The TRA has recently set out its first findings on the use of ‘particular market situation’ (PMS) in the calculation of normal value in the aluminium extrusions case. I have written a separate blog item on this (Treatment of China in UK trade remedy investigations). In the context of doing this I thought it would be useful to set out the guidance on this issue provided by the panel in DS529 (Australia - AD measures on A4 Copy Paper).
The panel in DS529 is the first to consider the particular market situation provisions in detail. Of course this is only one panel decision. It was not appealed and it is likely that there will have to be more panel and Appellate Body reviews of the issue for robust WTO jurisprudence to emerge. I think that there remain significant ambiguities in the panel's analysis which will need further consideration for there to be greater clarity around this provisions.
The following summarises some of the key points as I see it from this panel report (paragraph numbers from the panel report are given in parentheses):
What is a particular market situation (PMS)?
Definition of particular market situation (PMS)
PMS is one of the circumstances in which an investigation authority can reject the use of domestic prices and costs in the calculation of normal value, against which export prices will be compared in order to calculate a dumping margin.
The term particular market situation has never been defined by any previous panel or Appellate Body. The panel's analysis is useful in making a first interpretation of PMS.
The panel points out that PMS is only relevant if it renders domestic sales unfit to permit a proper comparison and agrees with an old GATT panel.
…we agree with the observation of the GATT panel in EEC – Cotton Yarn that a "particular market situation" is only relevant insofar as it has the effect of rendering domestic sales unfit to permit a proper comparison. (7.21)
Thus, a PMS can be defined as one that renders domestic sales unfit to permit a proper comparison.
This remains a broad definition. The panel also found that the circumstances constituting a PMS cannot be comprehensively identified (7.21). In addition, it stated that a market situation must be "distinct, individual, single, specific" but that it does not have to "unusual" or "exceptional" (7.22).
The panel's analysis, therefore, suggests that PMS has broad potential in terms of the situations that might be identified by investigating authorities.
Can a government action constitute a PMS?
Indonesia argued that a situation arising from government action was necessarily disqualified from constituting a ‘particular market situation’. It argued that government behaviour is expressly regulated by the Subsidies & Countervailing Measures agreement and is not covered by the anti-dumping (AD) agreement. Thus, according to Indonesia, government actions, such as a subsidy, cannot constitute a PMS.
Having created a broad definition of circumstances that could constitute a PMS the panel rejects this argument. It points out that, where a subsidy reduces export price below normal value thus creating a situation of dumping, the AD agreement allows action regardless of the fact that the dumping is caused by a subsidy.
The panel reiterates its conclusion that the examination of a PMS requires a fact-specific and case-by-case analysis. However, the panel clearly concludes that a government action (including a subsidy) can constitute a PMS (7.56).
Proper comparison and subsidies affecting both domestic and export sales
Indonesia argued that a low-priced (subsidised) input used in both export and domestic sales still allows a proper comparison. The logic of Indonesia’s argument can be explained as follows:
If the subsidised input is only used for domestic sales, while export sales contained non-subsidised inputs, the Indonesian argument is the domestic sales could be rejected, and a constructed normal value (CNV) used, because the effect of the subsidised input on domestic and export sales is different.
If the subsidised input is only used for export sales, this would also create a situation where the export price is below the domestic price (normal value) i.e. dumping.
However, where the subsidised input is used in both domestic and export sales, Indonesia is arguing that a proper comparison can be made because the subsidy affects both domestic and export prices.
Indonesia seems to effectively be interpreting the term ‘proper comparison’ used in Article 2.2 as being equivalent to the term ‘fair comparison’ used in Article 2.4. A fair comparison ensures that, when the comparison between domestic and export price is made, the two prices are comparable i.e. like is being compared with like in terms of the various characteristics of the product.
The panel attempts to define 'proper comparison' but it doesn't really explore whether there is a difference between proper and fair comparison. I believe that this leaves considerable ambiguities that need to be resolved in future disputes.
