Revised/31 May 1999
Rodney de C. Grey
For UNCTAD

The purpose of this report is to examine the scope for devising provisions under the aegis of the World Trade Organization (WTO) which would bring anti-dumping policy into a greater degree of consistency with competition policy. Anti-dumping policy is directed at reducing the impact of an anti-competitive practice: namely, selling abroad for less than in the domestic market. Thus anti-dumping policy is, in a rather special sense, a part of competition policy. Competition policy is usually directed at actions taken in regard to the domestic market, and, commonly, by domestic firms. However, all countries with properly articulated competition policies address actions taken abroad, often by foreign-controlled and foreign-based entities, which have anti-competitive effects in the domestic market. Thus the question arises as to whether the anti-dumping system in any given country could be subsumed by that country's competition law and policy, along with other measures against other anti-competitive practices by foreign entities. A more modest question is: why are the standards and administrative rules of the anti-dumping system not more consistent with competition law?

These questions present themselves in the context of the forthcoming multilateral trade negotiations (MTN) to be launched by the WTO, in which it is assumed that a number of countries will seek changes in the GATT "Anti-dumping Agreement,"[1] an agreement which is designed to provide uniform rules for the use of anti-dumping duties, the accepted method for sanctioning injurious dumping.

Put more specifically, the issue is why one particular type of anti-competitive practice - price discrimination - should be dealt with under different rules, tests and standards when it is practiced by exporters in other countries than when it is practiced by domestic firms. Can anything be done to reconcile the two approaches? In fact, the standards and texts of anti-dumping policy diverge from those of competition policy only for historical reasons: anti-dumping measures evolved in a trade and tariff policy context, by different administrative agencies than those that addressed competition policy issues.

Before stating some specific proposals, offered in the context of the next MTN, certain working assumptions of this report should be made clear.

One key assumption is that in so far as the anti-dumping system penalizes price discrimination in import trade more severely than similar price discrimination in domestic commerce (under competition law), the anti-dumping system is protectionist, to that extent and by design. This seems to this writer to be a self-evident proposition, acceptable both to those who seek to make the anti-dumping system less restrictive, and those who wish to maintain and strengthen the anti-dumping system. And to the extent that the anti-dumping system is more restrictive then the competition policy system, its application would appear to be in breach of the National Treatment obligations of the GATT (Article III). [One says "would appear" because one could conceivably argue that GATT Article VI, the article conferring authority to levy anti-dumping duties, can be read as an exception to Article III. Only a WTO panel looking at a test case could decide; however, it seems to this writer that Article III overrides Article VI, and that Article VI does not permit the levying of a tax on imports which is in excess of that levied, in parallel circumstances, on domestics products.]

Another assumption is that governments may quite properly to take some offsetting action or remedial action when there is a sharp increase in the volume of imports at prices which appear unusually low and which displace domestic production to an intolerable degree.

In such circumstances a government has to determine whether this intolerable surge in imports is due to an essentially anti-competitive action by a foreign firm or firms, or whether it reflects longer-term and relatively permanent changes in the structure of production. In the former case measures designed to offset the anti-competitive action may be appropriate; in the latter case governments should focus on adjustment measures - particularly those designed to protect the labour force from the effects of decisions to invest in productive facilities which they could not reasonably be expected to have foreseen when they accepted employment. Of course, price competition, perhaps to a very uncomfortable degree, may be the result of dislocations in the world economy, such as sharp changes in exchange rates and sharp changes in demand in particular markets, as have been manifested during the past year or so by dislocation in the world markets for steel products. Measures designed to deal with anti-competitive actions are not well-adapted to dealing with problems of adjustment to major changes in the structure of production of a particular product, nor to dealing with dislocations in markets due to rapid and substantive changes in exchange rates (a detailed review of the use of anti-dumping measures in the steel sector - in the US, since 1969 - will make this all too clear.)[2] The purpose of any proposals here is not to preclude action by governments to deal with dislocations in domestic markets which impose unreasonable burdens on workers, or which are manifestations of anti-competitive actions. However, as to how governments should best intervene, this writer accepts the logic of what has been called the "injury only" view: that is, that the essential task of the trade policy administration is to consider to what extent an industry is "injured," to then assess what measures of adjustment are reasonable in the circumstances, and only then to decide what, if any, restriction (or tax) an imports is justified. This view was expressed many years ago by a distinguished US. trade policy lawyer, Noel Hemmendinger:

"My thesis is that the Antidumping Act and the Countervailing Duty Law [�], are seriously defective in conception and need to be fundamentally reexamined. This is true in two respects. First, they attempt to deal with international economic conflicts through adversarial litigation, hobbling the essential elements of administrative discretion and negotiated solutions. Second, they assume an economic world free of governmental interventions, which has never existed, and more to the point [�] a world of fixed currency rates more or less rationally related to comparative advantage, which has been non-existent at least since 1971. I suggest that a variation on the Escape Clause would be the appropriate U.S. legal mechanism for addressing import trade problems."[3]

Put in GATT terms, this is the thesis that restrictions in imports which cause intolerable problems of adjustment to domestic producers should best be dealt under the safeguard provisions of Article XIX, rather than under Article VI. Clearly, this is not a negotiable solution to the problem, but the "injury only" view, so expressed, is a valuable and reliable perspective on how problems of import competition should best be addressed.

Yet another assumption, or working premise, is that the option of simply subsuming measures against international price discrimination under competition law and polices is not, at this stage, a realistic goal in WTO negotiations. One reason for this, which many would advance, is that in many countries - notably the U.S., Canada, the E.U., Australia, and more recently, in a number of developing countries - business men have become strong -- too strong -- supporters of the anti-dumping system. They realize it gives them protection against imports and they like that. The rhetoric of "fair trade" has convinced them that they are entitled to such protection - and it is substantial - as is afforded them by invoking anti-dumping measures. This is a practical, political reason why the anti-dumping provisions simply can't be scrapped, and subsumed under the competition policy apparatus.

