UNITED STATES - SAFEGUARD MEASURES ON
IMPORTS OF FRESH, CHILLED OR FROZEN
LAMB MEAT FROM NEW ZEALAND AND AUSTRALIA

AB-2001-1

Report of the Appellate Body

(Continued)

VI. Threat of Serious Injury

A. Standard of Review

97. At the outset of its findings, the Panel considered the standard of review appropriate for examination of the claims made by Australia and New Zealand. After citing our Report in Argentina - Footwear Safeguard , the Panel formulated the standard in the following terms:

� the standard of review that applies in safeguard disputes, as set out above, requires us to refrain from a de novo review of the evidence reflected in the report published by the competent national authorities. Our task is limited to a review of the determination made by the USITC and to examining whether the published report provides an adequate explanation of how the facts as a whole support the USITC's threat determination.⁵⁶

98. When the Panel came to examine the specific claims of Australia and New Zealand under Article 4.2, the Panel stated:

In examining the USITC's threat of serious injury determination we examine, first, whether the USITC evaluated "all relevant factors of an objective and quantifiable nature having a bearing on the situation of [the] industry", in particular, the factors listed in SG Article 4.2(a), as well as any other relevant factors. Second, we examine whether the approach followed by the USITC consisted of a fact-based, future-oriented consideration of increased imports and of the condition of the US domestic industry.⁵⁷ (emphasis in original)

99. Australia and New Zealand challenge two aspects of the Panel's standard of review. First, they argue that the Panel erred in its interpretation, and, therefore, formulation, of the legal standard to be used to review the determinations made by competent authorities in safeguard investigations. Second, they assert that, in reviewing the USITC's determination of a threat of serious injury, the Panel erred in its application of the standard of review.

100. As the Panel noted, we had occasion to examine, in Argentina - Footwear Safeguard, the standard of review appropriate to a panel's examination of claims made under the Agreement on Safeguards . In that appeal, we observed that:

[t]he Agreement on Safeguards � is silent as to the appropriate standard of review. Therefore, Article 11 of the DSU, and, in particular, its requirement that "� a panel should make an objective assessment of the matter before it, including an objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements", sets forth the appropriate standard of review for examining the consistency of a safeguard measure with the provisions of the Agreement on Safeguards.⁵⁸

101. As regards the standard of review contained in Article 11 of the DSU, we recall that, in European Communities - Hormones , we stated that "the applicable standard is neither de novo review as such, nor 'total deference', but rather the 'objective assessment of the facts' ".⁵⁹

102. In our Report in Argentina - Footwear Safeguard , we gave certain indications as to the application of the standard of review in Article 11 of the DSU in disputes where claims are made under Article 4 of the Agreement on Safeguards:

� with respect to its application of the standard of review, we do not believe that the Panel conducted a de novo review of the evidence, or that it substituted its analysis and judgement for that of the Argentine authorities. Rather, the Panel examined whether, as required by Article 4 of the Agreement on Safeguards, the Argentine authorities had considered all the relevant facts and had adequately explained how the facts supported the determinations that were made. Indeed, far from departing from its responsibility, in our view, the Panel was simply fulfilling its responsibility under Article 11 of the DSU in taking the approach it did. To determine whether the safeguard investigation and the resulting safeguard measure applied by Argentina were consistent with Article 4 of the Agreement on Safeguards , the Panel was obliged, by the very terms of Article 4, to assess whether the Argentine authorities had examined all the relevant facts and had provided a reasoned explanation of how the facts supported their determination.⁶⁰ (underlining added)

103. Thus, an "objective assessment" of a claim under Article 4.2(a) of the Agreement on Safeguards has, in principle, two elements. First, a panel must review whether competent authorities have evaluated all relevant factors, and, second, a panel must review whether the authorities have provided a reasoned and adequate explanation of how the facts support their determination.⁶¹ Thus, the panel's objective assessment involves a formal aspect and a substantive aspect. The formal aspect is whether the competent authorities have evaluated "all relevant factors". The substantive aspect is whether the competent authorities have given a reasoned and adequate explanation for their determination.

104. This dual character of a panel's review is mandated by the nature of the specific obligations that Article 4.2 of the Agreement on Safeguards imposes on competent authorities. Under Article 4.2(a), competent authorities must, as a formal matter, evaluate "all relevant factors". However, that evaluation is not simply a matter of form, and the list of relevant factors to be evaluated is not a mere "check list". Under Article 4.2(a), competent authorities must conduct a substantive evaluation of "the 'bearing', or the 'influence ' or 'effect' "⁶² or "impact" that the relevant factors have on the "situation of [the] domestic industry". (emphasis added) By conducting such a substantive evaluation of the relevant factors, competent authorities are able to make a proper overall determination, inter alia, as to whether the domestic industry is seriously injured or is threatened with such injury as defined in the Agreement.

105. It follows that the precise nature of the examination to be conducted by a panel, in reviewing a claim under Article 4.2 of the Agreement on Safeguards , stems, in part, from the panel's obligation to make an "objective assessment of the matter" under Article 11 of the DSU and, in part, from the obligations imposed by Article 4.2, to the extent that those obligations are part of the claim. Thus, as with any claim under the provisions of a covered agreement, panels are required to examine, in accordance with Article 11 of the DSU⁶³, whether the Member has complied with the obligations imposed by the particular provisions identified in the claim. By examining whether the explanation given by the competent authorities in their published report is reasoned and adequate, panels can determine whether those authorities have acted consistently with the obligations imposed by Article 4.2 of the Agreement on Safeguards .