The panel rightly focuses on whether the PMS allows the domestic price to be used as the benchmark against which the export price will be compared to establish the dumping margin.
The function of the "permit a proper comparison" test is to determine whether the domestic price can or cannot be used as a basis for comparison with the export price to identify the existence of dumping.(7.74)
The panel outlines the Indonesia argument that the key issue here is price comparability:
According to Indonesia…Article 2.2 requires examination of price comparability between domestic sales and export sales even in the context of a particular market situation (7.64)
It then spells out what the investigating authority must do:
...the investigating authority must examine the domestic sales in order to determine whether a proper comparison between the two prices is permitted in spit of the effect of the particular market situation.
The panel does not explore the meaning of the term 'price comparability' as used by Indonesia. The panel uses the term in a way that implies it means the same thing for both a proper comparison and fair comparison.
The panel then focuses on whether a subsidised input used in domestic and export sales allows a proper comparison. It finds that such a situation does not necessarily disqualify a finding of PMS.
…. we find that Indonesia has failed to demonstrate that a situation of a low-priced input used identically to produce merchandise for the domestic and export markets is necessarily disqualified from constituting a "particular market situation" within the meaning of Article 2.2 of the Anti-Dumping Agreement. (7.32)
Use of the term 'necessarily' suggests that there may be situations where the subsidised input used in identically in domestic and export products is disqualified from constituting a PMS. This is further implied by the panel by focusing on the question of whether there are differences in the impact of the subsidy on domestic and export sales:
….we consider that, in at least some cases, differences in the impact on domestic and export sales could prevent a proper comparison. (7.57)
We therefore consider that the "proper comparison" language calls for an assessment of the relative effect of the particular market situation on domestic and export prices. (7.75)
By focusing on whether there are differences in the impact on domestic and export sales, the panel is effectively treating the two types of comparison (proper and fair) as equivalent, without ever fully considering the arguments for this.
The panel is effectively creating an ‘effects’ test for interpretation of proper comparison. It is not fully agreeing with the logic of the Indonesian argument but has definitely guided the interpretation of proper comparison to be equivalent to fair comparison. The implication of this is that, if the effect of the subsidy on domestic and export price is identical, a proper comparison can be made. Although the panel does not explicitly state this it is clearly implied by the various statements it does make.
Why I think that the panel has not settled this issue
The panel did not explicitly address the most interesting question that remains unanswered in this ruling. This question is whether a proper comparison could be made if the effect of the subsidy is the same for both domestic and export sales? The implication of the panel’s focus on the effect on domestic and export sales is that a proper comparison could be made in that situation. But the panel did not make any ruling on this point.
Australia did not focus on the term proper comparison. Rather, Australia focused on arguing that the key issue is whether the domestic sales are suitable for establishing normal value.
Australia argued that the appropriate analysis of whether "because of the particular market situation ... such sales do not permit a proper comparison" requires determining whether the domestic sales are "suitable" for establishing a normal value that will provide a "reliable foundation" that will "permit" a "proper comparison" with the export price. (7.85)
The panel rejected this line of argument:
We consider that this approach fails to give meaning and effect to the phrase "permit a proper comparison". (7.87)
I think that there is significant merit to the Australian argument that the key issue is whether the sales are 'suitable' for establishing a normal value that will permit a 'proper comparison'.
At the same time, I think I can see why the panel feels like it has to reject the focus on normal value in terms of ensuring that the 'comparison' element is not missed. However, one of the key critical points that the panel neglects to undertake is an explicit consideration of the difference between proper and fair comparisons. Indonesia seems to conflate the issues of proper comparison (2.2) and fair comparison (2.4) and the panel seems to go along with it.
It is somewhat surprising that the panel did not consider the difference between proper and fair comparison despite the fact that several parties in the dispute had raised the point. The text below provides extracts from Russian, Thai and US submissions made in the context of DS529.
