategic Use of Antidumping Law to Enforce Tacit International Collusion,”NBER Working Paper No. 3016 (Cambridge, Massachusetts: NBER1989).

    • Search Google Scholar
    • Export Citation

    StaigerRobert W. and Frank A.Wolak“The Effect of Domestic Antidumping Law in the Presence of Foreign Monopoly,”Journal of International EconomicsVol. 32 (1992) pp. 26587.

    • Search Google Scholar
    • Export Citation

    StaigerRobert W. and Frank A.Wolak“The Trade Effects of Antidumping Investigations: Theory and Evidence,” in Analytical and Negotiating Issues in the Global Trading Systemed. by AlanDeardorff and RobertStern (Ann Arbor: University of Michigan Press1994).

    • Search Google Scholar
    • Export Citation

    SteeleH. Keith C.“The Australian Antidumping System,” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989).

    • Search Google Scholar
    • Export Citation

    TharakanP.K.M. and J.Waelbroeck“Antidumping and Countervailing Duty Decisions in the E.C. and in the U.S.: An Experiment in Comparative Political Economy,”European Economic ReviewVol. 38No. 1 (January1994) pp. 17193.

    • Search Google Scholar
    • Export Citation

    TiroleJeanThe Theory of Industrial Organization (Cambridge, Massachusetts: MIT Press1989).

    VermulstEdwin A.“The Antidumping Systems of Australia, Canada, the EEC, and the USA: Have Antidumping Laws Become a Problem in International Trade?” in Antidumping Law and Practice: A Comparative Studyed. by JohnJackson and EdwinVermulst (Ann Arbor: University of Michigan Press1989) pp. 42566.

    • Search Google Scholar
    • Export Citation

    VermulstEdwin A.“A European Practitioner’s View of the GATT System: Should Competition Law Violations Distorting International Trade Be Subject to GATT Panels?”Journal of World TradeVol. 27No. 2 (1993) pp. 5575.

    • Search Google Scholar
    • Export Citation

    WebbMichael“The Ambiguous Consequences of Antidumping Laws,”Economic InquiryVol. 30No. 3 (1992) pp. 43748.

    YanoMakoto“Voluntary Export Restraints and Expectations: An Analysis of Export Quotas in Oligopolistic Markets,”International Economic ReviewVol. 30No. 4 (1989) pp. 70723.