A second, and equally cogent reason, is that competition law is itself - as between jurisdictions - neither always fully developed - nor is there international agreement on policy. This reflects the fact that within many countries there are sharp, and often highly politicized differences of views about the economic and legal rationales of competition policy. The focus of this paper is on the reform of anti-dumping policy, under the aegis of the WTO, rather than on the reform of competition policy; it is sufficient to note that competition policy varies widely from one WTO member to another. We shall, none the less, look briefly at some features of the competition law system addressed to price discrimination, to see whether they might be adapted to the anti-dumping system.

A fourth working premise of this report is that, at the very least, the anti-dumping system should not itself involve or encourage actions which are anti-competitive. This is particularly the case in regard of the use of "undertakings"[4] by exporters to raise prices, to cease dumping (or "injurious" dumping). Undertakings involve, in practice, the creation of combinations of exporters to fix export prices (and, no doubt, to allocate markets) in the importing countries. These are arrangements which were they not entered into under the aegis of the administrative authorities of the importing country, would be actionable offenses under the competition laws of most countries with developed competition law regimes. Further, such "undertakings" often involve the informal or tacit participation and agreement of the "injured" domestic industry, under the umbrella of the administrative authorities. They are thus no more than thinly-veiled conspiracies to raise prices. Thus the fourth premise or working proposition of the report is that the least that should be done to bring the anti-dumping provisions into some measure of conformity with competition law is to avoid the use of the anti-dumping system to encourage anti-competitive practices, i.e. market restriction or allocation by foreign cartels.

In the balance of this paper we examine some specific issues which will have to be considered, and perhaps negotiated, to bring the anti-dumping provisions of the WTO A/D Agreement somewhat closer to competition law concepts. We will need to look at the concepts of price discrimination under the two systems, at the concept of predatory pricing, particularly as shown by competition policy in practice (drawing mainly an U.S. and Canadian practice), at the concept of an "injury" to an industry and "injury" to competition - two quite different concepts, at the concept of "cause" under the various national anti-dumping regimes, and at the related concept of "material" injury. Finally, we should consider suggestions as to how developing countries might negotiate to narrow the very large gap between anti-dumping law and competition law. But before doing so perhaps two disclaimers are in order, and we will need some brief comments on the history of the anti-dumping provisions of the GATT/WTO.

The first disclaimer relates to the question of how and where to discuss the issues of detail which, of course, arise when negotiations are put in hand. A good deal of the literature - both the literature supporting the anti-dumping system as a necessary good, and those opposing it as an unnecessary evil - address the elements that are involved in the measurement of price discrimination (that is, of the "margin of dumping"). Such an approach will inevitably involve prolonged negotiations and minuscule improvements in detail. The Uruguay Round made that quite clear. Insofar as such detailed aspects of the techniques of price comparison may be involved, it may be sufficient to note that, since the Tokyo Round, if not before, there has been created a very extensive literature on all the separate components of price. These issues cannot be summarized usefully - they are extremely detailed - and the "devil is in the details." (There was a time, prior to the Kennedy Round, when it was possible for one person to be familiar with all the literature on anti-dumping, with all the statutes, and with most if not all the jurisprudence. Even before the Tokyo Round, this became no longer possible; major studies of anti-dumping are, increasingly, the work of groups of experts). Fortunately, a number of recent studies include detailed bibliographies, which enable specialists to look at specific issues, particularly issues affecting price comparison.[5] The second disclaimer is as much the same lines: it is not possible, in this short paper, to make any detailed examination of how price discrimination is addressed in the competition law systems of major trading countries. This is, like the calculation of the "margin of dumping," a matter of great detail (and, in any event there are detailed studies available of this aspect of competition law in various jurisdictions (e.g. U.S., EU, Canada). We will note some major questions which arise, but a cursory review of the literature of anti-dumping and the literature of price discrimination might well suggest that to embark on an attempt at reconciliation between the two systems, on this issue, would be an enormously detailed, time-consuming exercise (supposing the major negotiating countries were willing to undertake it) and probably prove in the end to be a blind alley.[6]

The anti-dumping provisions (that is, the Agreement and national legislation) is not directed at predatory dumping, but at any dumping which causes injury, or threatens injury, to a domestic industry. In fact, there are few, if any, cases of anti-dumping action in the U.S., in Canada, or in the EU, where it is clear that predation - which involves intent - is evident. Predatory pricing is pricing designed to eliminate a business competitor, and thus is anti-competitive in nature. But how (unless there is documentary evidence) can we establish intent or design? (This was part of the problem with trying to enforce the U.S. anti-dumping legislation of 1916.) Thus there has been a search for some sat of objective test of predatory intent, in the context of competition law. As opinion has evolved, it has become accepted that sales below marginal costs suggest predatory intent, but not necessarily sales below total average cost but above marginal costs.[7] To put the issue more exactly the key academic text is that by Richard Posner:

"[...] the most useful definition of predatory pricing is the following: pricing at a level calculated to exclude from the market an equally or more efficient competitor. Only two practices fit this definition. The first is selling below short-run marginal cost. There is no reason consistent with an interest in efficiency for selling a good at a price lower than the cost that the seller incurs by the sale [...]

The second practice that is predatory under my definition is selling below long-run marginal cost with the intent to exclude a competitor. Long-run marginal costs are those that must be recovered to stay in business for the more or less indefinite future."[8]

It is readily apparent that this is a different standard than applied in the anti-dumping provisions. At issue is the calculation of the exporter's price in his domestic market; Article 2 of the Code deals with this issue. As summarized by a U.S. lawyer:

"Article 2:1 continues the limitation of the Tokyo Round Code that the home market (or third-country) sales used for comparison to export prices be made "in the ordinary course of trade." Unlike the Tokyo Round Code, however, the Uruguay Round Code makes explicit that sales below total cost are not "in the ordinary course of trade." These sales may be disregarded in determining normal value if they are made "within an extended period of time in substantial quantities and are at prices which do not provide for the recovery of all costs within a reasonable period of time."[9]

This is manifestly far from the standard of predatory pricing as proposed by Posner, and generally accepted in U.S. competition law. It is also far from the standard of predation in Canadian law. There predatory pricing is defined in the Competition Act as

"a policy of selling products at prices unreasonably low, having the effect or tendency of substantially lessening competition or eliminating a competitor, or designed to have such effect."

and the Canadian courts have held that

"[i]f an article is sold for more than cost, it can never be held to be unreasonable."[10]

We need not pursue the comparison between predatory pricing, as addressed in competition law, and dumping prices, any further; sufficient to say that anti-dumping deals with price discrimination, not predatory pricing. If the predatory pricing standard were to be applied in anti-dumping cases, it would be sufficient defense against the charge of dumping to show that the prices at issue were above marginal cost. Clearly this is, for the present, not a negotiable approach.