Extracts from WT/DS529/R/Add.1 – Addendum to the report of the panel in Australia – AD measures on A4 Copy Paper
INTEGRATED EXECUTIVE SUMMARY OF THE ARGUMENTS OF THE RUSSIAN FEDERATION (12) The use by the drafters of different wordings in different obligations ("proper comparison" in Article 2.2 and "fair comparison" in Article 2.4 of the Anti-Dumping Agreement) with different requirements should be respected. The structure of Article 2 of the Anti-Dumping Agreement and interconnection of its provisions support the understanding that the obligation to ensure fair comparison arises after the normal value and the export price are established. According to the WTO jurisprudence, "the subject-matter of Article 2.4, i.e. differences affecting the comparability of the normal value and the export price, can be contrasted with that of Articles 2.1, 2.2 – including its subparagraphs – and 2.3 which pertain to the methodology for determining the normal value and the export price".
INTEGRATED EXECUTIVE SUMMARY OF THE ARGUMENTS OF THAILAND IV. THE PROPER COMPARISON IS DISSIMILAR FROM THE FAIR COMPARISON 7. Thailand notes that the "fair comparison" under Article 2.4 of the Anti-Dumping Agreement seeks to ensure that a comparison between the normal value and the export price is fair by adjusting differences which may affect price comparability. The proper comparison under Article 2.2 of the Anti-dumping Agreement, on the other hand, relates to the step before the adjustment occurs when considering if the domestic sales price may be used as a basis for deriving normal value in order to ensure that the comparison to the export price is proper for the purpose of the determination of dumping.
EXECUTIVE SUMMARY OF U.S. THIRD PARTY RESPONSES TO QUESTIONS 22. The phrase "proper comparison" in Article 2.2 and the phrase "fair comparison" in Article 2.4 are aimed at different lines of inquiry. Article 2.2 establishes certain alternatives for establishing normal value, when there are no domestic sales in the ordinary course of trade or when, because of a "particular market situation" or a "low volume of ... sales in the domestic market of the exporting country", "such sales do not permit a proper comparison". In contrast, the text of Article 2.4 presupposes that an investigating authority has established a normal value pursuant to Articles 2.1 and 2.2 and an export price pursuant to Article 2.3 and obligates an investigating authority to make a "fair comparison" between normal value and export price when determining the existence of dumping and when calculating a margin of dumping.
The emphasis of the panel on the effect of the subsidy, and the implication that the same effect would allow a proper comparison, suggests that the panel implicitly made the same mistake as Indonesias in conflating the concepts of proper and fair comparison. It is a shame that it did not pick up these arguments and consider the difference between fair and proper comparisons.
Why proper and fair comparisons are different
I think it is useful to take a more detailed look at the difference between proper and fair comparison.
A fair comparison ensures that, when the comparison between domestic and export price is made, the two prices are comparable i.e. like is being compared with like in terms of the various characteristics of the product. Article 2.4 creates the obligation on investigating authorities to consider differences in product characteristics and terms of sale to ensure that the comparison is fair.
As the panel itself says, the term proper comparison is concerned with whether domestic sales can be used as the "basis of comparison" with the export price for determining normal value (7.74). However, the way in which the comparison might be affected is not explored.
A key point here is that the Article 2.4 fair comparison is not concerned with trade distorting factors. It is concerned with practical comparison issues in terms of physical characteristics and terms of sale.
The article 2.2 proper comparison is concerned with assessing the appropriateness of domestic sales for determining normal value.
Normal value is the benchmark against which the export price is assessed. It is the test used to assess if the export price is unfair. Thus, in order to consider the appropriateness of domestic sales as the basis of normal value, it is necessary to assess whether the value is an appropriate representation of a normal or fair price.
The existence of a subsidy is not a normal characteristic that is merely to be considered in terms of a fair comparison. Rather, it is a trade distorting factor that itself can create an unfair export price regardless of whether a fair comparison is possible or not.