106. We wish to emphasize that, although panels are not entitled to conduct a de novo review of the evidence, nor to substitute their own conclusions for those of the competent authorities, this does not mean that panels must simply accept the conclusions of the competent authorities. To the contrary, in our view, in examining a claim under Article 4.2(a), a panel can assess whether the competent authorities' explanation for its determination is reasoned and adequate only if the panel critically examines that explanation, in depth, and in the light of the facts before the panel. Panels must, therefore, review whether the competent authorities' explanation fully addresses the nature, and, especially, the complexities, of the data, and responds to other plausible interpretations of that data. A panel must find, in particular, that an explanation is not reasoned, or is not adequate, if some alternative explanation of the facts is plausible, and if the competent authorities' explanation does not seem adequate in the light of that alternative explanation. Thus, in making an "objective assessment" of a claim under Article 4.2(a), panels must be open to the possibility that the explanation given by the competent authorities is not reasoned or adequate.

107. In this respect, the phrase "de novo review" should not be used loosely. If a panel concludes that the competent authorities, in a particular case, have not provided a reasoned or adequate explanation for their determination, that panel has not, thereby, engaged in a de novo review. Nor has that panel substituted its own conclusions for those of the competent authorities. Rather, the panel has, consistent with its obligations under the DSU, simply reached a conclusion that the determination made by the competent authorities is inconsistent with the specific requirements of Article 4.2 of the Agreement on Safeguards .

108. In this case, as we have noted, the Panel formulated the standard of review by reference to our Report in Argentina - Footwear Safeguard , and the Panel also, explicitly, rejected any standard implying a de novo review of the evidence. Indeed, the Panel quoted the passage in our Report in Argentina - Footwear Safeguard to which we have just referred, and specifically drew attention to our statement, in that passage, that panels must examine whether competent authorities have examined all relevant factors and whether those authorities have provided a reasoned and adequate explanation for their determination.⁶⁴ Accordingly, we find that the Panel correctly interpreted the standard of review appropriate to the examination of the claims by Australia and New Zealand.

109. It will be recalled, though, that Australia and New Zealand have also appealed the Panel's application of the standard of review. For the most part, their appeal on the application of the standard of review is related to these participants' respective appeals that the Panel erred in finding that the USITC had acted consistently with Article 4.2 of the Agreement on Safeguards in determining that there existed a threat of serious injury to the United States' domestic lamb meat industry. We will, therefore, examine most of these arguments when we consider the issues relating to the existence of a threat of serious injury.

110. However, one aspect of New Zealand's appeal on the application of the standard of review raises a general procedural question we will address now. This pertains to the arguments that a panel is entitled to consider in reviewing competent authorities' determinations. The Panel said in this regard:

� to the extent that any of the alternative explanations put forward by Australia and New Zealand are in effect new analyses of the record evidence, they are not relevant to our review. Rather, these factual and legal arguments would be relevant to our review only to the extent that they were raised in the investigation, in which case we would need to consider whether the USITC gave a reasoned explanation of why the facts supported its conclusions in respect of them, and whether that explanation is persuasive.⁶⁵ (emphasis added)

111. Thus, the Panel confined its own review of the competent authorities' determination to an examination of that determination in terms of the factual and legal arguments put forward by the interested parties during the domestic investigation conducted under Article 3.1 of the Agreement on Safeguards .

112. In our report in Thailand - Anti-Dumping Duties on Angles, Shapes and Sections of Iron or Non-Alloy Steel H-Beams from Poland, in the course of our examination of the specificity of Poland's request for the establishment of a panel under Article 6.2 of the DSU, we said:

The Panel's reasoning seems to assume that there is always continuity between claims raised in an underlying anti-dumping investigation and claims raised by a complaining party in a related dispute brought before the WTO. This is not necessarily the case. The parties involved in an underlying anti-dumping investigation are generally exporters, importers and other commercial entities, while those involved in WTO dispute settlement are the Members of the WTO. Therefore, it cannot be assumed that the range of issues raised in an anti-dumping investigation will be the same as the claims that a Member chooses to bring before the WTO in a dispute.⁶⁶ (emphasis added)

113. Although the claim under examination in that appeal was different, the same reasoning applies in respect of the relationship between domestic investigations culminating in the imposition of a safeguard measure, and dispute settlement proceedings under the DSU regarding that safeguard measure. In arguing claims in dispute settlement, a WTO Member is not confined merely to rehearsing arguments that were made to the competent authorities by the interested parties during the domestic investigation, even if the WTO Member was itself an interested party in that investigation. Likewise, panels are not obliged to determine, and confirm themselves the nature and character of the arguments made by the interested parties to the competent authorities. Arguments before national competent authorities may be influenced by, and focused on, the requirements of the national laws, regulations and procedures. On the other hand, dispute settlement proceedings brought under the DSU concerning safeguard measures imposed under the Agreement on Safeguards may involve arguments that were not submitted to the competent authorities by the interested parties.

114. Furthermore, we recall that, in United States - Wheat Gluten Safeguard, we reversed a finding by the panel that competent authorities are obliged to evaluate only those other relevant factors, under Article 4.2(a), which were actually raised by the interested parties during the investigation before it.⁶⁷ We said there that competent authorities have an independent duty of investigation and that they cannot "remain[] passive in the face of possible short-comings in the evidence submitted, and views expressed, by the interested parties."⁶⁸ (emphasis added) In short, competent authorities are obliged, in some circumstances, to go beyond the arguments that were advanced by the interested parties during the investigation. As competent authorities themselves are obliged, in some circumstances, to go beyond the arguments of the interested parties in reaching their own determinations, so too, we believe, panels are not limited to the arguments submitted by the interested parties to the competent authorities in reviewing those determinations in WTO dispute settlement.

115. We wish to emphasize that the discretion that WTO Members enjoy to argue dispute settlement claims in the manner they deem appropriate does not, of course, detract from their obligation, under Article 3.10 of the DSU, "to engage in dispute settlement procedures 'in good faith in an effort to resolve the dispute'."⁶⁹ It follows that WTO Members cannot improperly withhold arguments from competent authorities with a view to raising those arguments later before a panel. In any event, as a practical matter, we think it unlikely that a Member would do so.

