12 February 2020

RELATED expertise

On 27 January 2020, the WTO Dispute Settlement Body, made up of all WTO Members, adopted the panel report in a dispute initiated by Indonesia regarding anti-dumping measures imposed by Australia on imports of A4 copy paper from Indonesia (“Australia –A4 Copy Paper or “DS529”).[1] Adoption of the DS529 report, without appeal by either Australia or Indonesia, concludes a highly anticipated dispute, the implications of which go well beyond the product involved and the countries themselves.

In this in-depth Trade Law Bulletin Moulis Legal Senior Associate Charles Zhan provides an analysis of the dispute and the panel report, and thoughts on its implications.

Dumping 101 – sounds simple but it isn’t

If an exporter sells a product to Australia at a price that is less than its home market price for the same product, or below-cost, and the price of the product causes injury to the Australian industry that makes the same kind of product, then the Australian government may impose a special customs duty (“anti-dumping measures”) to equalise the price with the exporter’s home market price or full product cost, whichever is higher. In that way it is hoped and intended that the Australian industry receives some price relief. This is permitted under the WTO Anti-Dumping Agreement,[2] which many WTO Members have implemented into their domestic laws.

A clash at the cutting edge of today’s trade tensions

The key issue raised by Indonesia in DS529 concerned the question of an exporter’s legitimate costs, for the purposes of working out the full product cost, and ultimately the existence of “dumping”.

The anti-dumping tools questioned by Indonesia were first used by the Australian investigating authority (“IA”) against imports of steel and aluminium based products from China in 2010. The same or similar practices have been imitated by other WTO Members since then, as an alternative way of counteracting imports from highly cost and price competitive countries. These countries have typically been those with economic or political systems, or industry “pockets”, that are considered by their trading partners not to have been “marketised” to an acceptable extent. Targets have included not only exporters from China, but also from Russia, Indonesia and even South Korea.

Can differences in economic systems and in market regulation form the basis for trade retaliation? Or does this undermine the fundamental principle on which world trade is based, that of comparative advantage? These are the key concerns underlying the DS529 dispute, and ones like it.

A peculiar solution for a particular situation

As mentioned, the “fair” or “normal” price level to which an anti-dumping measure seeks to lift the price of an imported product is the higher of the home market price and the full cost of the product. Under the Anti-Dumping Agreement, the home market price need not be used as the determinant of the normal price level (“normal value”) if there is what is called a “particular market situation” (or “PMS”) in the exporter’s home market,[3] or if the home market price is below cost.[4] In each case the full product cost, plus an uplift for profit, can be used as the normal value instead of the home market price.[5]

In Australia, PMS-based normal values – which have always been higher than both the home market price AND the exporter’s full product cost, have typically been calculated by the IA along the following lines:

  • First, find a PMS, on the rationale that the exporter’s input costs are artificially low, and that it is its government’s policies or regulations that are to blame. This allows the home market price to be (largely) disregarded in working out the normal value.
  • Secondly, because disregarding the home market price means the IA can use the full product cost for the normal value, and because it was one or other of those product costs that caused the PMS in the first place (eg. that cost being “artificially low”), the IA calculate the full product cost using higher “benchmark” costs. Such benchmark “surrogated” cost or costs are borrowed from a third country or amalgam of countries. This ignores the exporter’s actual costs, as deemed to be too low by the IA, and surrogates them with higher costs, leading to a higher normal value.

In DS529, Indonesia challenged the Australian IA’s PMS finding with respect to the Indonesian A4 copy paper market, which caused the Indonesian exporters’ domestic sales process to be ignored, and the resultant cost surrogation that was used to work out the normal value for Indonesian A4 copy paper. Indonesia was broadly successful in both respects.

However, what these successes may actually lead to in practice, for the Indonesian exporters directly involved and other keen observers, such as China and its exporters, is not clear.

