"حقایق موجود" ادعاهای دامپینگ : چه زمانی شرکت های خارجی در دادخواست ضد دامپینگ همکاری خواهندکرد؟

Foreign firms accused of dumping in the World Trade Organization (WTO) system may face punitive duties if they do not cooperate with domestic investigative authorities. The punitive tariffs are typically based on domestic firms' allegations, or so-called “facts available” dumping margins. This paper considers a number of questions relating to “facts available” dumping allegations. Given that governments' use of facts available margins are credible, why do foreign firms still sometimes choose not to cooperate? Furthermore, why do domestic firms, knowing that an administering authority might use their allegations to determine punitive duties, not always announce alleged dumping margins high enough to eliminate all imports? Is it always in the domestic firm's interests to force foreigners not to cooperate and thereby be “punished” with facts available margins? And what is the administering authority's optimal decision concerning compliance costs faced by foreign firms?

Antidumping has been one of the most frequently researched areas of trade policy over the last 20 years. This reflects the fact that antidumping duties are the most common form of import relief under the General Agreement on Tariffs and Trade (GATT) and the current World Trade Organization (WTO) system. Economists have analyzed many aspects of antidumping, both theoretical and empirical.1 One area of importance has not received scrutiny in the economics literature. WTO rules (and the GATT before it) allow governments to use dumping margin allegations provided by the domestic petitioners in assessing tariffs on foreign firms that do not cooperate in “unfair trade” investigation. In particular, the Antidumping Agreement concluded in the Uruguay Round of trade negotiations states that: In cases in which an interested party refuses access to, or otherwise does not provide, necessary information within a reasonable time or significantly impedes the investigation, [decisions] may be made on the basis of facts available.2 An annex to the agreement adds that the authorities will be free to make determinations based on facts “including those contained in the application for the initiation of the investigation by the domestic industry.”3 The justification for this provision is simple. The calculated dumping margin normally is based on a comparison of the export price with either the price charged in the exporter's home market or its production costs. However, foreign prices or costs are typically privately held information of the exporter so that foreign firms need an incentive to cooperate in an investigation. This incentive in practice has been provided by the potential use of the domestic firm's allegations (called “best information available,” or BIA, under the GATT rules) about foreign firms' dumping margin. Because the domestic firm has an obvious incentive to overstate the dumping margin when laying out its case, the exporter faces the possibility of very high duties if it does not cooperate. Critics of this system's application in the pre-WTO US system have noted that the average dumping margin when BIA was used was much higher than calculated rates. Baldwin and Moore (1991) report that, for the period between 1980 and 1990, the average final dumping margin imposed by the Department of Commerce (DOC), the relevant US authority, using BIA methods was 67% (36 cases), while the average duty based on foreign firm provided information was only 28% (188 cases). One of the main problems in the United States, critics such as Murray (1991), Palmeter (1991) and Finger and Artis (1993) have argued, was that the Department of Commerce often made compliance very difficult for foreign companies (including 200-page questionnaires, tight and inflexible deadlines, requirements to report data using US-style accounting techniques, and on-site investigations of all records by US investigators). In addition, if foreign firms were found by the DOC to be only “partially cooperative,” all information provided by the firms could be thrown out, with BIA used in its place. After considering these procedures, many critics saw the BIA provisions as a means by which domestic firms could use the system, not to elicit truthful announcements about foreign costs and pricing behavior, but instead to bring duties on foreign firms far in excess of the true dumping margin. Supporters argue that this “stick” is the only way to achieve compliance. Stewart (1991) and Mastel (1998) note that administrators determining dumping margins are confronted with the problem that their decision in principle relies largely on data provided by the foreign firm. Without BIA, foreign firms could thwart the investigation simply by withholding data so that these provisions are the only way for an administering authority to obtain the true level of dumping. Finally, supporters in the United States note that the DOC does not always use BIA so that it is unfair to allege that BIA margins are simply a means to force high tariffs onto innocent foreign firms. Indeed, information provided by Baldwin and Moore is consistent with this; only 16% of final dumping margins assessed during the 1980s in the United States used BIA information. In Moore (2002), I found that many foreign firms subject to