116. At the oral hearing before us, New Zealand indicated that, in its view, the Panel had failed to consider the econometric arguments it had set forth in Exhibit NZ-13 on the ground that these arguments had not been presented to the USITC. In view of our findings below, we do not find it necessary to examine the significance of Exhibit NZ-13.

B. The Determination of a "Threat of Serious Injury"

1. Background

117. Before the Panel, Australia and New Zealand both claimed that the USITC's determination of a threat of serious injury was inconsistent with Article 4.2(a) of the Agreement on Safeguards because the USITC did not properly evaluate "all relevant factors", as required by Article 4.2(a). As part of this claim, the parties asserted that the USITC did not have sufficient data to allow it to make a proper evaluation of the situation of the domestic industry.

118. The Panel found, first, that the USITC had "investigated" all the relevant factors mentioned in Article 4.2(a) of the Agreement on Safeguards.⁷⁰ The Panel next considered the approach the USITC took in determining whether there existed a "threat" of serious injury. The Panel concluded that there was "no conceptual fault with the USITC's analytical approach" and that this approach was "sufficiently fact-based and future-oriented".⁷¹ However, the Panel was "not persuaded" that the data used as a basis for the USITC's determination in this case was sufficiently representative of the domestic industry⁷², and the Panel, therefore, concluded:

� that the United States has acted inconsistently with Article 4.1(c) of the Agreement on Safeguards because the USITC failed to obtain data in respect of producers representing a major proportion of the total domestic production by the domestic industry as defined in the investigation.⁷³

Consequently, the Panel also found that the USITC's determination of a threat of serious injury was inconsistent with Article 2.1 of the Agreement on Safeguards . Australia and New Zealand appeal certain aspects of the Panel's findings on the threat of serious injury. So, too, does the United States, with respect to another aspect of these findings.

119. Although not identical in all respects, the appeals by Australia and New Zealand on this issue both challenge the Panel's interpretation and application of the term "threat of serious injury". Their arguments on the misapplication of this term are closely entwined with their arguments that the Panel erred, under Article 4.2(a), first, in concluding that the USITC had evaluated "all relevant factors" and, second, in concluding that the USITC had adopted a proper analytical approach to the evaluation of the data in a case of alleged "threat" of serious injury. Both participants assert that in reaching these conclusions, the Panel showed undue deference to the USITC. Accordingly, they argue that the Panel failed to apply the appropriate standard to its review of the USITC's determination.

120. For its part, the United States appeals the Panel's finding that, because the data before the USITC was not sufficiently representative of the domestic industry, the United States acted inconsistently with Article 4.1(c) of the Agreement on Safeguards . The United States argues that Article 4.1(c) simply defines the term "domestic industry" and does not impose any obligation on Members regarding the sufficiency of data about a "domestic industry". In any event, the United States adds, the USITC complied with the relevant obligations on data collection. As the United States sees it, Article 4.2(a) requires no more than that competent authorities evaluate all relevant factors of an "objective and quantifiable nature" having a "bearing" on the situation of the domestic industry, while Article 4.2(b) requires that those authorities' determination of the causal link be made on the basis of "objective evidence". The United States contends that the Agreement on Safeguards does not require that the data collected by competent authorities be representative of a particular proportion of the domestic industry.

121. Before addressing these appeals, we note that the Panel's analysis of these issues was based on the assumption that the USITC's findings on the definition of the "domestic industry" were consistent with the Agreement on Safeguards.⁷⁴ We have found that the Panel correctly concluded that the USITC improperly defined the "domestic industry". Even so, as the relevant findings of the Panel on the "threat of serious injury" have in fact been appealed, we think it appropriate for us to examine the "issues of law" and "legal interpretations" raised in this appeal regarding these findings. In doing so, we will use the same assumption employed by the Panel.

2. Meaning of the Term "Threat of Serious Injury"

122. We first address, briefly, the interpretation of the term "threat of serious injury", which is defined in Article 4.1(b) of the Agreement on Safeguards as follows:

(b) "threat of serious injury" shall be understood to mean serious injury that is clearly imminent, in accordance with the provisions of paragraph 2. A determination of the existence of a threat of serious injury shall be based on facts and not merely on allegation, conjecture or remote possibility; (emphasis added)

123. An integral element of this definition is the reference to "serious injury", which is defined in Article 4.1(a) of the Agreement on Safeguards as follows:

(a) "serious injury" shall be understood to mean a significant overall impairment in the position of a domestic industry; (emphasis added)

124. The standard of "serious injury" set forth in Article 4.1(a) is, on its face, very high. Indeed, in United States - Wheat Gluten Safeguard, we referred to this standard as "exacting".⁷⁵ Further, in this respect, we note that the word "injury" is qualified by the adjective "serious", which, in our view, underscores the extent and degree of "significant overall impairment" that the domestic industry must be suffering, or must be about to suffer, for the standard to be met. We are fortified in our view that the standard of "serious injury" in the Agreement on Safeguards is a very high one when we contrast this standard with the standard of "material injury" envisaged under the Anti-Dumping Agreement, the Agreement on Subsidies and Countervailing Measures (the "SCM Agreement") and the GATT 1994.⁷⁶ We believe that the word "serious" connotes a much higher standard of injury than the word "material".⁷⁷ Moreover, we submit that it accords with the object and purpose of the Agreement on Safeguards that the injury standard for the application of a safeguard measure should be higher than the injury standard for anti-dumping or countervailing measures, since, as we have observed previously:

[t]he application of a safeguard measure does not depend upon "unfair" trade actions, as is the case with anti-dumping or countervailing measures. Thus, the import restrictions that are imposed on products of exporting Members when a safeguard action is taken must be seen, as we have said, as extraordinary. And, when construing the prerequisites for taking such actions, their extraordinary nature must be taken into account.⁷⁸