Brakes put on use of particular market situation

The phrase “particular market situation” appears in Article 2.2 of the Anti-Dumping Agreement, but is undefined. The phrase was considered by a panel in the General Agreement on Tariffs and Trade (“GATT”) years in EEC – Cotton Yarn[6] but had never been interpreted by a WTO panel.

Notably, in the EEC – Cotton Yarn case, Brazil argued that its home market prices were too high due to currency exchange rate measures, and that the EEC should have recognized that and not used those high prices as the normal value, under the “particular market situation” umbrella. In contrast, in recent years PMS has been more frequently cited by investigating authorities, especially in Australia, as a way of rejecting home market prices which are considered to be “artificially low” due to policy, legal and/or regulatory interventions.

The panel in DS529 considered that a PMS must be something “distinct, individual, single, specific but that does not necessarily make it unusual or out of the ordinary — i.e. exceptional”,[7] but otherwise appears to have allowed Members broad leeway in deciding what might qualify as a PMS. Indonesia’s arguments:

  • that the identification of a PMS should not take into account government actions; and
  • that only situations having an exclusively unilateral impact on home market sales could form the basis for a PMS,

were not accepted by the panel.

But the WTO panel did put some bounds on the finding of a PMS – bounds that proved to be critical to Indonesia’s success on the facts of the case. The panel noted that the phrases “particular market situation” and “permit a proper comparison”, which appear in the same sentence in Article 2.2 of the Anti-Dumping Agreement, necessarily function together to give meaning to that part of the treaty:

In our assessment, the phrases “particular market situation” and “permit a proper comparison” function together to establish a condition for disregarding domestic market sales as the basis for normal value. Specifically, that domestic sales “do not permit a proper comparison” must be “because of the particular market situation”. If domestic sales do permit a proper comparison, then they cannot be disregarded as the basis for normal value, regardless of the existence of the particular market situation and its effects, whatever those may be. We find no functional purpose is served by incorporating into the meaning of “particular market situation” part of the function that will necessarily be served by the terms “because of” and “not permit a proper comparison”. Accordingly, we find that “capable of preventing a proper comparison” is not a necessary qualification for a situation to constitute the “particular market situation”.[8]

For the panel, the Australian IA’s identification of the existence of a PMS was only a job half-done. The analysis must continue on to the question of whether the PMS prevented a “proper comparison” of home market prices and export prices. That additional inquiry remains an imperative obligation on the part of the investigating authority concerned.

In DS529, the panel found that the IA’s disregard of the Indonesian exporters’ home market prices was inconsistent with the requirements of Article 2.2, because the IA considered the PMS automatically rendered the home market sales as being “not suitable”. The IA failed to engage with the question of whether the home market price and the export price could be properly compared nonetheless:

We find a deficiency in the [IA]’s examination in this case because it focused exclusively on the domestic sales and domestic prices, without taking into account the export prices with which the domestic prices would be compared. In particular, the examination does not address the question whether the domestic prices could be properly compared with the export prices despite the effects of the particular market situation.[9]

The DS529 panel report concludes:

On the basis of the above findings, we determine that the [IA]’s disregard of Indah Kiat’s and Pindo Deli’s domestic sales (and consequently of their domestic prices) as the basis for normal value was inconsistent with the requirement to examine whether sales in the exporting country’s market do not “permit a proper comparison” because of “the particular market situation” in Article 2.2 of the Anti-Dumping Agreement. Specifically, where a particular market situation was found to affect domestic market sales prices solely as a result of a decreased cost for an input that was used identically to produce merchandise for the domestic and export markets, the investigating authority was obligated to assess the effect of the particular market situation on the domestic price in relation to the effect on the export price when determining whether domestic prices permitted a proper comparison with those export prices.[10]

However the panel did not say that a low input cost – or more specifically a decrease in an input cost – that equally affected the manufacture of the product concerned (whether sold in the exporter’s home market or exported) would necessarily allow for a proper comparison. The panel said:

7.80     In our view, how domestic prices and export prices of an individual exporter are affected notwithstanding an equal decrease in input costs is likely to depend significantly upon a number of factors, including the prevailing conditions of competition in each market and the existing relationship between price and cost. We consider that an exporter may find itself with different options in respect of how to take advantage of an input cost decrease depending on market conditions in each market. This is similar to a situation when a cost increase occurs and the exporter faces differing market conditions in domestic and export markets such that the exporter is able to pass on the cost increase to customers in one market but unable to do so in the other.