    • Search Google Scholar
    • Export Citation
1The principal author.
2Administered protection, sometimes called contingent protection, refers to antidumping, countervailing duties, and emergency protection under the GATT’s principal safeguards clause (Article XIX). The term also applies to formal government-to-government voluntary export restraint agreements (VERs). These forms of protection are distinguished from directly legislated protection in the form of tariffs, quantitative restrictions, and other nontariff barriers.
3This was due principally to a large jump in the yearly number of revocations of outstanding antidumping measures. From 1979–85, a total of 34 old actions were revoked. This compares with 50 revocations in 1986, 60 in 1987, 66 in 1988, and 43 in 1989. The increase in revocations may be attributable to the significant depreciation of the Australian dollar, which improved the competitiveness of Australian manufacturers, as well as to a more active review process by Australian Customs. These revocations preceded the existence of a sunset clause for antidumping measures in Australia (Banks (1993)).
7The sample includes Australia, Austria, Brazil, Canada, the EU, Finland, India, Japan, Korea, Mexico, New Zealand, Poland, Sweden, and the United States. In addition, the following countries have recently introduced or reactivated antidumping legislation: Argentina, Bolivia, Chile, China, Colombia, Egypt, Indonesia, Israel, Jamaica, Malaysia, Morocco, Peru, the Philippines, South Africa, Thailand, Trinidad and Tobago, and Venezuela (Low and Yeats (1994)).
8From 1980 through 1985, of 1,019 antidumping cases initiated and reported to the GATT, the developing countries initiated none (Finger (1993b), p. 4). From 1985 through 1989, developing countries initiated 34 of 539 cases. In a single year preceding June 30, 1993, Mexico, India, Brazil, and Korea alone initiated a total of 38 cases; that is, one more than Canada, 2 more than the EU, and a bit more than half of U.S. initiations (GATT, Basic Instruments and Selected Documents, various issues).
9The sample of 13 countries and the EU reporting to GATT had 657 antidumping initiations over this three-year period.
10The U.S. and EU figures are based on final duties (or price undertakings) reported to the GATT Committee on Antidumping Practices from July 1, 1992 through June 30, 1993. Because Mexico reports duties for the entire stock of outstanding antidumping actions, the frequency distribution of antidumping duties for Mexico is based on this stock as of June 30, 1993.
11Some of the details of current antidumping rules and practices in Australia, Canada, the EU, and the United States are discussed below in the section on the results of the Uruguay Round.
13On the issue of competition policies and antidumping laws, see Kelly, McGuirk, and others (1992). For a review of the history and recent attempts to bring a competition policy standard to U.S. antidumping laws, see Davidow (1991). In the United States, for example, Section 2 of the Clayton Act, as amended by the Robinson-Patman Act of 1936, prohibits, inter alia, price discrimination within the U.S. provided that it also injures competition.
14See Davidow (1991) and Finger (1993b). Unlike the United States, Canada’s original antidumping regime—the world’s first antidumping system was introduced in Canada in 1904 (Finger (1993b))—did not establish antidumping on an antitrust foundation.
15See, for example, the discussion in Deardorff (1989).
16See Deardorff (1989), p. 35.
17OECD (1989) argues that what evidence there is of predatory pricing is largely anecdotal. It points out that primary and secondary allegations of predatory pricing in U.S. antitrust cases have fallen significantly during the ten-year period 1973–83. And while it also notes that the absence of clear sightings cannot be interpreted as conclusive evidence that predatory pricing does not exist, the study concludes that “[p]erhaps all that can be said is that cases of predation may arise but at most only very infrequently” (p. 81).
19The theoretical industrial organization literature continues to study conditions under which predatory pricing might be successful. Improved methodologies for identifying predatory intent empirically have also been proposed. See, for example, Fudenberg and Tirole (1986) and the discussion in Tirole (1989).
21See Vermulst (1993), p. 71.
22See Article 1907 of the CUSFTA. The agreement to work within a specific timetable toward the implementation of a substitute system of rules was not originally carried into the NAFTA. At the initiative of Canada, an agreement was reached in December 1993 to try to negotiate by December 31, 1995 new rules governing dumping and subsidies.
23The Uruguay Round text on safeguards now allows limited discrimination among suppliers under exceptional circumstances.
24Antidumping actions are linked with the notion that something “unfair” is being corrected. Safeguard actions carry no such cover. Thus antidumping provides import-competing firms and protection-granting governments the “balm of labeling the exporter unfair” (Finger (1993b), p. 58).
25While the GATT (pre-Uruguay Round) did not formally require that exporting countries be compensated by a country invoking the safeguards clause (Article XIX)—only that exporting countries be given an opportunity to consult—in practice, countries invoking Article XIX have granted to interested exporting countries alternative market-opening concessions by way of compensation, so as to avoid the suspension of equivalent concessions or other obligations (i.e., retaliation) by the affected exporters. This has been referred to as the “compensation requirement” (Jackson (1989), pp. 167-68). The Uruguay Round agreement on Safeguards (GATT (1994c)) has changed matters somewhat. The agreement indicates that (Section III:16) “members concerned may agree on any adequate means of trade compensation” and, in the event agreement is not reached and subject to qualified circumstances, the suspension of equivalent concessions by interested exporting countries is not to occur during the first three years of a safeguards action.