125. Returning now to the term "threat of serious injury", we note that this term is concerned with "serious injury" which has not yet occurred, but remains a future event whose actual materialization cannot, in fact, be assured with certainty. We note, too, that Article 4.1(b) builds on the definition of "serious injury" by providing that, in order to constitute a "threat", the serious injury must be "clearly imminent ". The word "imminent" relates to the moment in time when the "threat" is likely to materialize. The use of this word implies that the anticipated "serious injury" must be on the very verge of occurring. Moreover, we see the word "clearly", which qualifies the word "imminent", as an indication that there must be a high degree of likelihood that the anticipated serious injury will materialize in the very near future. We also note that Article 4.1(b) provides that any determination of a threat of serious injury "shall be based on facts and not merely on allegation, conjecture or remote possibility." (emphasis added) To us, the word "clearly" relates also to the factual demonstration of the existence of the "threat". Thus, the phrase "clearly imminent" indicates that, as a matter of fact, it must be manifest that the domestic industry is on the brink of suffering serious injury.

126. We recall that, in Argentina - Footwear Safeguard , we stated that "it is essential for a panel to take the definition of 'serious injury' in Article 4.1(a) of the Agreement on Safeguards into account in its review of any determination of 'serious injury'."⁷⁹ The same is equally true for the definition of "threat of serious injury" in Article 4.1(b) of that Agreement. Thus, in making a determination on either the existence of "serious injury", or on a "threat" thereof, panels must always be mindful of the very high standard of injury implied by these terms.

3. Evaluation of Relevant Factors under Article 4.2(a) of the Agreement on Safeguards

127. So far, we have examined the interpretation of the term "threat of serious injury" in the abstract. However, the Agreement on Safeguards also imposes obligations on competent authorities with respect to the process by which they arrive at a determination of serious injury or threat thereof. Article 4.2(a) of the Agreement provides that, in making an injury determination, the competent authorities must "evaluate all relevant factors". This appeal raises two general interpretive questions concerning the way in which competent authorities actually conduct their "evaluation" of "all relevant factors". The first of these questions is whether the "evaluation" by the competent authorities, under Article 4.2(a), must be based on data that is sufficiently representative of the domestic industry. The second question is whether there is an appropriate temporal focus for the competent authorities' "evaluation" of the data in determining that there is a "threat" of serious injury in the imminent future.

(a) Sufficiency of the Data

128. The Panel found that, in order to make a determination regarding the "domestic industry", as defined in Article 4.1(c) of the Agreement on Safeguards, competent authorities must rely on data that is sufficiently representative of that industry.⁸⁰ The United States appeals this interpretation and asserts that the sole requirements in the Agreement on Safeguards on data collection are that the factors to be evaluated must be "of an objective and quantifiable nature" and that these factors have a "bearing on the situation of [the domestic] industry".

129. We note that no provision of the Agreement on Safeguards specifically addresses the question of the extent of data collection, and in particular, whether competent authorities must have before them data that is representative of the domestic industry. However, we note as well that, under Article 4.2(a) of the Agreement on Safeguards , competent authorities are required to investigate whether the "domestic industry" is facing a situation of "serious injury". To do so, competent authorities are obliged to "evaluate" all relevant factors of an "objective and quantifiable" nature. Moreover, in conducting this evaluation, Article 4.2(a) requires competent authorities to evaluate the "bearing" that the relevant factors have on the "situation of [the domestic] industry". On this basis, competent authorities must make an "overall" determination as to whether the "domestic industry" is seriously injured, or threatened with such injury.⁸¹

130. We recognize that the clause "of an objective and quantifiable nature" refers expressly to "factors", but not expressly to data. We are, however, convinced that factors can only be "of an objective and quantifiable nature" if they allow a determination to be made, as required by Article 4.2(b) of the Agreement on Safeguards , on the basis of "objective evidence". Such evidence is, in principle, objective data. The words "factors of an objective and quantifiable nature" imply, therefore, an evaluation of objective data which enables the measurement and quantification of these factors.

131. The term "domestic industry" is defined as meaning, at least, the producers of "a major proportion of the total domestic production" of the products at issue. In our view, the requirement for competent authorities to evaluate the "bearing" that the relevant factors have on the "domestic industry " and, subsequently, to make a determination concerning the overall "situation of that industry ", means that competent authorities must have a sufficient factual basis to allow them to draw reasoned and adequate conclusions concerning the situation of the "domestic industry". The need for such a sufficient factual basis, in turn, implies that the data examined, concerning the relevant factors, must be representative of the "domestic industry". Indeed, a determination made on the basis of insufficient data would not be a determination about the state of the "domestic industry", as defined in the Agreement, but would, in reality, be a determination pertaining to producers of something less than "a major proportion of the total domestic production" of the products at issue. Accordingly, we agree with the Panel that the data evaluated by the competent authorities must be sufficiently representative of the "domestic industry" to allow determinations to be made about that industry.

132. We do not wish to suggest that competent authorities must, in every case, actually have before them data pertaining to all those domestic producers whose production, taken together, constitutes a major proportion of the domestic industry. In some instances, no doubt, such a requirement would be both impractical and unrealistic. Rather, the data before the competent authorities must be sufficiently representative to give a true picture of the "domestic industry". What is sufficient in any given case will depend on the particularities of the "domestic industry" at issue. In this case, the Panel's conclusion that the data before the USITC was not sufficiently representative is, in our view, a finding that turns on the particularities of the United States' lamb meat industry, as defined by the USITC, and we see no reason to disturb this finding of the Panel. We note, moreover, that the USITC itself acknowledged that the data before it for growers did not represent a "statistically valid sample".⁸²

133. We, therefore, uphold the Panel's finding that the USITC, and, hence, the United States, acted inconsistently with the Agreement on Safeguards by making a determination regarding the "domestic industry" on the basis of data that was not sufficiently representative of that industry. However, we find that, in so doing, the United States acted inconsistently with Article 4.2(a) of the Agreement on Safeguards , read together with the definition of "domestic industry" in Article 4.1(c). Article 4.1(c) contains nothing more than a definition of the term "domestic industry" and does not, by itself, impose any obligation on WTO Members. We, therefore, disagree with the Panel's ultimate conclusion on this point that the United States acted inconsistently with Article 4.1(c) alone.