The anti-dumping provisions speak of "injury" to the domestic industry, and this, as a practical matter, means some firms in the industry having to reduce prices and or lose sales.

In the U.S. system, if there is dumping, and if there is the condition called "injury" - that is, the domestic producer finds he must cut prices or lose sales, the test of "injury" is met, although the dumped imports may be only one of a number of factors influencing the market behaviour of the domestic producer. (This aspect of the anti-dumping system is examined below under causality. The EU and Australia are said to apply a more rigorous standard.) But whether one looks at the U.S. system, which appears to be the most trade restrictive, or at others, this is a very different approach from how the effect of price discrimination in domestic markets is considered in competition law. The relevant U.S. law (Section. 2 of the Robinson - Patman Act) makes it unlawful to engage in price discrimination

"[�] where the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce, or to injure, destroy, or prevent competition with any person who either grants of knowingly receives the benefit of such discrimination, or with customers of either of them:"[11]

On the face of it, this is an "injury to competition" or "injury to the structure of competition" test, not merely a question of whether the competing domestic industry has had to cut prices or lose sales. However, a close examination of case law shows that lost sales, or forced price cutting, as a result of price discrimination, may go some way to meeting the test of injury to "competition."

The essential view of competition law authorities in the U.S. is, however that of the Attorney General's 1955 National Committee to Study the Antitrust Laws:

"[T]his Committee recommends that analysis of the statutory "injury" center on the vigor of competition in the market rather than hard-ship to individual businessmen. For the essence of competition is a contest for trade among business rivals in which some must gain while others lose, to the ultimate benefit of a consuming public. Incidental hardships to an individual businessmen in the normal course of commercial events can be checked by a price discrimination statute only at the serious of risk stifling the competitive process itself."

The anti-dumping concept of injury is simpler than the typical competition law test, but in practice the difference may not be as great as critics of the anti-dumping system have suggested.[12] The problem with the anti-dumping concept of injury to domestic producers is better discussed as the issue of "causality," but here we should note that "injury" to an industry is taken, in U.S. law and in other jurisdictions, to mean the existence of a member of negative developments lost sales, reduced market share, reduced profits that is, an assessment of the general health of the industry, rather than the notion of a specific adverse effect, such as is implied when we speak of an individual having been injured by a particular or specified external factor or event, such as being hit by an automobile.

Another way of stating this is the assertion that injury is "separable" concept - that is, an industry can be suffering from various injuries, from various causes. This is, indeed, the layman's view of what is an injury. In the Tokyo Round negotiation (of the Subsidies - Countervail Agreement - in which this issue arose) the U.S. representative agreed that "injury" is a "separable concept" and indeed that phrase was advanced by the U.S. representative. If injury is separable, then the issue is the general state of health of the industry is, strictly speaking, irrelevant - and the injury test would then seem radically different, and rather less easy for domestic producers to satisfy.

The solution to the dilemma of "cause" could go some way to make the anti-dumping provisions less restrictive of legitimate trade and more nearly consistent with competition law standards. The issue is not simple. This writer's view of the meaning of the relevant phrases of GATT Article VI is that the dumped imports must be shown to be the cause of a material injury to the domestic industry. This view of the causal link is apparently accepted by the EU administrative officials, but only a detailed examination of EU anti-dumping cases could show whether this is so in practice. The Australian practice is described as follows:

"It is not sufficient that the local industry be suffering detriment and that there be some dumping and/or subsidization of the imported goods. It is necessary that the dumped or subsidized goods themselves be the cause of the material injury."[13]

U.S. practice is quite different. The key phrase in the Code is in Article 3.5 "It must be demonstrated that the dumped imports are through the effects of dumping ["] causing injury within the meaning of the Agreement." What is at issue. In the U.S. view - as shown in a long series of U.S. International Trade commission decisions in anti-dumping, subsidy and escape clause cases - in that injury is being in a state of ill-health, or having suffered some adverse effects; it is not, as we noted above, a reference to a particular harm or damage that can be attributed to a specific event, whether internal (e.g. mismanagement) or external (e.g. a change in demand). Given this rather elastic concept of "injury," the important question then is to decide on the causal connection. Article VI of the GATT, and Article 3.5 or the Agreement deal, inter alia, with the causal relationship. The provision reads as previous:

3.5 It must be demonstrated that the dumped imports are, through the effects of dumping ["] causing injury within the meaning of this Agreement. The demonstration of a causal relationship between the dumped imports and the injury to the domestic industry shall be based on an examination of all relevant evidence before the authorities. The authorities shall also examine any known factors other than the dumped imports which at the same time are injuring the domestic industry, and the injuries caused by these other factors must not be attributed to the dumped imports. Factors which may be relevant in this respect include, inter alia, the volume and prices of imports not sold at dumping prices, contraction in demand or changes in the patterns of consumption, trade-restrictive practices of and competition between the foreign domestic producers, developments in technology and the export performance and productivity of the domestic industry."