It is absurd that a distortion cannot be addressed if it affects domestic and export price equally
The idea that a subsidy affecting domestic and export sales equally cannot be addressed is absurd. The reason why the domestic price can be rejected when there is a particular market situation is because it may lead to an underestimate of the margin of dumping. Whether a price is 'fair' or not is determined by comparing it to the normal value. An export price that benefits from a subsidy is not fair whether it affects the domestic price equally or not.
The logic of this absurdity can be summarised as follows:
Take an example of a company that decides to adopt a very aggressive pricing strategy and sells both domestic and export sales at a price less than cost of production. If domestic and export price are the same, on a simple comparison there is no dumping.
In this situation, the anti-dumping agreement permits domestic price to be rejected i.e. it can be considered not in the ordinary course of trade. Although the WTO Agreement doesn't say this it would not be unreasonable to say that sales not in the ordinary course of trade are not suitable to make a proper comparison. This is not an issue of fair comparison. Normal value could be constructed on the basis of cost of production and dumping would legitimately be found.
Whether the low domestic price is caused by commercial strategy or government subsidy is irrelevant to the dumping calculation, as confirmed by the panel in this ruling (7.44).
The absurdity arises from Indonesia’s argument that there is a distinction between dumping caused by commercial strategy or government intervention.
The ordinary meanings of the terms proper and fair are not identical. If the drafters of the anti-dumping agreement had intended for these concepts to be the same, they would have used the same language for both. Instead, the term proper comparison cannot be made redundant in its own right by giving it the same meaning as fair comparison. There must be a difference between a proper comparison and a fair comparison. In the case of a subsidy that affects both domestic and export sales equally, the distortion ensures that there cannot be a proper comparison even though there is no comparison issue in relation to fairness.
To be fair on the panel, it did not need to rule specifically on whether the impact of a PMS is identical for domestic and export price definitively means that a proper comparison can be made. I think this is implicit in many of the things that the panel says. However, future panels and Appellate Body reports will have to address this issue for there to be great clarity.
Conditions for using PMS to disregard records kept by the exporter
The panel considered the meaning of Article 220.127.116.11 of the WTO anti-dumping agreement and the extent to which investigating authorities can replace the actual cost items in the records kept by the exporter.
18.104.22.168 ……..costs shall normally be calculated on the basis of records kept by the exporter or producer under investigation, provided that such records are in accordance with the generally accepted accounting principles of the exporting country and reasonably reflect the costs associated with the production and sale of the product under consideration……
Costs should normally be based on the figures recorded in the accounts of the exporter. Two tests have to be met in order for the accounts to be used; a) in accordance with GAAP b) reasonably reflect the costs associated with the production and sale of the product.
In DS473 EU – biodiesel, the EU rejected the Argentinian exporters’ cost of production on the grounds that subsidised inputs meant that the costs in the producers’ records did not reasonably reflect the costs associated with the production and sale of biodiesel (i.e. b) above). The Appellate Body confirmed that the fact that inputs were subsidised was not sufficient to find that the accounts did not reasonably reflect the costs. As long as the subsidy was accurately recorded in the costs, the costs could not be rejected on the grounds that they are not reasonable i.e. there is not a reasonable test here.
Since this decision there has been some ambiguity about whether subsidised inputs could be replaced when constructing normal value. Many commentators have pointed out, however, that the word ‘normally’ means that there are other situations apart from a) and b) where the costs can be rejected.
The panel in DS529 provides useful clarity in relation to this point.
we consider that the term "normally" in Article 22.214.171.124 was used by the drafters deliberately to introduce a difference to the meaning of the sentence and cannot be reduced to a mere reference to the conditions that follow the words "provided that" (7.115)
The panel does not define what situations may fall outside of the ‘normal’ situation. But it is implicit in its overall rulings that particular market situation may constitute such a non-normal circumstance.
we do not believe that this dispute requires us to define precisely under what circumstances an investigating authority would be allowed to depart from the obligation to use the exporter's records on the basis of the term "normally (7.117)
The panel declined to be specific about that the non-normal situations that would justify the exporter's records not being used. This makes sense as each situation should be judged on the particular circumstances that apply.