134. Accordingly, we modify the Panel's conclusion, in paragraph 8.1(e) of the Panel Report, by holding that the United States acted inconsistently with Article 4.2(a) of the Agreement on Safeguards in making a determination regarding the "domestic industry" on the basis of data that was not sufficiently representative of that industry.

(b) Temporal focus of the data evaluation

135. Before the Panel, the parties disagreed as to which part of the period of investigation was the most relevant in "evaluating" the state of the domestic industry when making a "threat" determination. The Panel opined that, "due to the future-oriented nature of a threat analysis, it would seem logical that occurrences at the beginning of an investigation period are less relevant than those at the end of that period."⁸³ The Panel went on to conclude that:

� the USITC was correct to focus on the most recent data available from the end of the investigation period. We also consider that data from 1997 and interim-1998 cover an adequate and reasonable time-period if complemented by projections extrapolating existing trends into the imminent future so as to ensure the prospective analysis which a threat determination requires.⁸⁴ (emphasis added)

Therefore, we consider that, by basing its determination on events at the end of the investigation period (i.e., one year and nine months) rather than over the course of the entire investigation period, the USITC analysed sufficiently recent data for making a valid evaluation of whether significant overall impairment was "imminent" in the near future.⁸⁵ (emphasis added)

136. We recall that, in making a "threat" determination, the competent authorities must find that serious injury is "clearly imminent". As we have already concluded, this requires a high degree of likelihood that the anticipated serious injury will materialize in the very near future. Accordingly, we agree with the Panel that a threat determination is "future-oriented". However, Article 4.1(b) requires that a "threat" determination be based on "facts" and not on "conjecture". As facts, by their very nature, pertain to the present and the past, the occurrence of future events can never be definitively proven by facts. There is, therefore, a tension between a future-oriented "threat" analysis, which, ultimately, calls for a degree of "conjecture" about the likelihood of a future event, and the need for a fact-based determination. Unavoidably, this tension must be resolved through the use of facts from the present and the past to justify the conclusion about the future, namely that serious injury is "clearly imminent". Thus, a fact-based evaluation, under Article 4.2(a) of the Agreement on Safeguards , must provide the basis for a projection that there is a high degree of likelihood of serious injury to the domestic industry in the very near future.⁸⁶

137. Like the Panel, we note that the Agreement on Safeguards provides no particular methodology to be followed in making determinations of serious injury or threat thereof.⁸⁷ However, whatever methodology is chosen, we believe that data relating to the most recent past will provide competent authorities with an essential, and, usually, the most reliable, basis for a determination of a threat of serious injury. The likely state of the domestic industry in the very near future can best be gauged from data from the most recent past. Thus, we agree with the Panel that, in principle, within the period of investigation as a whole, evidence from the most recent past will provide the strongest indication of the likely future state of the domestic industry.

138. However, we believe that, although data from the most recent past has special importance, competent authorities should not consider such data in isolation from the data pertaining to the entire period of investigation. The real significance of the short-term trends in the most recent data, evident at the end of the period of investigation, may only emerge when those short-term trends are assessed in the light of the longer-term trends in the data for the whole period of investigation. If the most recent data is evaluated in isolation, the resulting picture of the domestic industry may be quite misleading. For instance, although the most recent data may indicate a decline in the domestic industry, that decline may well be a part of the normal cycle of the domestic industry rather than a precursor to clearly imminent serious injury. Likewise, a recent decline in economic performance could simply indicate that the domestic industry is returning to its normal situation after an unusually favourable period, rather than that the industry is on the verge of a precipitous decline into serious injury. Thus, we believe that, in conducting their evaluation under Article 4.2(a), competent authorities cannot rely exclusively on data from the most recent past, but must assess that data in the context of the data for the entire investigative period.⁸⁸

139. In this case, the Panel interpreted Article 4.2(a) of the Agreement on Safeguards to mean that the USITC was entitled to "base its determination" of a "threat of serious injury" on data pertaining to the last 21 months of the five year period of investigation. In our view, as we will see below, the Panel's interpretation of the temporal aspects of the competent authorities' evaluation, under Article 4.2(a), placed too much emphasis on certain data from the most recent past, while neglecting other, even more recent data. Also, the Panel did not ensure that the data was assessed in the context of the data for the entire period of investigation. The Panel's approach improperly excluded the possibility that short-term trends in the data, evident in the last 21 months of the period of investigation, could possibly be a misleading indicator of the likely future state of the domestic industry, when viewed in the context of the data for the entire period of investigation.

4. Panel's Review of the USITC's Determination of a "Threat of Serious Injury"

140. We have examined, so far, the interpretation of the appropriate standard of review, and the interpretation of the term "threat of serious injury" defined in Article 4.1(b) of the Agreement on Safeguards . We have also examined certain interpretive questions relating to the competent authorities' evaluation, under Article 4.2(a) of that Agreement, in making a determination that there is such a "threat". With all these considerations in mind, we will now examine the heart of the appeal by Australia and New Zealand on this point: whether the Panel applied the appropriate standard of review to the USITC's evaluation of the state of the domestic industry, under Article 4.2(a), and to the USITC's determination that there existed a "threat of serious injury".

141. We have already said that, in examining a claim under Article 4.2 of the Agreement on Safeguards , a panel's application of the appropriate standard of review of the competent authorities' determination has two aspects. First, a panel must review whether the competent authorities have, as a formal matter, evaluated all relevant factors and, second, a panel must review whether those authorities have, as a substantive matter, provided a reasoned and adequate explanation of how the facts support their determinations.