The more-recent, and authoritative discussion of this provision (which this writer has written about elsewhere) is the 1995 article by David Palmeter. As he points out, the U.S. ITC position is that "the Commission must determine whether imports are a cause of material injury," and the U.S. Court of International Trade has stated that "A cause which even minimally contributes to material injury is sufficient." Palmeter sums up the issue:

"The key to the issue is how the seemingly innocent phrase, "a cause of material injury," is interpreted. One way of interpreting it would be to emphasize the adjective "material" as in "whether imports are cause of Material injury." Under this interpretation, the injury caused by the dumped imports alone must be Material. This would seem to be the interpretation closest to thement of the first sentence of Article 3:5: "It must be demonstrated that the dumped imports are, through the effects of dumping casing injury within the meaning of this Agreement." under the minimal cause doctrine, the ITC reads this phrase differently to emphasize the indefinite article "a" as in "whether imports are a cause of material injury" under this interpretation, of imports are one of a thousand causes then they are a cause, and the requirement of Article 3:5 has been met ["]

This practice goes a long way toward reading the causation requirement of Article 3:5 completely out of the Code. None the less, it is a practice which, the United States argues, received the sanction of the panels formed to review the ITC's decision in Fresh and Chilled Atlantic Salmon from Norway.[14] (Emphasis added)

The wording referred to (Article 3.5) is not significantly different from the wording of the Tokyo Round version of the Agreement, which was in turn an attempt to clarify what had been agreed in the Kennedy Round. Speaking as one of the negotiators of the Kennedy Round Agreement and of the Tokyo Round Agreement, this writer wishes to confirm the view expressed by Palmeter above as to the requirement of Article 3.5. As for the proposition that the GATT Anti-dumping Committee Panel on Norwegian Salmon confirmed the U.S. view, it is now possible for outside observers to decide for themselves whether or not this is the case, because the Panel reports have now been published.[15] Palmeter argues that, in effect, the Panel did not deal properly with this issue, and a close reading of the Reports confirms this. (That is not to say that this case at issue - Norwegian salmon - was a good case with which to examine the issue, but in this writer's view the Panel did not address the legal argument adequately.)

It is virtually self-evident that, if the Australian/EU interpretation of the causality obligation was adopted by all signatories, the provisions would be less restrictive of trade and would be more consistent with the approach in competition law to assessing the causal connection between an instance of price discrimination (in the domestic market) and the alleged impact on competition.[16]

This issue is noted by Palmeter in the citation above. It seems preferable, for analytic purposes, to discuss this as a separate question. The issue above is: Has the dumping been the cause of an injury to an industry? Here the question is. Is that injury material? (And how does it differ from serious injury, under Article XIX?)

The various degrees of injury at issue under the two key GATT articles, (VI and XIX) and the Agreements could be more clearly set out. As this writer sees the matter,[17] at one extreme there is that degree of adverse impact on domestic producers which is "negligible," which does not warrant any intervention, which should not be actionable. We are considering, not the quantity of imports, nor the margins of dumping, but the degree of impact: how much injury? Further along in the progression there is that degree of adverse impact which is "material," that is the key word in Article VI, and in the Agreement. Nothing in the GATT wording or in the history of the drafting of these provisions (at least, in the experience of this writer) suggests that "material" begins where "negligible" ends, although such an approach, of course, commends itself to protectionists. In the absence of any GATT (or Article VI Agreement) provision defining "material," the U.S. Congress legislated a definition in the Trade Agreements Act of 1979: "in general, the term 'material injury' means harm which is not inconsequential, immaterial or unimportant."[18] Given that precedents and practices under the two GATT Article Vi Agreements tend to become internationalized (that is, given the likelihood of producers seeking definitions, precedents and standards in the practices of other countries), this U.S. definition is a definition which has created a "merely more than de minimis" injury test.

Further along in this progression of adverse impacts, there is that degree of impact which is "serious" and which under GATT Article XIX may justify the withdrawal of a tariff concession. It is implicit in the GATT that the withdrawal of concession can be justified only by a degree of impact considerably greater than that which has to be determined to exist to warrant action against "unfair" imports.

It should be dear from these few comments that it would be in the interests of developing countries, and perhaps of developed countries as well, that the word "material" be given a positive and meaningful sense - to signify a degree of adverse impact that is substantially more than merely trivial or negligible. As matters stand, in U.S. practice, at least, the word "material" is without meaning.

The wording of the Uruguay Round Anti-dumping Agreement appears to make it mandatory that domestic authorities, when investigating the impact of dumping, consider "trade restrictive practices of and competition between the foreign and domestic producers." This phrase is from Article 3.5. Not all countries (e.g. Canada) have taken all Agreement provisions into their domestic legislation; accordingly, in domestic law in those countries there may be no requirement that the requirements of Article 3.5 be considered by the administrative authority. Clearly, here is one step that could be taken to bring competition policy considerations into anti-dumping policy without changing the international rules: all that appears to be required is that the existing Agreement provisions be properly implemented. (The same reasoning applies in relation to the application of countervailing duty).

It is not clear however, that the wording of the Agreement requires domestic administrative authorities to deny the protection of the anti-dumping provisions to a domestic producer which is a monopoly, or which is in a dominant position in the domestic market. On this point, it would have appeared that the EC Extramet decision, requiring the EC authorities to take into account EC competition policy rules when assessing injury by dumping, would have set an important precedent. In the case before the European Court of Justice, in June 1992, it was held that the institutions, of the Community must take into account EC competition rules when determining damage by dumped imports (the Extramet decision.[19]) Prior to this decision, the EC Commission appeared to give little weight to competition policy rules in making anti-dumping decisions. In one case, the Commission held that the Community producers had been injured, despite the fact that some aspects of their pricing policy had been criticized by the French anti-trust authorities. What was involved was a refusal to supply a product by the sole producer in the EU. The customer switched to imported supplies and the producer lodged a complaint of dumping. The customer charged the supplier with abuse of a dominant position, under the EU competition law provisions. The Commission imposed an anti-dumping duty on the customer who had turned to imports: this was appealed to the ECJ, which made the decision noted above. All well and good - competition law policy could not be ignored in anti-dumping cares. But the Commission anti-dumping authorities were not prepared to be disciplined by the Court. As Patrick Messerlin reported:

"["] without any formal re-initiation of the case, the Commission re-opened the file and quickly ended it with dumping margins six times higher than those assessed in the initial case-despite the strong anecdotal evidence that the anti-competitive behaviour of Pechiney was still going on" (emphasis in original).[20]

If it were to be agreed that firms in a dominant position, or firms held to be abusing market power in the domestic market, should be denied the protection of the Agreement provisions, there would be some logic in considering just what of sanctions should be applied when the dumping at issue is being carried out by a firm in the exporting country which exercises market dominating power, or, for example, is able to dump exports because it serves a protected domestic market - protected by tariffs, quantitative restrictions, procurement preferences, and the like, which enable it to charge higher prices. Should the anti-dumping system, in taking competition policy into consideration, distinguish between price discrimination in an import market by an exporter in a relatively competitive domestic market, and price discrimination by an exporter which enjoys a highly protected domestic market or a position of domestic market dominance. One could envisage that remedies more effective (and less cartelizing) than anti-dumping duties, such as "cease and desist" or "exclusion" orders, might be invoked. One should recall that anti-dumping provisions, when first proposed by Canada in the early part of the century, were held to be necessary because of the export practices of protected "trusts or combinations."