7.81     Accordingly, we are not persuaded that a low-priced input used identically to produce merchandise for domestic and export markets will necessarily have the same effect on domestic prices and export prices and therefore necessarily permit a proper comparison. Rather, we find that whether the exporter’s domestic sales permit a proper price comparison with the export price is a question that can only be ascertained through an examination of relevant factual circumstances.

This seems to leave open the possibility that a change in an input cost – in the example given by the panel, a decrease – having the necessary “artificiality” could create some incomparability between the home market sales and the export sales depending on the exporter’s price response to the change. This seems odd, but the concept of a PMS has always been an ambiguous and odd feature of the Anti-Dumping Agreement. Given these arguendo comments by the panel it is likely to retain some mystery unless and until a working Appellate Body is returned.

Does it really matter whether a PMS exists, and whether it prevents a proper comparison of home market and export prices? In Australia, the answer is “no”, because a PMS finding is not the only legal route to cost surrogation, as we now explain.

The real issue is cost surrogation

The panel’s rulings and the debate that brought those rulings about are of systemic importance, and accordingly have been greeted by trade lawyers with excitement. Giving meaning to PMS is of contemporary relevance to those WTO Members who have recently legislated harsher dumping margin determination methodologies with PMS as a “gateway” thereto.

That said, the real issue is the dumping margin determination that leads to the imposition of the anti-dumping measures. The existence of a PMS does not by itself dictate the finding of a higher or lower dumping margin – it simply allows the IA to determine normal value using the full cost of production and sale, plus a profit, or using the exporter’s prices to a third country, in its discretion. But cost always has to be worked out in determining whether the home market prices are above cost or not, and if they are not then it is the full cost of production and sale, plus a profit (“CNV”), or the third country export price that will be used as the normal value instead.

Accordingly, how the cost of production, which is central to deciding the “above cost” question and working out the CNV, is the “kingpin” issue. Specifically, for the Indonesian exporters in DS529, and other exporters who have been in the same situation, the real issue lies in the “surrogation” of higher costs in working out the CNV. Higher costs increase the likelihood of below cost home market sales and of a CNV that is higher than the home market price. Cost surrogation has been the centerpiece of the Australian IA’s treatment of Chinese steel and aluminium products since 2010, and underpins many of Australia’s 17 anti-dumping measures currently in place against Chinese imports, and one other against ammonium nitrate from Russia.

Article 2.2 of the Anti-Dumping Agreement provides that a CNV should be calculated using the “cost of production in the country of origin” plus a reasonable amount for administrative, selling and general costs, and for profits. Article provides further rules about the determination of the cost of production:

…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.

Is an investigating authority allowed to derogate from what appears to be an obligation to use an exporter’s actual cost records and instead determine one or other of those costs by reference to an “out of country” benchmark, if the investigating authority perceives there to be some “distortion” in the exporter’s actual cost? Of late there have been a rash of WTO cases that have considered this question.

In EU-Biodiesel (Argentina)[11] and Ukraine – Ammonium Nitrate,[12] the WTO Appellate Body confirmed that the conditions provided in Article, particularly the phrase “reasonably reflect the costs associated with…”, are not to be read as imposing an additional “reasonableness” standard on the cost record being examined. That is to say, the exporter’s cost record should be used as long as it properly reflects the actual costs incurred, and the record is kept in accordance with the applicable generally accepted accounting principles (“GAAP”). Thus, whether or not the cost looks too low, or too high, or “unreasonable” to an investigating authority is irrelevant in determining if the two explicit conditions under Article have been met.