26See, for example, the discussion in Finger and Dhar (1994).
27See the discussion in Jackson (1989), pp. 149-54.
29Because a successful antidumping petition needs to satisfy an injury criterion, and because the indicators of injury (domestic sales, employment, recent capital expenditures, profits, etc.) are largely under the control of petitioning firms, there may be a short-run incentive in oligopolistic markets to manipulate these indicators of injury (“spurious injury”) in order to enhance the prospect of winning an antidumping order (Leidy and Hoekman (1991) and Baldwin (1992)).
30See GATT (1994b), p. 57.
33In this scenario, the act of dumping is viewed as a necessary prior condition before governments will intervene to help negotiate and enforce a voluntary export restraint agreement.
34Bertrand competition describes a situation in which firms select prices, taking a competitor’s price selection as given, and the market determines the quantities that clear the market given the choice of prices.
36Cournot competition, in contrast to Bertrand competition, describes a market setting in which firms select quantities, taking their competitor’s quantity choice as given, and the market determines a market-clearing price.
38Spurious injury distortions introduced by the prospect of injury-contingent protection were examined in Leidy and Hoekman (1991), Baldwin (1994) reports indirect empirical support for the “spurious-injury” hypothesis.
39Subsequent references to the Uruguay Round Final Act agreement on antidumping (Agreement on Implementation of Article VI) may be found in GATT (1994c).
40See GATT (1986), p.132.
41See Horlick (1993), p.15.
42Administrative procedures can impose significant costs on exporters even when preliminary duties are less than 1 percent. Boltuck, Francois, and Kaplan (1991, p. 163), argue that “[s]imply placing imports under bond, even at 0.5 percent, can cut off access to the U.S. market.”
44See Horlick (1993), p. 14.
46An exporter must be able to show that there were no sales at “less than fair value” for at least two years, and no likelihood of a resumption of such sales. Alternatively, an exporter may request an injury ruling to show that the domestic industry would not suffer material injury if the duty were revoked.
48The duration of Australia’s sunset provision was changed in the Customs Legislation Act of 1992 from three years to five years (GATT (1994a)).
50For example, Articles 6.6 and 8.5 call for public notice (without elaboration) upon initiation of an antidumping investigation and once a preliminary finding has been reached.
51Article 12 of the Uruguay Round Agreement on the Implementation of Article VI is entitled Public Notice and Explanation of Determinations.
53Council Regulation 2423/88, Article 11(10) and 12(1).
54See, for example, Bellis (1989). Also see Eymann and Schuknecht (1993), p. 230, who point out that “[i]n none of the 904 cases considered during the 1980s did the Commission rule against the imposition of antidumping measures on the basis of injury to users and consumers.”
57The agreement also specifies that a cumulative assessment of the effects of the imports must be “appropriate in light of the conditions of competition between imported products and the conditions of competition between the imported products and the like domestic product” (Article 3.3). This language would appear to impose no concrete constraint on the use of cumulation.
59Article 17.6 of the Uruguay Round agreement on the Implementation of Article VI (GATT, 1994c) says that: “(i)… If the establishment of the facts was proper and the evaluation was unbiased and objective, even though the panel might have reached a different conclusion, the evaluation shall not be overturned, (ii) … Where the panel finds that a relevant provision of the agreement admits of more than one permissible interpretation, the panel shall find the authorities’ measure to be in conformity with the agreement if it rests upon one of those permissible interpretations.”
60Under past practice a panel ruling was not adopted until a consensus was reached. This enabled the “losing” party to block or delay approval of panel reports. Under the new WTO procedures, a panel report will be adopted within 60 days of issuance unless one of the Parties to the dispute states its intention to appeal, or the WTO Dispute Settlement Body decides by consensus not to adopt the report.
61See, for example, the discussions in Bellis (1989), Magnus (1989), Steele (1989), and Murray (1991).
62Indeed, in U.S. antitrust cases, only pricing below average variable costs is considered relevant in price discrimination cases (Vermulst (l993, p. 72)).
63See GATT (1944c), Article 2.2.1 of the Agreement on Implementation of Article VI.
65See, for example, Boltuck and Litan (1991) for an exhaustive discussion of the many procedural biases in antidumping policy as practiced in the U.S.
66See Bhagwati (1988), p. 116.
67The whole of the public choice literature would recommend such an approach.
68Finger (1993b) argues that protection through antidumping should be treated as an exception to liberalization and that by requiring a transparent cost-benefit analysis, the politics of considering a request for an exception will be more balanced by strengthening the voice of those opposed.

    Other Resources Citing This Publication