Set out above are the issues which could be negotiated to achieve the stated objective: to make anti-dumping policy more consistent with competition policy. We will summarize them, in the form of a conclusion, and make some additional proposals. But before doing so, it would be useful to look briefly at the history of the GATT/WTO Anti-Dumping Agreement, in order to better understand how the issues set out above have arisen. That history may provide some guidance, some clues, for the next attempt at revising the Agreement.

The Anti-Dumping Agreement negotiated in the closing phases of the Kennedy Round - in the winter of 1966-67 - was the first attempt under GATT auspices to create a binding code of conduct, a set of contractual rules, governing the use of what was labelled a 'non-tariff barrier'. Previous GATT negotiations, apart from the drafting of the GATT Articles (which derived from the commercial policy provisions of the Havana Charter) and the GATT revision in 1955, had focussed on tariff rates. Reductions in tariff rates could be negotiated in what appeared to be a quantifiable fashion, the agreed results could be set out in schedules of rates, and simple rules could be developed for the conduct of negotiations. However, it was obviously more difficult to assess benefits and cost (to adopt the mercantilist language of GATT negotiations), or of concessions obtained and concessions given, when formulating a set of common administrative rules. Each major participant had to consider just how the emerging set of rules could be presented as an achievement by its negotiators. And there were some sharp differences as to objectives as between the major negotiating parties. The GATT member country which initially pressed for the negotiation of a set of rules governing the administration of anti-dumping duties was the U.K. (not at that time a member of the European Community). The principal objective of the U.K. was to set some limits on the ability of the US authorities (the Treasury, and its Bureau of Customs) to hold a given category of goods being imported 'subject to appraisement' if it later appeared that they were dumped and that there was injury to domestic producers by reason of' that dumping. There were no rules requiring that this very disruptive administrative measure be limited to a fixed period of time. Most importers, if they wished to continue to import, either built up large contingent liabilities for duties, or came to terms with the authorities - even if they felt that the dumping of which they were being accused was not injurious. The U.K. wanted to set some limits on this US practice, and US representatives were prepared to concede, in the context of the GATT multilateral agreement, that some limits should be placed on the duration of 'provisional measures'. The U.K. was also interested in getting the Canadians to adopt an overt test of injury - and the Canadians were prepared to concede that - if they could secure some reform of the US system. The Canadians were believed to have a so-called 'automatic' system, without an overt and open public inquiry into whether or not domestic producers were being injured by dumping. The U.K. and the US both argued that the Canadian system, which relied on the application of a number of complicated legal tests, did not meet the injury requirements of the GATT Article VI. The Canadians were prepared to follow the US model (the Tariff Commission, later the International Trade Commission) of an independent inquiry, and reduce the discretion of the customs authorities in regard to dumping, if they were allowed to act against so-called 'sporadic' or hit-and-run dumping and if the US was prepared to limit harassment of legitimate or non-injurious dumped imports. The Commission of the EC, for its part, realized that the creation of a GATT agreement requiring application to the Community as a single market, would add to the powers of the Commission. Behind the Commission, only the French were really active users of anti-dumping measures. For the French, the main problem - as they saw matters - was the scope for French subsidiaries of foreign firms to compete in the French market by importing components at artificially low transfer prices from their parent firms (i.e. from their US parents). This was the problem of 'hidden dumping'. It is unlikely that at the time of the Kennedy Round EC officials could foresee that anti-dumping would become the protectionist 'weapon of choice' for EC producers and the principal device which the Commission could use to meet domestic producers' complaints about import competition.

The Kennedy Round agreement, and its subsequent application by major GATT signatories, revealed a number of not entirely foreseen results.[21] One was that as each major negotiating party insisted on incorporating in the Agreement the special features of its domestic law which it wished to preserve, all parties to the agreement acquired rights to use all the various different devices. Thus the US kept the right to apply 'provisional' measures, and so did every other country. The Canadians insisted on provisions in regard to 'sporadic' dumping (Article 10.6, .7, .8), and so other countries acquired the same right. The US insisted on a regional industry provision (Article 4.1 (ii)), and so did Canada.[22] The EC opposed the concept of an independent outside body inquiring into injury (as that would severely limit the discretion and authority of the Commission) and so no such institutional provision was put into Article 6.

A second feature of that negotiation was that the drafting was not always very precise, or alternately, where it was precise the words meant different things in different legal systems (e.g. the concept of cause, discussed above). The Agreement was, in effect a compromise between the detailed precise drafting of the US, which the Canadians also favoured, and the much looser legal drafting of the European Commission. Much of the later disputes about anti-dumping resulted from this rather loose drafting - and, of course - once administrative practices developed which relied on the imprecision in the Agreement, administering officials developed vested interests in those practices. For example, it took until 1994 - 27 years - to add precision to the concept of sales below cost and to the concept of 'the ordinary course of trade' (Article 2) - even if that precision authorized a basis for price comparison quite different from the standards of competition law (as noted above).

Experience with this first GATT non-tariff agreement might be compared with other agreements in the area of trade law and international commercial law. 'For example, most conventional trade agreements have relied on phrases and expressions which have been in use - and been tested - since the late 18th century. Other instruments, such as the UNCITRAL agreement on sales contracts, are the results of accumulated commercial practice and detailed drafting over a period of years. The Anti-Dumping Agreement, as a legal instrument, suffers in comparison.

Yet another result of the Agreement, which should have been forseen, was that once the Agreement was in place, it would become the preserve of the administering officials, rather than of trade policy or economic policy officials. Only the administrators, and trade policy consultants and members of the trade law bar, now understand the various anti-dumping systems - so, as a practical matter, they have captured these systems. (Accordingly, it will be important to those countries which may wish to reform the system - that is, to liberalize the system, to make it less restrictive - that they endeavour to get the negotiation into the hands of policy officials and out of the grip of professional anti-dumpers.)