DS529 opens the gate to non-exporter cost data, a little

In DS529, Australia sought to justify its disregard of the Indonesian exporters’ cost records, and its surrogation of higher Brazilian and South American export prices of pulp to certain third countries in the Indonesian exporters’ CNV, on its view that the exporters’ costs did not “reasonably reflect competitive market cost” under Australian regulations. Australia drew attention to the phrase “shall normally” at the beginning of Article, and argued that the PMS-related cost distortion created a situation outside of the normal and ordinary circumstances envisaged by that phrase.

The panel accepted that the word “normally” should be given full effect, and could indeed provide a possible basis for an investigating authority to depart from the obligation to use an exporter’s cost record, in addition to departure as permitted under the two exclusionary conditions also mentioned in Article However, the Panel emphasised that the two conditions still needed to be considered, and be given full effect in the investigating authority’s determination:

We conclude that in relying on “normally”, the investigating authority should give meaning to the whole of the obligation in Article, first sentence, and should therefore examine whether the records satisfy the two explicit conditions and provide a satisfactory explanation as to why, nonetheless, it finds compelling reasons to disregard them.[13]

In light of Australia’s argument that the investigating authority rejected the exporters’ cost records based on its finding, under the Australian regulation, that the pulp cost did not “reasonably reflect competitive market cost”, rather than making a finding with respect to the two explicit conditions required by Article, the panel considered that the investigating authority failed to comply with the obligation under Article

Surrogation contemplated, “country of origin” cost principle not disrupted

A separate, and perhaps more fundamental issue the panel was required to determine was whether the Australian investigating authority’s surrogation of third country costs in the Indonesian CNV was a separate breach of the “cost of production in the country of origin” requirement under Article 2.2 of the Anti-Dumping Agreement. This concerns the replacement of the Indonesian exporters’ cost for wood pulp – being a raw material used for the production of A4 copy paper – with Brazilian and South American export prices of pulp to China or Korea. The two Indonesian exporters used wood pulp which they either self-produced in Indonesia or purchased from a related party who self-produced the pulp in Indonesia. The wood pulp used by the Indonesian exporters was not sourced from Brazil or South America, nor from China or Korea.

Generally speaking, when the normal value is a CNV, “cost surrogation” has a direct and substantive impact on the dumping margin calculation. This is especially so where an investigating authority is only given high-cost alternatives by the domestic industry applicants and is pressured to use them, or is prevailed upon by government policy to itself go out and find higher costs that could be used as surrogate costs. Costs that might more reasonably act as a proxy cost, taking into account the particular economic and production conditions of the exporter and of the exporting country, are not used, nor is that even a recognised test pursuant to which they should be selected.

In this regard, the panel noted that it had already determined that Australia’s departure from the use of the exporters’ financial records was in breach of the obligation under Article, and that it was therefore unnecessary to address the question of whether costs that can be used are only those “in” the country of origin. The panel did however note and endorse the position established in the Appellate Body report in EU-Biodiesel (Argentina):

7.132    The expression “cost of production in the country of origin” in this provision has been understood as “a reference to the price paid or to be paid to produce something within the country of origin”. Normally, and as reflected in the obligation set out in the first sentence of Article, the cost of production in the country of origin should be calculated on the basis of cost information from an exporter’s own records. However, as explained by the Appellate Body in EU – Biodiesel (Argentina):

In circumstances where the obligation in the first sentence of Article to calculate the costs on the basis of the records kept by the exporter or producer under investigation does not apply, or where relevant information from the exporter or producer under investigation is not available, an investigating authority may have recourse to alternative bases to calculate some or all such costs. [footnote omitted]