The panel did rule that the investigating authority should give meaning to the whole obligation in 126.96.36.199 when relying on the word ‘normally’ i.e. it should examine whether the records satisfy the two explicit conditions (7.117). If the two conditions are met, the authority can then decide if it is a normal or non-normal situation. Australia found a PMS and rejected domestic sales as the basis of normal value. However, it did not check the two conditions meaning that it "failed to give effect to the whole of the obligation in that provision" (7.126).
The panel is perhaps taking an overly procedural approach to this issue. The WTO Agreement requires the two conditions to be assessed in a normal situation and the panel is extending the same requirement to non-normal situations. However, while the accounts may be in accordance with GAAP' and 'reasonably reflect the costs', they can still be rejected in the case of a non-normal situation such as a PMS. In other words, the two conditions become redundant if an investigating authority has determined that a PMS does not permit a proper comparison. It is not clear why the panel would insist that these 2 conditions still be assessed.
Benchmarks can replace distorted costs but with conditions
The panel confirms that information from other sources can be used to replace distorted costs but imposes a condition that the benchmark should be appropriately adapted:
…where the investigating authority uses information other than that contained in the records kept by the exporter or producer to construct the cost of production, it has to ensure that it adapts the information appropriately. (7.159)
What this means in practice is not 100% clear. The panel reasonably seems to be suggesting that the authority cannot just apply any old benchmark. Rather, it must 'adapt' it to be appropriate to the exporter's country of origin. This might mean that adjustments to the benchmark may have to made in order that it is relevant to the country of origin. The panel gives the example of ensuring that the profit amounts reflect the specific circumstances of the exporter concerned. Another example might be if the country of origin has a massive supply of a particular raw material, one would expect the non-distorted cost might still be cheaper than the cost of the same raw material from a country that has only a limited supply. Thus, although the actual cost can be rejected because it is distorted, the investigating authority must still assess whether the comparative advantage that exists would give the exporter a comparative advantage in the absence of the distortion.
The panel adds further conditions on the use of such information in terms of not automatically replacing all costs where there are costs unaffected by the distortion.
assuming arguendo that Article 2.2 allows the investigating authority to replace distorted costs in constructing "the cost of production in the country of origin", this provision also requires the investigating authority to consider available alternatives for replacing recorded costs so as to use the costs that are unaffected by the distortion to the extent possible. (7.162)
In practical reality, choosing benchmarks is not always straightforward and it will not always be possible to find benchmarks that exactly reflect the situation in the country of origin for the particular exporter. However, what the panel is clearly saying is that investigating authorities must explain the “choice of the final benchmark in light of” alternatives (7.163).
The panel is also saying that that you can’t just replace aggregated costs with a benchmark and that any undistorted costs for each exporter must be used.
When dealing with market economy situations, this makes sense. If one input is subject to government distortion, it is logical that only the distorted cost can be rejected and replaced by a benchmark. Other inputs and costs not affected by the distortion can be used.
However, when dealing with situations with systematic distortions across the whole economy, the justification for replacing all costs with benchmarks becomes stronger. In the case of non-market economies, or economies in transition, it may still be possible to reject all costs and use benchmarks on an aggregated basis. That is, if it is established that the distortions are systematic and affect all costs, no costs "are unaffected by the distortion" and thus can all be rejected and replaced. But, if the Australia copy paper panel is correct, it is necessary for authorities to determine whether any costs are unaffected by the distortions in such situations.
This panel is useful in providing some guidance on WTO provisions covering PMS. However, there are some major ambiguities and there needs to be more jurisprudence for more certain guidance to emerge.