One development after the Round was that the US Congress was unwilling to make the changes in legislation (particularly on the issue of 'cause') which the Agreement entailed. This was in part a quarrel between the Administration and the Congress over turf - resulting from what appeared to have been a failure on the part of officials to consult Congressional leaders fully before and during the negotiations; it was also in part a difference over policy - the Administration was less protectionist that some Congressional leaders.[23] This was all unexpected by the negotiators of other countries, who had assumed, quite properly, that the US negotiators had cleared their lines with domestic business groups and with the Congress. The result of this confusion was that the US implemented only those administrative changes which could be achieved by regulation, not requiring Congressional approval. The further result was that Congress became much more aware of the possibilities of the anti-dumping system being used to meet pressures for protection, and much more determined to keep control of trade negotiations. (It was thus that the so-called 'fast track' system was invented.)

In the years between the Kennedy Round Agreement (and its non-implementation by the US) and the launching of the Tokyo Round, in 1975, the use of anti-dumping measures - by Canada, the US, by Australia, and by the EC - increased significantly. It was as though the Kennedy Round Agreement had become a sort of open, general hunting license - for a perfectly legal way to harass importers and impose a discriminatory duty on imports.

Negotiations in the Tokyo Round focussed on subsidies and countervailing measures; the group of anti-dumping measures was not very active, and it was only after the Tokyo Round Agreement on Subsidies and Countervailing Measures had been virtually completed that the negotiators turned their attention to bringing the wording of the Anti-Dumping Agreement (on 'cause' and on 'injury') into line with the language agreed in regard to countervail. The most important of the changes was the dropping of the Kennedy Round language about 'principal cause'. This followed a detailed (but informal) discussion in the countervail group. There it was argued that what was at issue was injury which resulted solely from the subsidization (or dumping) of imports, that injury was a 'separable concept' - to use a phrase of the US representative. Like most of the real negotiation of this and other non-tariff issues, this discussion took place, not in a plenary meeting, of which there would be a record, but in an informal group (but attended by the GATT Secretariat). Somehow this understanding among the main negotiators did not survive the voyage to Washington, and no trace of it appears in the Congressional documents, in the Statement of Administrative Action, nor in the account by two American officials which was later published.[24] Hence the continuing confusion, and differences of interpretation, on these issues, in regard both to anti-dumping and to countervail. The other major development was the introduction of the world 'material' into US legislation. As we have indicated above, the insistence of the EC that the US legislation be so amended was counter-productive.

After the Tokyo Round, and leading up the Uruguay Round, there was extensive use of anti-dumping (and of countervail, by the US) and extensive discussion in GATT groups of experts of various aspects of the anti-dumping system, or systems. These discussions were between professional anti-dumpers - who were thus enabled to exchange much information about how to make their various systems more precise, more detailed - and thus more perfectionist. It is this sort of discussion that must be curtailed if there is to be any serious negotiation leading to the anti-dumping regime becoming less protections (i.e. more consistent with competition policy). The Uruguay Round negotiation led to minor improvements; that they were only minor was to be expected from the tenor of discussions in the groups of experts referred to above, and given that no major negotiating country was committed to liberalizing the system. The protectionism that has previously been contained by the tariff system had been diverted into a new channel - involving much more detailed administration, and much to the profit of the rapidly enlarging trade law bar in Brussels, Washington and elsewhere. Thus in addition to the obvious interest of producers in finding some device affording protection against imports, two new and powerful interest groups were created in the period from the end of the Kennedy Round to the end of the Uruguay Round - the administering officials and the trade law bar.

The Uruguay Round did produce, as we have said, some detailed, marginal changes: the rules on 'standing' - that is, who has the right to launch a complaint of dumping, and certain precise reforms in the techniques of calculating the margin of dumping, notably, the rules regarding the margin as between an import price and the average value of goods sold in the exporter's domestic market.[25] These changes did little more that add a further measure of legitimacy to a system which provided administered protection.[26]

As for the definition of the de minimis margin of dumping (2% of the export price) introduced into the Agreement (Article 5), US trade experts explained in testimony before a House of Representatives Committee that this would "have no impact on the vast majority of US cases."[27]

The very brief comments set out above on what is in fact a long, complicated and essentially unrecorded history, retained only in fragmentary accounts of a few plenary meetings, evasive agreed texts, and the fading recollections of negotiators, should serve as warning to negotiators who may formulate the more ambitious objective of moving the GATT/WTO Agreement on Anti-Dumping in the general direction of competition policy. Set out below are some suggestions.

(1) To concentrate on the minutiae of the calculation of the margin of dumping would not be profitable. Negotiators should not be diverted into such a detailed and time-consuming exercise. As long as markets are protected by transportation cost, by the existence of established brands and trade marks, by the existence of distribution systems which are even partly in the control of domestic producers, there will be dumping, and it will not be beyond the wit of administrators to find a positive measure of such dumping. The calculation of margins is not a major issue, it is subordinate, but it is what professional anti-dumpers will be prepared to talk about and around for the duration of the next MTN.

(2) Another issue which might be thought negotiable, but which might turn out to be a blind alley, is the question of 'injury' to what, and to whom. Academic critics of anti-dumping never titre of pointing out that the anti-dumping provisions (GATT Article VI and the Agreement) speak of injury to the 'industry' in the importing country, in contrast to competition law which addresses injury 'to competition'. As suggested above, this difference is perhaps less evident in practice: under competition law evidence of injury to competitors would be lost sales, lost market share, and being obliged to cut prices. If competitors are injured, then competition is injured. It might thus well be that trying to reconcile the two apparently different approaches to injury would merely serve to emphasize that injury under the anti-dumping provisions is not all that different conceptually from injury to competition under competition law.