142. As regards the formal aspects of the Panel's review, we observe that the Panel found that "the USITC has investigated all the relevant injury factors".⁸⁹ (emphasis added) The Panel reached this conclusion after describing, in summary form, the data before the USITC concerning the relevant factors.⁹⁰ Although Australia argues that the Panel erred in finding that the USITC had evaluated all of the relevant factors⁹¹, we consider that the Panel was correct to conclude that, as a formal matter, the USITC did evaluate each of the relevant factors, with respect to some part of the domestic industry, as the USITC defined that industry. However, we see the essence of Australia's appeal, on this point, as directed more toward the Panel's review of the substantive aspects of the USITC's evaluation of the relevant factors, and to the USITC's determination that the domestic industry was threatened with serious injury.⁹² New Zealand also appeals the substantive aspects of the Panel's review.⁹³ Both of these appellants assert that the Panel erred in finding that the USITC's explanation of its evaluation of the relevant factors was adequate to support its determination that there existed a threat of serious injury to the domestic industry.

143. On our reading of the Panel Report, we see that the Panel examined, first, whether the USITC formally evaluated all the relevant factors. Next, the Panel examined "the USITC's analytical approach" and concluded that there was "no conceptual fault" with that approach because it was "sufficiently fact-based and future-oriented".⁹⁴ The Panel reached this conclusion after reviewing the "[p]rojections relevant to a threat of injury finding" and the "[r]elevant time-period for the threat analysis".⁹⁵ The Panel also made an "[e]valuation of data pertaining to the period from January 1997 to September 1998".⁹⁶ This period was the last 21 months of the period of investigation, on which the Panel said the USITC could "bas[e] its determination".⁹⁷ In the course of evaluating the data from this period, the Panel noted:

� the complainants do not, as such, challenge the USITC's findings that there were declines in 1997 and interim-1998 for most of the indicators referred to by the USITC in its determination.⁹⁸

144. After summarizing the remainder of the parties' views on the data, the Panel then observed, correctly, that the competent authorities are not required "to show that each listed injury factor is declining" but, rather, they must reach a determination in light of the evidence as a whole.⁹⁹ Without any further analysis, the Panel concluded:

Therefore, in the light of the specific evidence, explanations and prospective analysis reflected in the USITC report, we consider the USITC's reliance, among other difficulties, on factors including the domestic industry's market share, production, shipments, profitability and prices as a sufficient basis for determining whether threat of serious injury exists. We also consider that the USITC's analysis of the overall picture of trends reflected in and projected from the most recent data (especially from 1997 and interim-1998) along with the projections concerning further increases in imports (assuming arguendo that the data on which these trends and projections were based were representative of a major proportion of the producers forming the relevant industry), seem to confirm the USITC determination that a "significant overall impairment" in the overall position of the domestic industry was clearly imminent.¹⁰⁰ (underlining added)

145. Finally, the Panel examined, and upheld, the claim by Australia and New Zealand that the USITC did not have sufficiently representative data to make a determination about the lamb meat industry, as defined by the USITC.¹⁰¹

146. The only part of the Panel Report where the Panel purports to conduct a substantive review of whether the USITC provided a reasoned and adequate explanation of how the facts supported its determination is in the section dealing with the "[e]valuation of data pertaining to the period from January 1997 to September 1998".¹⁰² However, even there, the Panel did not demonstrate any substantive review of the factors which it considered provided "a sufficient basis" for the USITC's determination - "market share, production, shipments, profitability and prices". The Panel seemed to regard it as sufficient for its own conclusion, first, that most economic indicators were in decline in 1997 and 1998 and, second, that not every economic indicator need be in decline.

147. Australia and New Zealand made a number of substantive arguments before the Panel about the USITC's evaluation, and about its determination that the domestic industry was "threatened with serious injury" at the end of the period of investigation in 1998. These arguments were that: lamb meat prices actually rose in late 1998 and in 1999; the USITC's price comparisons were inappropriate; the USITC did not properly evaluate capacity, capacity utilization, inventories and productivity; lamb growers' production, sales and productivity increased over the entire period of investigation; shipments of live lambs increased in 1998; and, growers' productivity and employment levels increased in 1998. The Panel summarized the substantive arguments made by Australia and New Zealand regarding the USITC's evaluation and determination.¹⁰³ Yet, nowhere do we see that the Panel actually addressed them.

148. These arguments by Australia and New Zealand were evidently intended to cast doubt on the adequacy of the explanation provided by the USITC as to how the facts, in the form of the data, supported the USITC's "threat of serious injury" determination. In our view, by failing to review the USITC's determination in light of these detailed substantive arguments, the Panel failed to examine critically whether the USITC had, indeed, provided a reasoned and adequate explanation of how the facts supported its determination that there existed a "threat of serious injury".

149. In consequence, we find that the Panel has not applied the appropriate standard of review, under Article 11 of the DSU, in examining whether, as a substantive matter, the USITC provided a reasoned and adequate explanation of how the facts support a determination of "threat of serious injury" under Article 4.2(a) of the Agreement on Safeguards .

150. Having found that the Panel did not properly review the USITC's determination that there existed a "threat of serious injury", we now turn to examine ourselves the claims of Australia and New Zealand on this issue. We will focus on the arguments by Australia and New Zealand relating to prices, and we will base our determination exclusively on the facts presented in the USITC Report, which form part of the Panel record and are uncontested.

