Seeking Transparency in Antidumping Actions through Procedural Review: The GATT/WTO Jurisprudence and Its Implications for China (Part I)

Dongsheng ZANG

Perspectives, Vol. 2, No. 5

"Dumping" under international trade law in a simple definition means to export products from one country to another at a price below its normal value. The General Agreement of Tariff and Trade (GATT), now part of the World Trade Organization (WTO) agreements, allows member countries to take actions to counteract the effects of dumping once a proper determination of both dumping and injury is made. The first national antidumping law was adopted in Canada in 1904; Australia, New Zealand, Great Britain adopted similar laws in the following two decades. The first modern antidumping law in the United States was the 1916 Act, and the current basic structure was adopted in 1921. In mainland China, the Foreign Trade Act of 1994 authorizes that anti-dumping measures be taken when dumping is found and injury is determined. Based upon the above Foreign Trade Act, the State Council adopted mainland China's first antidumping and countervailing regulations in March 1997, which prescribes basic rules for definition and calculation of dumping, injury determination, investigation procedures, antidumping measures, etc. Chinese Taipei established their antidumping rules earlier. Although their structures are not identical, they share a lot of commonalities with those in mainland China.

I. China's Uneasy Relations with Antidumping Actions

Antidumping is not a foreign term for the Chinese during the past two decades. From time to time, public media reports about China's exports being under antidumping investigations by China's major trading partners, typically the United States and the European Union. It is widely believed that mainland China might have become the largest victim of antidumping actions by her trading partners. Recently the International Finance Daily reported that in the past decade, World Trade Organization (WTO) member countries launched 246 antidumping investigations of mainland China's exports to them, and affirmative final results were found in approximately 60 percent of these investigations. As a direct consequence of the anti-dumping actions, according to the report, China's exporting industries could have suffered at least 10 billion U.S. dollars.

In the 1980s, when the Japanese economy was expanding rapidly in the international market, Japan was the top target of antidumping investigations in the United States. Even then, mainland China, Taiwan, together with South Korea, were among America's top ten targets of its antidumping actions. Statistics from the U.S. Department of Commerce suggests that in the 1990s mainland China has replaced Japan to become the top target of its antidumping actions, and it may remain so in the near future. According to a United States report filed with the WTO, by June 30, 2000, there are 42 antidumping orders against mainland China in effect in the United States, 22 against Chinese Taipei, 16 against South Korea, 46 against Japan. There is no doubt that the Greater China economies are the largest antidumping target in the world. The sense of victimization is not only a great frustration to the believers of David Ricardo's comparative advantage in theory and the exporting industries in practice; it may also pose a serious challenge to the export-oriented strategy in its economic development, which is often considered as the secret of the East Asian miracle by some economists. Currently, China is working hard on negotiations for accession to the WTO, and it is hopeful that China can join the WTO in the year of 2001. Before the accession, China only has observer status at Geneva. For a long period of time, China will probably continue to face a dilemma with respect to antidumping investigations. On the one hand, not being a member of the WTO club, China cannot use the WTO dispute settlement mechanism, and therefore cannot fully take advantage of the WTO regulatory framework. On the other hand, by the time she's acceded to the WTO, she will be under the WTO rules to make her home market more open and will be under the scrutiny of the WTO framework. She would need a set of more effective domestic antidumping rules to protect the domestic market. In both two cases, she definitely needs to know more about the WTO antidumping rules. There are additional reasons to know more about WTO antidumping rules. More often than not, in wake of waves of antidumping actions in their foreign markets, governmental agencies in Beijing are handicapped by the inaccessibility of WTO dispute settlement mechanism; on the other hand, domestic exporters or manufacturers usually have little enthusiasm in defending their legal rights through taking part in the antidumping proceedings in the importing countries. A typical reaction of the exporters is to decide not to respond at all. Thus, it seems that how to handle China's difficulty with antidumping actions initiated by her trading partners is more than a governmental matter.