(3) What might be more profitable would be to seek agreement that only a really intolerable impact on domestic producers warrants the application of a discriminatory duty - that is to say , the adjective 'material' should be given some real meaning. This will not be easy to negotiate, given that it was the US Congress (i.e. the Senate) which put in place the definition of material which governs in US law. But if it could be argued that 'material' injury is an adverse impact substantially more than merely 'immaterial', something much more that what in effect is merely the normal manifestation of competition. That would move the anti-dumping provisions a useful distance toward the more rigorous standards of competition law. A negotiation of this issue would necessarily involve consideration of the difference between material injury, under GATT Article VI (and in the two Article Vi Agreements) and the concept of 'serious injury' under Article XIX - the so-called 'safeguard' provision. This will be difficult negotiating ground, but logic and reason would be on the side of those who would wish to make the 'material injury' test a standard of adverse impact that would be meaningful. As noted above, it is this writer's view that there is a progression in the GATT language about the adverse impact of imports on domestic producers. Because dumping is thought of as an uncompetitive practice, the degree of adverse impact justifying action against the imports in question is logically less than that degree of impact which justifies action against 'fairly traded' imports - i.e. imports which are coming in at prices and in quantities which have an intolerable impact on domestic producers (and on their employees) - and thus which justify the withdrawal of a concession (but on a non-discriminatory basis). Beyond that there is that degree of disruption of markets which has been held to justify restrictive action not provided for in the GATT - for example, the actions taken against textiles and garments under the MFA and prior agreements in this sector. But this does not amount to saying that the lesser degree of adverse impact from 'unfairly traded' imports is properly defined as merely more than 'immaterial'. The application of a discriminatory duty to imports can only be warranted if the impact on the domestic producers is something more than that following from normal competition for markets.[28] Now that those signatories of the Agreement which have been majors users of the anti-dumping provisions are finding that they face action by other countries they may be more willingness to consider making the system more realistic and therefore, less protectionist. A working group could be established to draft an interpretative note putting some needed flesh on the word 'material'.

(4) Another profitable area for the next negotiation is to sort out the confusion between the notion of injury being done to the domestic producers or the domestic industry, and the notion of the general health of the industry. In US practice, injury means much the same as ill-health - or what has been called 'overall injury'. What Article VI (and the Agreements) speak of is 'injury' caused by dumping (and that injury must be material). This writer described the issue in the following terms, after the Kennedy Round:

"What is at issue is that a practical matter, an industry seeking relief from dumping is likely to be under the influence of various adverse factors - changes in demand, changes in costs, or changes in the character of import competition. Some of these are external factors impinging or having an impact on the fortunes or health of the industry. Others will be in a sense internal factors. But whether or not the industry is healthy or depressed, it is entitled to relief from dumping if injury - say a marked decline in profits, in employment, and in sales - can be shown to have been caused by dumped imports. Having satisfied themselves that there is an injury caused by the dumped imports and not by some other factor, the authorities must then decide whether or not such injury was material."[29]

Put another way, this is the question of whether 'injury' is a separable concept. This is a major issue for negotiation; sorting this out will, as a practical matter, move the anti-dumping provisions an important distance toward the concepts of competition law (even though the language may be rather different).

(5) Finally and it will be perceived that these issues are closely related - there is the problem of 'cause'. As explained above, that word means, in US dumping law, that the dumping was a cause of injury, it was one of the causes. Thus injury is the sum of the effects of all the adverse factors impinging on producers, and dumping, if it is one of those factors, is a cause of injury. This is the most protectionist interpretation of the wording possible, but in this writer's view, it is not consistent with Article VI of the GATT. It has not been effectively challenged because the professional anti-dumpers who dominated the Uruguay Round discussions did not want to attack it, and, in any event, the attention being given to a major agreement on subsidies virtually precluded real reform of the injury and causality concepts being employed by the USITC. All students of law are aware that causality is an extraordinary difficult concept; there is an extensive literature. It would be a mistake to import this legal hair-splitting into a negotiation of these GATT phrases. The GATT Articles are attempts to find some relatively simple language which could mean the same things to administrators in many different jurisdictions - it should be interpreted with common sense. An understanding that when real damage, in an economic sense, is being inflicted on an industry or on producers by dumped imports because they are dumped - whether or not that industry is suffering from other outside or inside events, and if it is clear that it is the dumping which gives rise to the competitive harm, then it should be permissible to apply anti-dumping remedies. To achieve this result will require, not a simple change in the wording of the Agreement, but an extensive interpretive note.

It is the conclusion of this paper that reform of the GATT anti-dumping provisions to bring the somewhat nearer the concepts of competition law - and incidentally make them less protectionist in effect - will turn on careful reworking of the concepts of cause, of injury, of material. They are the substantial concepts of Article VI, and only more defensible interpretations of those concepts will make the application of that Article more consistent with competition law, and thereby less protectionist.

Much criticism of the GATT/WTO Agreement, and of the various national anti-dumping systems derived from it, has focussed on the cost to a national economy of such a system. Attempts have been made to calculate the costs to consumers, and, more particularly, to down-stream users of the allegedly dumped imports. E.g. the controversial USITC study.[30] There has been much criticism of the more bizarre aspects of the calculation of the margin of dumping, which, it is alleged, leads to findings of dumping when there is no dumping. Economists concern themselves with these elements in the system perhaps because they can be expressed numerically, with the appearance of precision. It is clear that is very difficult to measure the various costs imposed by an anti-dumping duty, and thus defenders of anti-dumping policy are able to make fairly damaging criticism of such attempts at measurement. It is evident, of course, that anti-dumping, like other barriers to trade, may impose costs - but it is difficult to be really precise as to their magnitude in the longer term. As for the rules about calculating margins, some of the more obviously protectionist practices were negotiated away in the Uruguay Round. It is suggested in this paper that a further round of detailed negotiation of such matters between professional anti-dumpers is likely to yield meager results. This writer has taken part in two rounds of GATT anti-dumping negotiations and has drafted legislation on anti-dumping and has advised various governments on anti-dumping policy. Careful reading of the discussions in legislative bodies (e.g. the US Senate Finance Committee, the Trade Policy Sub-Committee of that Committee), intensive discussions with senior members of such bodies (e.g. US Senators Long and Ribicoff), with their staff and with senior members of the trade law bar in Washington and Brussels - and examination of USITC reports on such matters - have persuaded this writer that it is only by reworking the Agreement texts on cause, on material, on injury, that the system will be materially altered. Of course, these issues arise in regard to the application of countervailing duties as well.