151. The USITC made the following statement about prices:

We find that financial performance across all industry segments has worsened due largely to falling prices. Commission questionnaires show a marked decline in prices for various lamb meat products beginning in mid-1997. Although prices recovered to some degree in 1998, prices remained depressed through September 1998, the end of the period surveyed. Weighted average U.S. delivered prices for virtually all of the products surveyed were substantially lower beginning with the third quarter of 1997. In several instances prices for several of the products were 20 percent or more below comparable quarters in 1996 and early 1997.¹⁰⁴ (emphasis added)

In summarizing the data on prices, the USITC Report states:

Respondents have questioned whether the domestic industry is injured when slaughter prices, the price felt most directly by the segment of the industry that petitioners have stated are most injured, ended the period of investigation higher than they began. Respondents argued that prices are returning to normal after a temporary increase brought about by decreased domestic supply. Petitioners state that the Commission should focus on the price decline in 1996 and 1997 and that 1993 is not a good year for comparison because it was one of the worst years on record. To examine this question, staff converted monthly slaughter lamb prices (Jan. 1985-Aug. 1998) to constant 1985 dollars using the BLS producer price index to remove the effects of inflation. The average real price per CWT was $62.96 for the full-period (Jan. 1985-Aug. 1998) and $56.19, $55.61, $64.86, $71.50, $73.32, and $64.73, respectively, for years 1993-interim 1998. Thus, prices were below the full-period average in 1993 and 1994, increased above the average level in 1995-97, and decreased to slightly above the average in interim 1998.¹⁰⁵

152. Australia and New Zealand argued, before the Panel, that it was not appropriate for the USITC to use prices from 1996 and 1997 as the benchmark for comparison with prices in 1998, at the end of the period of investigation, because prices in 1996 and 1997 were unusually high. The fall in prices between 1996 and 1998 was, they argued, a misleading indicator of price trends because prices were simply returning to their normal levels and were not in general decline. In addition, Australia and New Zealand argued that, in any event, the most recent price data indicates that prices were rising in 1998, at the end of the period of investigation.¹⁰⁶ This rise in prices indicated, they argued, that the situation of the lamb meat industry was improving and that it was not faced imminently with "serious injury".

153. We note that the price data set out in the USITC Report, which we have just summarized, indicates that prices in 1996 and 1997 were considerably higher than at any other point during the period of investigation.¹⁰⁷ The data also shows that prices were significantly lower in August 1998 than in 1996 and 1997.¹⁰⁸ Prices in 1998 were, nonetheless, markedly higher than the prices in 1993 and 1994, the early part of the period of investigation.¹⁰⁹ In addition, the price data in the Part II of the USITC Report, and in United States' Exhibit US-41, submitted to the Panel, indicates that there was a sharp rise in prices in the last few months of the period of investigation, in mid-1998.¹¹⁰

154. Accordingly, the uncontested data demonstrates that, during the period of investigation, lamb meat prices generally rose until 1996/1997, then dropped until mid-1998, and rose again until the end of the period of investigation. At that time, prices were higher than they had been at the beginning of the period of investigation.

155. We emphasize that we are not in a position to reach any definitive conclusions on the significance of these price trends for the situation of the domestic lamb meat industry. However, these trends raise doubts for us about the adequacy of the USITC's explanation of the "bearing" of prices on the situation of the domestic industry.

156. In the passage we quoted previously, from page I-20 of the USITC Report, the USITC expressed the view that the "worsen[ing]" financial performance of the domestic industry was "due largely to falling prices" for lamb meat.¹¹¹ It is clear from this passage that the USITC determined that prices were "falling" through a comparison between prices in 1998 and prices in 1996/1997. However, it seems to us that there is a legitimate doubt as to which prices during the period of investigation should have been used as the appropriate benchmark. That doubt stems from the fact that prices in 1996 and 1997 were around 30 percent higher than they had been in 1993 and, during those two years, were also at their peak for the period of investigation. In these circumstances, we consider that the USITC Report should have explained why prices in 1996 and 1997 were the appropriate benchmark rather than prices in 1993, 1994 or 1995. The USITC provides no such explanation and, instead, assumes that prices in 1996 and 1997 were the appropriate benchmark. We do not wish to suggest that prices in 1996 and 1997 could not be used as the benchmark, or that prices from another year should have been the benchmark. Our point is that the USITC has not justified its decision - which was key to its overall evaluation of prices and, thus, also, of the financial performance of the domestic industry - that prices in 1996 and 1997 were the appropriate benchmark for comparison with prices in 1998.

157. For similar reasons, we are not satisfied that the USITC explained adequately its conclusion that "prices remained depressed through September 1998" because, compared with price levels in 1993 and 1994, prices in September 1998 were markedly higher, and were not "depressed".¹¹² (emphasis added) Again, the USITC's conclusion overlooks entirely the evolution of prices across the entire period of investigation, and fails to explain why the overall rise in prices between 1993 and 1998 is not relevant to the determination.

158. In addition, we have already observed that there was a sharp rise in lamb meat prices in the last few months of the period of investigation.¹¹³ However, the USITC's consideration of this rise in prices was confined to the observation that, "[a]lthough prices recovered to some degree in 1998, prices remained depressed through September 1998".¹¹⁴ (emphasis added) The USITC did not elaborate further on the importance to the domestic industry of the rise in prices in 1998. Nor did the USITC explain the likely future evolution of prices in light of these price rises which were, in some cases, rather significant. The USITC did not, therefore, explain, at all, whether it considered that prices would continue to rise; whether the price rises would be reversed; or whether prices would remain at the level reached at the end of the period of investigation.

159. We see a contradiction between the most recent price rises, in 1998, and the USITC's conclusion that the price data supports its determination that the domestic industry is threatened with serious injury. A rise in prices, particularly if significant, should, in the ordinary course of events, be beneficial for an industry. Conceivably, such a rise could lead to an increase in revenues, and could increase margins and profits, and, possibly, also, production levels, if the price rises are sustained. Thus, if an industry is not yet in a state of serious injury, and that industry has enjoyed rising prices in the most recent past, it is, at least, questionable whether the industry is highly likely to suffer serious injury in the very near future. In such a situation, the competent authorities should devote particular attention to explaining the apparent contradiction between the most recent price rises and their view that the industry is still threatened with serious injury. In this case, the USITC offered no such explanation.