This paper is an effort to clarify some key characters of the GATT/WTO antidumping adjudication and its underlying jurisprudence. My focus is how GATT/WTO has tried to regulate use and misuse of antidumping actions in member states through what I call procedural review. In technical aspects, I wish, through analysis of the relevant doctrines in procedural review in different part of GATT/WTO history, to demonstrate a shift of GATT/WTO jurisprudence which is being made more explicit in recent years. Much of the information in my analysis will probably be good news for the exporting industries in China, but I do not want to leave an over-optimistic impression that an accession to the WTO would solve all the current concerns. There are, still, good reasons to be concerned, and I will discuss these concerns and their implications at the end of this paper.

II. The Diplomacy Period

What China has bitterly experienced with the anti-dumping investigations is not unique in the history of international trade; rather, it is quite typical. In the earliest anti-dumping case under GATT Article VI, Swedish Anti-dumping Duties, Italy, the complainant, argued that the basic price standard in Swedish anti-dumping system systematically discriminated the low-cost producers. Not finding the Swedish calculation methodology inherently discriminatory, the GATT panel nevertheless agreed with Italy that the antidumping proceedings, which were usually complicated and time-consuming even as early as in 1954, "[had] a serious discriminatory effect if consignments of the goods exported by the low-cost producers had been delayed and subjected to uncertainties by the application of that system and the case for dumping were not established in the course of the enquiry." It turned out that the GATT panel was telling a very important truism about the real economic and legal functions of the anti-dumping investigations. The idea of anti-dumping may sound reasonable-the importing countries are entitled to protect their own domestic markets, but the more technical and yet crucial questions cannot be ignored: who is going to define key terms such as "industry," "normal value," "injury," etc. and how? It is for this reason that what China's exporting industries have experienced in the last two decades is not unique; they are just facing a frustration that almost every exporting country has been and probably is still facing today. From the 1970s, antidumping measures were taken more frequently. There were a lot of complaints about antidumping regulations and their administration in the U.S. Inside the U.S., free trade economists and lawyers never hesitate to condemn the protectionist inclination in the antidumping regulations and their administration.

The problem, however, is more than the protectionist nature of antidumping regulations and their administration; without external restraint, all governmental agencies have incentives to protect their domestic market from outside competition. Article VI of GATT 1947 was designed to be an external restraint, but it proved that it was too vague and simple that its structure had left too much room for the investigating authorities of the importing countries to manipulate for protectionist purposes. This problem was not unrecognized, however. In April 1959, a Group of Experts was established to study the legal issues of antidumping and countervailing duties under Article VI. The Group warned against "immoderate use" of anti-dumping law, and stressed "the need for governments to use their anti-dumping powers with great care . . ." However, it is also clear that at this moment, procedural restraint of the investigating authorities was not the top priority for the experts in Geneva; rather, the priority was given to the substance of key conceptions in Article VI. The "main concern" of the Group of Experts was to fill in the gaps in the substantive framework, or in their own words, the "technical requirements of existing legislation". The idea was that the abuse of antidumping laws by member countries can be curbed by standardization of the substances of the rules. Many of these efforts were made through a political and diplomatic approach-diplomatic coordination of lawmaking and administration and political dialogues about trade issues. Various groups were set up to study the issues on antidumping, coordinate national legislations, and propose practice standards to member countries. These groups include the Group of Experts established in 1959, as mentioned above, a Committee on Anti-dumping Practices, established in 1969; a Working Party on Acceptance of the 1967 Antidumping Code, which was set up in September 1970. Through meetings, discussions in these groups, and regular reports to the CONTRACTING PARTIES, in a careful and informal way, great efforts were made, consciously and unconsciously, to harmonize the national legislations and practices. One of the milestone achievements in this period of time is what was then called Kennedy Round Anti-dumping Agreement, which we now usually call the 1967 AD Code. This document is an official codification of many of the standard practices that were extensively discussed and debated in those study groups and their reports. Compared with GATT Article VI, it offered much more detailed anti-dumping rules, definitions, standards for key concepts such as "dumping," "industry," "injury," and causation. The 1967 AD Code also embodied the diplomatic philosophy of this period by making it an obligation of a member to keep other members informed of any changes of its antidumping law, and to report to the other members the administration of its laws on an annual basis.