new sunset review procedures in the United States were originally subject to BIA. In particular, 131 firms in a sample of 314 firms subject to sunset reviews did not cooperate sufficiently in the initial investigations to avoid BIA. Those subject to BIA procedures had average antidumping duties of 64% compared to only 34% for cooperating firms. Furthermore, of those 131 firms, 71 did not cooperate in subsequent administrative reviews. Clearly, there are substantial numbers of foreign firms that do not find it in their interest to cooperate with US authorities at various stages of antidumping administration. While complaints about BIA led to some reform of the process in the Uruguay Round, the use of domestic allegations in antidumping duty calculation remains an important part of the system. The most notable reform was that “reasonable attempts” to fulfill the information requests must be recognized. However, the domestic administering authorities still may use petitioners' allegations when foreign firms do not cooperate “sufficiently.”4 US law implementing the Uruguay Round reflected these changes. One particularly cosmetic change was that the DOC no longer uses the term “best information available” but instead “facts available.” Yet the DOC's new regulations made clear that the use of “adverse inferences” could be justified if the foreign respondent was perceived as deliberately uncooperative, an approach fully supported by the Congress and the President. In addition, the Department of Commerce continues to use margins calculated using particular now-prohibited BIA methods under the GATT system in the recently inaugurated US “sunset review” process, which was designed to evaluate whether antidumping orders should remain in place 5 years after their initiation (Moore, 1999). Certainly, the “stick” of “facts available” data remains a very significant threat. In recent cold-rolled steel antidumping cases, the DOC used “facts available” in all but two of the cases for the final dumping margin.5 The margins (which ranged from 46% to 64% for Brazil, 53% for Japan, and 109–163% for Slovakia) yielded significantly lowered imports from the targeted countries.6 Despite the importance of BIA in the administration of antidumping law, no systematic study of the use of BIA appears in the economics literature. Economists analyzing the US antidumping process typically argue that, because the Commerce Department almost always finds a positive dumping margin in its investigations (nearly 95% of cases), the focus should be on the injury decisions by the US International Trade Commission. However, this ignores the fact that the DOC determines the specific duty level, often using “facts available” margins, and thereby the restrictiveness of the import regime. This paper will begin to fill this gap. A number of questions will be addressed. First, if the use of BIA is credible (which it surely must be, at least in the United States given the history of AD administration), why do foreign firms still sometimes choose not to cooperate? Second, if domestic firms know that an administering authority might use their allegations, why not always announce alleged foreign margins high enough to eliminate all imports? Is it always in the domestic firm's interests to force foreigners not to cooperate and thereby be “punished” with BIA margins? Finally, what is the administering authority's optimal decision concerning compliance costs faced by foreign firms? I will address these questions in a Bertrand–Nash model in which a foreign firm endogenously chooses whether to cooperate in an antidumping investigation.7 Its choice will depend on the domestic firm's announcement about the alleged foreign dumping margin while the authority sets the costs of the foreign firm complying with its investigation. The approach in this model is similar to Prusa (1992) who models the choice of firms between withdrawing an antidumping petition and pursuing it to its final conclusion. However, my model analyzes an earlier part of the antidumping process, in particular, whether the foreign firm will cooperate with the investigation in the first place. Three types of behavior by the domestic authority will be addressed. I first assume that the authority sets an unchanging level of compliance cost. I then analyze the outcome when the authority chooses the compliance cost strategically but is indifferent to the impact of the duty on domestic consumers. Finally, I assume that the authority cares about both consumer and producer welfare. Each of these scenarios yields a different combination of compliance cost, tariff rate, and frequency of foreign firm cooperation. The focus is on aspects of the relationship cited by critics, i.e., that “facts available” information is a means for domestic firms to achieve higher duties. I do not consider how the authority might use “facts available” to elicit accurate information about foreign costs. While this latter issue is important, it is best analyzed using a game between the authority and foreign firm with asymmetric information. This would be a useful extension to the present research.8