160. We wish to emphasize again that our remarks about the price data are not intended to suggest that the domestic industry was not threatened with serious injury. Rather, our conclusion is simply that the USITC has not adequately explained how the facts relating to prices support its determination, under Article 4.2(a), that the domestic industry was threatened with such injury.

161. Accordingly, we find that the United States acted inconsistently with Article 4.2(a) of the Agreement on Safeguards and, hence, also with Article 2.1 of that Agreement.

To continue with VII. Causation

Return to Index

56 Panel Report, para. 7.3.

57 Ibid., para. 7.140.

58 Appellate Body Report, supra, footnote 15, para. 120.

59 Appellate Body Report, supra, footnote 25, para. 117.

60 Appellate Body Report, supra, footnote 15, para. 121.

61 Clearly, a claim under Article 4.2(a) might not relate at the same time to both aspects of the review envisaged here, but only to one of these aspects. For instance, the claim may be that, although the competent authorities evaluated all relevant factors, their explanation is either not reasoned or not adequate.

62 Appellate Body Report, United States - Wheat Gluten Safeguard, supra, footnote 19, para. 71.

63 We note, however, that Article 17.6 of the Anti-Dumping Agreement sets forth a special standard of review for claims under that Agreement.

64 Panel Report, para. 7.1.

65 Panel Report, para. 7.207.

66 Appellate Body Report, WT/DS122/AB/R, adopted 5 April 2001, para. 94.

67 Appellate Body Report, supra, footnote 19, para. 56.

68 Ibid., para. 55.

69 See, Appellate Body Report, United States - Tax Treatment for "Foreign Sales Corporations", WT/DS108/AB/R, adopted 20 March 2000, para. 166.

70 Panel Report, para. 7.177.

71 Ibid., paras. 7.222 and 7.224.

72 Ibid., para. 7.225.

73 Ibid., para. 8.1(e).

74 Panel Report, para. 7.119.

75 Appellate Body Report, supra, footnote 19, para. 149.

76 "Material injury" is the standard provided for in Article VI of the GATT 1994, Articles 5 (footnote 11) and 15 (footnote 45) of the SCM Agreement, and Article 3 (footnote 9) of the Anti-Dumping Agreement.

77 We find support for our view that the standard of "serious injury" is higher than "material injury" in the French and Spanish texts of the relevant agreements, where the equivalent terms are, respectively, dommage grave and dommage important; and da�o grave and da�o importante.

78 Appellate Body Report, Argentina - Footwear Safeguard , supra, footnote 15, para. 94.

79 Supra, footnote 15, para. 139.

80 Panel Report, para. 7.221.

81 Appellate Body Report, Argentina - Footwear Safeguard , supra, footnote 15, para. 139.

82 USITC Report, p. I-17.

83 Panel Report, para. 7.192.

84 Ibid., para. 7.193.

85 Ibid., para. 7.194.

86 We observe that the projections made must relate to the overall state of the domestic industry, and not simply to certain relevant factors.

87 Panel Report, para. 7.184.

88 We note that, at footnote 130 of our Report in Argentina - Footwear Safeguard, supra, footnote 15, we said that "the relevant investigation period should not only end in the very recent past, the investigation period should be the recent past." In this Report, we comment on the relative importance, within the period of investigation, of the data from the end of the period, as compared with the data from the beginning of the period. The period of investigation must, of course, be sufficiently long to allow appropriate conclusions to be drawn regarding the state of the domestic industry.

89 Panel Report, para. 7.177.

90 Ibid., paras. 7.142 - 7.171. We note that this section of the Panel Report is entitled "Whether the USITC evaluated in this investigation all injury factors listed in SG Article 4.2(a)". After an introduction
(paras. 7.139 - 7.141), the Panel presented a "Summary of the injury data collected by the USITC"
(paras. 7.142 - 7.171), before turning to examine whether the USITC had evaluated all relevant factors
(paras. 7.172 - 7.178).

91 Australia's other appellant's submission, paras. 63 - 80.

92 Australia's other appellant's submission, paras. 63 - 80.

93 New Zealand's other appellant's submission, paras. 3.12 - 3.22.

94 Panel Report, paras. 7.222 and 7.224.

95 Ibid., headings 3(a) and 3(b), pp. 64 and 65.

96 Ibid., heading 3(c), p. 66.

97 Ibid., para. 7.194.

98 Ibid., para. 7.199.

99 Panel Report, para. 7.203. We note that, earlier in its Report, the Panel stated that competent authorities "may arrive at a threat determination even if the majority of firms within the relevant industry is not facing declining profitability, provided that an evaluation of the injury factors as a whole indicates threat of serious injury." (Panel Report, para. 7.188, emphasis added) In Argentina - Footwear Safeguard , we said that the competent authorities' determination of "serious injury" must be based on "the overall picture" of the domestic industry and that the determination must be made "in light of all the relevant factors". Accordingly, in evaluating "the overall position of the domestic industry", no single relevant factor can be accorded decisive importance and, instead, all of the factors must be examined and weighed together. (Appellate Body Report, Argentina - Footwear Safeguard , supra, footnote 15, para. 139)

It follows that the Panel was correct to state that the competent authorities' determination must be based on "an evaluation of the injury factors as a whole". Moreover, it is theoretically possible, as the Panel said, that an industry might be threatened with serious injury, even though "a majority of firms � is not facing declining profitability". Profits are simply one of the relevant factors mentioned in Article 4.2(a) and to accord that factor decisive importance would be to disregard the other relevant factors. However, in our view, it will be a rare case, indeed, where the relevant factors as a whole indicate that there is a threat of serious injury, even though the "majority of firms in the industry" is not facing declining profitability.