There was a debate within GATT whether it should abandon diplomatic approaches and move towards more legalism. Compared to later developments that I'll discuss later, the antidumping rules in this period were closer to what Professor Hudec had suggested. That is, GATT generally preferred standards to rules, and member countries wanted these standards to be normative than to be legally binding. However, Professor John Jackson rightly pointed out that compared to other multilateral agreements, GATT already represented a strong tendency toward legalism. As a result, dispute resolution under GATT did not play an active role during this period, as member countries were reluctant to file a formal complaint despite their tremendous displeasure.

III. The Establishment of GATT Panel Review

The 1967 AD Code, however, did not go into full operation due to legal/political reasons in the United States. The direct legal problem was that the President was not under express delegation from Congress to enter into the negotiation. On the more substantive matter, the 1967 AD Code was perceived by the Congress inconsistent with U.S. domestic law. Subsequently Congress enacted implementing legislation that authorized the U.S. Treasury Department and International Trade Commission to implement the 1967 Code only to the extent it was consistent with prior domestic practice. This reservation effectively undercut the 1967 Code's acceptance in the GATT community. By the time the GATT members had agreed on a new code, the 1979 AD Code, it was under a quite different social and political context.

First, there was a revival of neo-liberalism in economic theory by the end of 1970s and in the early 1980s. The following liberalization of trade policy advocated or encouraged by the international institutions such as the IMF and the World Bank left the GATT a new role to play in regulating trade issues: it became an agent to implement liberalization trade policies. One such policy supported by the World Bank was transparency. In practice, the assumption of this revived neo-liberalism and its corresponding jurisprudence was, and still is, that regulations are considered barriers to free trade.

Second, the GATT dispute settlement mechanism in this period was undergoing a profound transition from diplomacy toward legalism. Before the 1979 GATT dispute settlement procedures was adopted, dispute settlement at GATT was based upon Article XXIII:2 of GATT 1947, which even did not say anything about panels or working parties. In practice, however, both were set up to handle trade disputes between GATT members. Since early 1950s, the GATT panels gradually displaced the "working party" as the normal procedure for dispute resolution. Though for a long time, GATT panel dispute settlement shared a lot of common characters with conciliation and arbitration, it made a big move towards formal adjudication as a consequence of the Tokyo Round negotiations, when the "Understanding" was adopted. Adjudication procedures were further formalized and strengthened in 1985 and 1989. The 1989 "Improvement Rules" made further rooms for professionalization of the GATT panels by allowing non-governmental individuals to join the governmental panelists, and it was even declared that "[t]he roster of non-governmental panelists shall be expanded and improved."

Third, as part of the larger movement in this period, there were growing concerns over procedural issues within antidumping rules and their administration. Though on substantive issues, the 1979 AD Code was largely based upon its predecessor of 1967, it nevertheless had clear and important departures from the traditional GATT approach. Characteristically, Article 3, determination of injury, was restructured and as a result, was made more like prescribed procedures for decision-making than guidelines in the 1967 Code. Another important breakthrough was the 1979 Code, which had a dispute resolution clause, Article 15, that enabled parties in an antidumping dispute to submit the issue to a GATT panel. The significance of these departures was soon demonstrated by a GATT panel decision in New Zealand-Imports of Electrical Transformers from Finland.