This paper provides tentative conclusions regarding issues that have been overlooked in the antidumping literature in economics. The incentives for foreign firms to cooperate in antidumping investigations have been analyzed when the firms face the possibility that an administering authority will use allegations by the domestic petitioner when setting antidumping duties. The need for study in this area is evident given the very high margins that have at times arisen in US antidumping cases when so-called “facts available” have been used by the Department of Commerce. A number of questions have been addressed. Why do foreign firms sometimes not cooperate in antidumping investigations given the clear authority for governments to counter with punitive tariffs? In addition, should domestic firm always allege prohibitive tariffs, given the possibility that the domestic authority might use these allegations against noncooperative firms. Finally, what are the domestic authority's motivations and strategies when setting compliance costs? Within the framework introduced here, foreign firms have ambiguous and conflicting incentives concerning possible cooperation. On one hand, high compliance costs reduce the attractiveness of pursuing a “cooperative” antidumping path but a decision to forgo cooperation means that the “stick” of facts available margins may be placed on their imports. A foreign firm may choose not to cooperate even if it is certain to win its case as long as compliance costs are positive. This might be particularly problematic for small firms where compliance costs are high relative to any expected benefits from cooperation. In general, foreign firms that do not cooperate may have nothing “to hide.” Instead, they may simply be responding to the incentives presented by two possible paths of antidumping litigation. For some values of compliance costs, a domestic firm would rather face a cooperating foreign firm while for others it would rather compel the foreign firm to face the “facts available” dumping margin. It is not always in the domestic firm's interests to allege extremely high dumping margins by its foreign competitors, as long as the foreign firm has the option of cooperating with the domestic investigation. Certainly, the presence of a path of possible cooperation limits a domestic firm's ability to completely manipulate the antidumping process through allegations of very high foreign costs. The most important results are obtained when the decisions of the domestic authority are modeled. I have identified three distinct patterns of tariffs, compliance costs, and degree of foreign competition. In Model 1, if the domestic authority cannot or does not want to vary compliance costs, perhaps because of legislative or judicial constraints, there will be not be cooperation from all foreign firms, but with higher (lower) duties for noncooperating (cooperating) firms. If instead the authority has the freedom to set compliance costs, the outcomes depend critically on the degree to which consumer interests enter into its objective function. If the authority is indifferent to consumer losses (as in Model 2), compliance costs will be set so high that there will never be cooperation and domestic firms' allegations of dumping margins will lead to prohibitive tariffs. If the authority does care about consumer effects (as in Model 3), then nonprohibitive tariffs will be observed but compliance costs will vary greatly, with extremely low costs when foreign firms cooperate and higher compliance costs when firms do not cooperate. In principle, the data should reveal which prediction best fits actual practice. In the United States, for example, facts available information is used often but not exclusively. This would suggest that either the Model 1 (nonstrategic behavior) or Model 3 (consumer-sensitive) scenario. In principle, one should be able to distinguish between Models 1 and 3 by considering actual compliance costs. If the costs vary nontrivially across cases, there is support for Model 3; if costs remain unchanged, the DOC may simply be precluded by judicial review from acting strategically. Unfortunately, it is difficult to collect useful data on compliance costs. Nonetheless, this may be a useful exercise for later research. Model 3 as a possibility may surprise observers familiar with complaints about alleged DOC bias in favor of domestic petitioners in antidumping cases. The patterns predicted in the model are; nonetheless, not consistent with a world in which the DOC arbitrarily sets compliance costs to benefit domestic firms so that some of the DOC's critics' accusations may be unfounded. Certainly, the model abstracts from many complications. Modifications would include making fully endogenous the probability of antidumping success and the level of tariff imposed in the antidumping petition so the entire government's problem is solved. However, as argued above, varying these “parameters” may alter the particular profit levels facing the domestic and foreign firms but they would not change the basic problem that domestic firms make dumping allegations in a world where the foreign firm might respond by cooperating. One might also design an asymmetric information framework wherein the domestic authority provides an incentive compatible mechanism to induce truth telling by the foreign firm. Of course, such a mechanism should be credible so that one would always see cooperation. Nonetheless, this research takes a first step towards building “facts available” incentives into a model of antidumping procedures.