1 "Agreement on Implementation of Article VI of the General Agreement of Tariffs and Trade 1994" in The Results of the Uruguay Round of Multilateral Trade Negotiations (GATT Secretariat, Geneva, 1994) (Results) pp. 168-196 (The "Anti-dumping Agreement")

2 For a description of earlier U.S. measures in the sector, see Hufbauer, Berliner, Elliot: Trade Protection in the United States/31 case studies, Washington, Institute for International Economics, 1986, pp. 154-184. (This account has extensive bibliographical references, including in regard to so-called "trigger-price" mechanism.)

3 Noel Hemmendinger: "Shifting Sands: An Examination of the Philosophical Basis of U.S. Trade Laws," in Jackson, Cunningham, Fontheim (eds): International Trade Policy: The Lawyer's Perspective, New York, Matthew Bender, 1985, pp. 2-1 - 2-10. The "injury only" view is argued, in the subsidies context, in John J. Barcelo III, "An Injury-Only Regime (For Imports) and Actionable Subsidies," in Wallace, Loftus, Krikmian (eds.): Legal Treatment of Domestic Subsidies, Washington International Law Institute, 1984. A more recent, and vigorous statement of this view is Robert B. Reich: "Trading Insecurities," in Financial Times, May 28, 1999.

4 "Undertakings," in the sense discussed in this report, are provided for in Article 8 of the Anti-dumping Agreement. No provision was made for such "undertakings" when Canada introduced revised anti-dumping legislation following the Kennedy Round negotiation. Not surprisingly, after the Tokyo Round, when some revisions to that legislation were required, there was pressure from domestic industry to incorporate a provision for accepting "undertakings." The possibilities of thus creating legalized expert cartels in countries exporting to Canada, with the consent or, at least, the tacit acceptance of the terms of such "undertakings" by Canadian competitors, was attractive to the business community.

5 We may note three recent studies: James Bovard: The Fair Trade Fraud, New York, St-Martin's Press, 1991 (A biography is to be found in the extensive chapter end-notes, and they includes useful references to newspaper articles); Richard Boltuck, Robert E. Litan (eds.); Down in the Dumps, Washington, Brookings, 1991 (a work by various authors, but without a bibliography); Greg Mastel: Antidumping laws and the U.S. Economy, Armonk (N.Y.) and London, M.E. Sharpe, 1998 (a vigorous defence of the U.S. anti-dumping system) with an extensive bibliography; Richard M. Hockman and Petros Mavroids: "Dumping, Anti-dumping and Antitrust," in Journal of World Trade, pp. 27-51 (contains a useful short list of references); J. Michael Finger (ed.): Antidumping/How It works and Who Gets Hurt, Ann Arbor, University of Michigan, 1993 some useful references). See also the annual volume of the Fordham Corporate Law Institute, published by Matthew Bender, New York, for the Institute.

6 For the U.S. system, see the classic work by Earl W. Kintner: A Robinson - Patman Primer/a Guide to the Law Against Price Discrimination, New York, Macmillan, 1979 (2nd Edition). This contains a detailed bibliography and texts of relevant U.S. legislation.

7 See the discussion in Kintner, op. cit., beginning at p. 127. For a general discussion, see OECD Secretariat: Predatory Pricing, Paris, 1989; the references constitute a useful bibliography.

8 Richard A. Posner: Antitrust Law/An Economic Perspective, Chicago, University of Chicago Press, 1976, pp. 188-189.

9 David Palmeter: "United States Implementation of the Uruguay Round Antidumping Code" 29 Journal of World Trade, No. 3, 1995, p. 46.

10 Canadian Competition Act Sector 50 (1) and Director of Investigation and Research: Predatory Pricing Enforcement Guidelines, Ottawa, 1992, pp. 2-3.

11 Cited Kintner, op. cit., p. 366, and see Chapter 5.

12 See Kintner's discussion of "injury to competition" under the Robinson-Patman Act, loc. cit.

13 Cited Palmeter, op. cit.; this article by Palmeter is the clearest discussion of this issue.

14 Palmeter, op. cit. p. 6.

15 Contracting Parties: Basic Instruments and Selected Documents (BISD), Supplement no 41, vol. 1, pp. 229-450.

16 See Kintner, op. cit., p. 140; p. 146, and Canada: Predatory Pricing, p. 13.

17 See Rodney de C. Grey: Injury, Damage, Disruption, UNCTAD/MTN 217, Oct. 1981 for a fuller discussion.

18 U.S.: Trade Agreements Act, 1979; See also Rodney de C. Grey: United States Trade Policy Legislation/A Canadian View, Montreal, Institute for Research on Public Policy, pp. 43-46 for a detailed discussion of "material" during and after the Tokyo Round.

19 "Extramet Industry SA v. Council of the European Communities" [No C-358-89] 11 June, 1992.

20 Cited Hoekman & Mavroides, op. cit., p. 48

21 For a short account of the Kennedy Round negotiations of an anti-dumping Agreement, the only one by a participant, see Rodney de C. Grey: The Development of the Canadian Anti-Dumping System, Montreal, Private Planning Association, 1973. See also the Report of the U.S. Tariff Commission to the Senate Finance Committee on this negotiation: "Committee Print," 90th Congress, 2nd Session, March 13, 1968.

22 Grey, op. cit., note 21, p. 48, p. 73

23 There is an extensive literature; see Grey, op. cit., note 18 for a discussion and references.

24 Richard R. Rivers and John D. Greenweld: "The Negotiations of a Code on Subsidies and Countervailing Measures: Bridging Fundamental Policy Differences" in 11 Law and Policy in International Business, 1979, pp. 1447-95.

25 Palmeter: op. cit.

26 loc. cit.

27 See "Testimony of Counsel to ECAT Before the Committee of Ways and Means / Sub Committee on Trade, U.S. House of Representatives" Jan. 23, 1992.

28 See Grey, op. cit., note 17

29 See Grey, op. cit., note 21

30 USITC: The Economic Effects of Antidumping and Countervailing Duty Orders and Suspension Agreements, Washington, USITC, 1995.