The legal question in Transformers raised in front of the GATT panel was simple and yet fundamental: whether the determination of injury should be subject to scrutiny of either the exporting party or the CONTRACTING PARTIES in general. In 1982, at the request of a domestic producer, the New Zealand customs authorities initiated an antidumping investigation against a Finnish exporter regarding its sale of two custom designed and manufactured electrical transformers. In February 1984 the New Zealand authorities made the finding that the two transformers in question were imported at less than their normal value and that their importation had caused material injury to the transformer manufacturing industry in New Zealand. An antidumping duty was subsequently imposed. New Zealand contended before the GATT panel that its determination were not to be subject to Finland or the GATT panel, because "the clear words of the Article [VI:6(a) of GATT 1947] did not permit any other party or body to make the determination formally required by this Article or to review its basis. That was a matter falling exclusively within the competence of national authorities." Finland, however, disagreed. It made a utilitarian argument by contending that New Zealand's view would give the "importing country full discretion regarding the injury determination. Such an interpretation would nullify GATT practice and discipline concerning injury criteria. It would open the door to anarchy without the possibility of international surveillance."

This is not the first GATT panel review of antidumping measures. In Swedish Anti-dumping Duties, a GATT panel was asked to review a Swedish antidumping decree of 1954. However, in terms of legal doctrine, Swedish Anti-dumping Duties is too ambiguous on the matter of "reviewability." Sweden made an argument similar to that made by New Zealand in Transformers: that it was "the right of the national authorities to decide whether dumping had really taken place," therefore Sweden was doubtful that the CONTRACTING PARTIES could not "consider that question." And the Panel quickly affirmed this by saying that, "[t]he Panel agreed that no provision of the General Agreement could limit in any way the rights of national authorities in that respect." However, the Panel did not want to give up, it rephrased the issue, that, "it would be reasonable to expect from the contracting party which resorts to the provisions of Article VI, if such action is challenged, to show to the satisfaction of the Contracting Parties that it had exercised its rights consistently with those provisions. The Panel felt therefore that, in order to decide whether Italy had suffered an impairment of benefits, it would be appropriate and necessary to examine whether a case had been made out that Italian stocking exporters had resorted to dumping practices." To make it more confusing, the Panel reviewed the price calculation process without using the typical procedural standards; and at the end of the review, it did not make any finding; rather, the Panel was more like a mediator who finally found a best solution to the difficult question by suggesting that Sweden should send a "responsible official" to Italy to seek more information. Swedish Anti-dumping Duties stands for the typical diplomatic approach under traditional GATT jurisprudence that have been identified by Hudec. Even the language used was much "soft"-the Panel used such words or phrases as "felt," "it appeared to the Panel," "[t]he Panel considered that it was not competent to . . ." much more often than using "noted," or "held," if at all.

This jurisprudence in Swedish Anti-dumping Duties, however, was not what the GATT panel wanted in the 1980s; rather, they were probably the very obstacles that the "new" GATT panel wanted to overcome. The GATT panels were being empowered both by the neo-classical economics and the Tokyo Round trade negotiations, they had new roles to play, new grand plans in mind: they wanted and had to be more assertive. This was exactly what the Transformers Panel did, when it made the following clear statement: ". . . the Panel believed that if a contracting party affected by the determination could make a case that the importation could not in itself have the effect of causing material injury to the industry in question, that contracting party was entitled, under the relevant GATT provisions, . . . that its representations be given sympathetic consideration and that eventually, if no satisfactory adjustment was effected, it might refer the matter to the Contracting Parties, as had been done by Finland in the present case. To conclude otherwise would give governments complete freedom and unrestricted discretion in deciding anti-dumping cases without any possibility to review the action taken in the GATT. This would lead to an unacceptable situation under the aspect of law and order in international trade relations as governed by the GATT." (emphasis added) The Panel expressly stated that the GATT panel review was an entitlement. This was more than what Finland had asked for-the Panel was declaring that procedural protection as an entitlement was a matter of principle, not a matter of utilitarianism. The Panel was quick in adding, after the above statement, that it was actually following the Swedish Anti-dumping Duties. But this seemed to be cosmetic at most. There were at least three substantial differences between the two cases that would help to indicate the significance of Transformers. First, in Transformers, the central issue in question was a procedural one-how the New Zealand authorities made findings of dumping and injury, while the dispute in Swedish Anti-dumping Duties was more focused on a substantive issue in the Swedish decree, the basic price system. Therefore, if the GATT Panel in both cases dealt with "reviewability" issue, what the Swedish Antidumping Duties Panel did was, primarily, to review a piece of legislation, which was a typical job for GATT panel and committees or expert groups since the 1950's; what the Transformers Panel did, however, was, primarily, to review the administration of New Zealand's domestic trade rules, the legislation as such was not in dispute. Therefore, in terms of legal doctrine, the standard of review in Transformers was much higher; it pointed to a great departure from GATT's traditional function as a coordinator; the new GATT panel was much more intrusive. Second, in its review of administrative finding process, the Swedish Anti-dumping Duties Panel noted that Swedish authorities did not have "definite evidence" to support its conclusion, but was nevertheless reluctant to make a finding; while the Transformers Panel found inconsistency with GATT because "New Zealand had not been able to demonstrate that any injury suffered by its transformer industry had been material injury caused by the imports from Finland." Clearly, the burden of proof of consistency was shifted from the complainant to the respondent. This is related to the third perspective, i.e., the empowerment of the GATT panel. In contrast to all its modesty in Swedish Anti-dumping Duties, the Transformers Panel exercised its authority to interpret the GATT anti-dumping practice and empowered itself. In its review of injury determination, the Panel analyzed the shares of the two Finnish transformers among the overall sales of transformers in New Zealand and made a finding that they were "insignificant", thus "the cause of this injury could not be attributed to the imports in question from Finland. . ." But what was the standard of "significance"? It was, after all, what defined by the Panel, first by redefining the scope of "industry." Thus, one important consequence of the move towards legalism was to shift authority from the member states to the GATT panel, just like in English legal history, Bentham's legislation reform shifted the power from the courtroom to Parliament. In terms of legal doctrine, what matters in Transformers is not who is showing more wisdom in the way of defining "industry," but who is doing the defining job.

The Transformers case is another milestone in the history of GATT anti-dumping regulatory framework. It established that the GATT panel had the power and authority to review member state antidumping rules and administration of those rules. It articulated for the first time the standard of review by the GATT panels in dispute settlement. However, this is only a starting-point in the development of GATT/WTO jurisprudence in this aspect. It didn't offer elaborated doctrine, except for a brief statement of the principle. It left a lot of room open, as we shall see in the following parts of this paper.

IV. The Rise of Procedural Review at GATT

The 1979 Code and Transformers had the infrastructure for procedural review ready, but it was not the sufficient reason for the full operation of GATT dispute settlement system; there had to be the some driving forces to push the system into operation. In antidumping cases, the first round of procedural review did not occur until early in the 1990s. By "procedure review," I do not mean that the parties or the GATT panels were only concerned about procedural issues; there were tremendous concerns over substantive issues, and this is still true even today. What interests me is that, in addition to the substantive policy issues, the parties were increasingly raising procedural issues and the GATT panels were making "common law" by interpreting the procedural rules in the 1979 Code. The efforts put on the procedures not only suggest the growing power of the GATT panels, in that sense the GATT panels were becoming more self-conscious and self-assertive; what is no less true is the other side of the coin: the efforts put on refining procedures are also the means to gain power and legitimacy-at the end of the day, what could be better differences between a diplomat (a politician) and a judge (a professional)? Thus the efforts on procedures suggest the growing professionalism in the way GATT adjudication was conducted.

A. Initiating an Antidumping Investigation

The standing question in a typical antidumping investigation is who has the right to petition for such an investigation. Article 5.1 of the 1979 Code provides that, "[a]n investigation to determine the existence, degree and effect of any alleged dumping shall normally be initiated upon a written request by or on behalf of the industry affected." From the text of Article 5.1, it is not clear whether an investigating authority of a member state has complete freedom to initiate an investigation. Who is qualified to represent the industry in question? Should the investigating authority consider the "representativeness" of the petitioner? Is the investigating authority acting inconsistently with Article 5.1 if it does not make a finding on "representativeness" in the first place? These questions were raised in the GATT antidumping cases in the early 1990s, many of which were against the U.S.: U.S.-Steel Products from Sweden, U.S.-Cement from Mexico, and U.S.-Salmon from Norway. The first two cases were not adopted by the Contracting Parties as a consequence of objections by the U.S., but similar procedural issues were discussed in the last case, which was adopted in 1994.

Standing may sound very much a pure procedural issue, but here, as elsewhere, it is not difficult to observe the close relationship between the procedural questions and the substantive issues. In U.S.-Steel Products from Sweden, Sweden made this clear when it argued that "Sweden was of the view that the failure of the [U.S.] Department of Commerce to examine on its own initiative the representativeness of a petition deprived exporters of the protection afforded by the Agreement, and Article 5 in particular, against frivolous investigations." Even though, the question raised by Sweden was more technical: Sweden argued that it had not contended that the U.S. legislation as such was inconsistent with the Agreement, "[w]hat Sweden criticized, in light of the requirements of the Agreement, was the passive role of the Department of Commerce with respect to the question of the standing of a petitioner." Thus, what Sweden was asking for was that the DOC should have made a finding as to the standing of the petitioners, which leaves the impression that Sweden was insisting on a pure procedure request in DOC's decision-making here. The U.S. counter-argued that the procedural requirement demanded by Sweden was not contemplated by the plain text of Article 5.1, but its main defense missed the point-it was basically trying to show why the substantive issues were more important and the procedure issues could either be ignored or be cured. It argued that the petitioner, Specialty Tubing Group (including its member companies), was "unquestionably comprised of manufacturers" of like products, that the countervailing investigation, which was going on by the time the petition for antidumping investigation was filed, would have given the investigating authorities all the data about the domestic producers; that there was no objection during the whole process of antidumping proceedings; that a co-petitioner, the United Steel Workers of America, which joined the Specialty Tubing Group after the antidumping investigation had been started, represented "a preponderance of" workers in the industry. All these arguments made by the U.S. did not convince the GATT panel. The Panel's reading of Article 5.1 was quite simple: if Article 5 allowed a petition to be filed by someone "on behalf of" the domestic industry, that meant the request should have been authorized or approved by the industry, and that was why the investigating authority should have satisfied itself about the representativeness before the investigation was to be initiated. What I want to suggest here is not how the U.S. misargued their case in front the Panel; my interest is not in the wisdom of the Panel's decision either. What is more interesting is the emphasis put on the procedures here: the Panel's interpretation of Article 5.1 seemed have very little to do with the substantive facts that the U.S. took so much pain to establish. What the U.S.-Steel Products from Sweden Panel decided was a procedural requirement-a procedural requirement in its strong sense: the procedure is required even though substantive issues can be solved or cured by other ways. And this is a sharp contrast with the GATT panel in Sweden Anti-dumping Duties.

B. Positive Evidence and Objective Examination

One important change from 1967 Code to the 1979 Code on injury determination is that Article 3.1 of the 1979 Code required that "[a] determination of injury for purposes of Article VI of the General Agreement shall be based on positive evidence and involve an objective examination. . ." As I have hinted earlier, Article 3 itself a great move toward procedural control. Article 3.1 has a stronger element of administrative law-it sounds like the "substantive evidence" rule under the United States Administrative Procedure Act. But, what is "positive evidence"? What is the standard for it? Who defines it? There seems to be no direct answer from the text of Article 3.1 to these important questions.

The question of positive evidence was raised but not addressed in U.S.-Cement from Mexico. The GATT Panel waited until later to address the issue, in U.S.-Salmon from Norway. In this latter case, Norway made a long argument against the U.S. antidumping investigation based on the positive evidence rule in Article 3. It dug into all the technical details in the decision-making process, data, calculation methodology, etc., as to the determination of dumping, injury, and causation between dumping and injury, claiming the lack of positive evidence. Here again, there is a close relationship between procedure and substance. It is probably more obvious than the previous case we discussed. Under Article 3.2, the investigating authority is obliged to consider whether there is "significant increase in dumped imports" in injury determination. In this case, Norway argued that the significance in the data collected by the investigating authority, the US ITC, was limited because it failed to consider the decline of imports right after the investigation period. Is this a procedural issue or a substantive issue? It is hard to tell.

Therefore, for the GATT panel, it is a great adventure to be involved in such details. Strategically, it has to secure at least two critical elements for its continuous functioning: legitimacy and relevancy. Like any adjudicatory body in a domestic legal system, on the one hand, it cannot totally ignore the real issues, the substantive issues. It has to step into conflicts. Otherwise it would become irrelevant. On the other hand, it has to restrain itself. It has to make principled decisions. It has to find its secure harbor for it to avoid political attacks, which would undermine its legitimacy. So far, the best strategy in combination of these two elements in the twentieth century is professionalism, the typical slogan of which in the field of law in the past two or three centuries is procedure. This seemed to be what was beneath United States-Salmon from Norway. In the first place, the GATT panel took the challenge by making astonishingly detailed inquiries (by the standard of GATT in that period of time). In that sense, it was demonstrating equally surprising courage as the panel in United States-Steel Products from Sweden did. What is more important, however, is that it also demonstrated reason in its inquiries. It made an important distinction between a de novo review and a procedural review: it articulated that "[t]he question of whether a determination of injury was based on positive evidence therefore was distinct from the question of the weight to be accorded to the facts before the investigating authorities." Thus, by making the distinction explicit, the Panel chose the principle of deference for procedural review.

C. Administrative Record and Public Notice

In Korea-Resins from the U.S., in question was whether the GATT panel could take into account a confidential transcript which was not included in the public notice issued by the South Korean investigating authorities. The U.S. arguments were based upon Articles 3.4 and 8.5 of the 1979 Code. Article 8.5 required public notice given when a determination had been made. "In the view of the Panel, effective review . . . of an injury determination against the standards set forth in Article 3 required an adequate explanation by the investigating authorities of how they had considered and evaluated the evidence with regard to the factors provided for in that Article." "An explanation of how in a given case investigating authorities had evaluated the factual evidence before them pertaining to the factors to be considered under Article 3 clearly fell within the scope of the requirement in Article 8:5 that authorities articulate in a public notice 'the findings and conclusions reached on all issues of fact and law considered material by the investigating authorities, and the reasons and basis therefore.' This provision served the important purpose of transparency by requiring duly motivated public decisions as the basis for the imposition of anti?dumping duties."

In retrospect, it seems that the idea of "transparency" was still somehow ambiguous: transparency for whom-the private parties, or the governmental agencies? For what purpose? The ambiguity on one hand stems from the fact that at this early stage, the GATT panel did not have a clear goal in its procedural review. In this sense, the ambiguity tells a lot about the ambiguity in procedural review in general. On the other hand, the ambiguity seems to suggest the GATT's underlying inclination for centralization through its panel review. In this sense, it probably suggests, at this stage of development, the GATT adjudication was still much influenced by the style of inquisitory adjudication in the European Continent. If this was true, this style was going to change soon.

(The author is an S.J.D. candidate at Harvard Law School. This paper was presented at the Conference on China and the World Trade Organization held at Australian National University in Canberra, Australia, March 16-17, 2001. All footnotes and references are omitted and are available from the author upon request.)