7.133.   We recall that the ADC did not use Indah Kiat’s and Pindo Deli’s pulp costs to calculate their respective costs of production of A4 copy paper for the purpose of constructing normal value. We have found in the previous section that in disregarding Indah Kiat’s and Pindo Deli’s costs, the ADC acted inconsistently with the first sentence of Article Accordingly, in the light of the above Appellate Body statement from EU – Biodiesel (Argentina), with which we agree, there was no legal basis for the ADC to have used third-country export prices of pulp as a proxy for Indah Kiat’s and Pindo Deli’s pulp costs when constructing normal value of A4 copy paper under the terms of Article 2.2. It follows that the ADC’s use of Brazilian and South American export prices of pulp to China and Korea as a starting point for the calculation of the costs of pulp in Indonesia was inconsistent with Article 2.2.

Although declining to rule on Indonesia’s complaint about the use of costs from outside Indonesia, the panel did provide some comment in relation to how cost replacement – if warranted – should be performed. Noting that one of the Indonesian exporters concerned was an integrated producer of both wood pulp and the final copy paper product, and therefore did not have to purchase wood pulp on the open market, the panel found that it was inappropriate to use a “purchase price” of raw material to replace the cost of production without adjusting for the profit embedded in such a purchase price.

Indonesia also argued that any cost surrogation should have been performed at the actual production input level, and be limited to address the alleged distortion. Accordingly, Indonesia argued that for the integrated producer that produced copy paper from woodchip, the replacement cost should have been limited to woodchip cost. It was the woodchip cost that the Australian IA had argued was the input material directly impacted by the alleged PMS, by reason of the Indonesian government’s industrial policies, including export bans on logs. The panel agreed, and determined that it is inappropriate to replace the cost of an integrated producer at a semi-finished product/material level, where it is known that the distortion is at the upstream, unprocessed material level. The panel stated in its report:

7.163    The circumstances of the investigation, in our view, called for the ADC to consider an alternative to replacing Indah Kiat’s cost of producing pulp with the pulp benchmark which replaces all the costs used in producing pulp with external information. We note the ADC’s above findings to the effect that the source of the distortions was in Indonesia’s timber market. …In light of the evidence on the ADC’s record and the ADC’s own findings regarding the source of the distortion, we find that the ADC should have considered using a replacement cost for woodchips in combination with Indah Kiat’s other costs for producing pulp which were not found to be affected by the distortion (labour, energy, etc.). If the ADC had undertaken such analysis, it should have explained its choice of the final benchmark in light of this alternative. The Final Report, however, contains no such explanation.

7.164    We are careful not to substitute our own judgment for that of the ADC as to what costs could have been feasibly utilized on the basis of the information before it. However, we recall that pursuant to the affirmative obligation under Article 2.2 to use the “cost of production in the country of origin”, it is incumbent upon the investigating authority to explore the alternative methodologies that would allow it to arrive at “the cost of production in the country of origin” by utilizing those components of the producer’s costs that are unaffected by the distortion, assuming arguendo that Article 2.2 allows for replacement of costs distorted by the effects of a particular market situation.[14]

Where to from here?

The panel report in DS529 was highly anticipated. Australia’s “defeat” captured a lot of attention and was quickly reported by the media. But like many WTO disputes, the real impact of the decision and how it will now shape Australia’s trade remedy practices in terms of implementation of the panel report in the case concerned, and the legality of existing anti-dumping measures based on similar reasoning and in future cases that will raise similar issues, is not clear.

Particular market situation – down but not out

The Dispute Settlement Body meeting note for the adoption of the panel report provides some intriguing insights:

Indonesia said a number of important issues related to the conduct of anti-dumping proceedings was raised in the DS529 dispute, including the determination of “normal value” (i.e. the domestic market price) for the product under investigation and the calculation of certain costs of production, with the panel finding in Indonesia’s favour on these points. Indonesia thanked Australia for its spirit of cooperation in agreeing not to appeal the panel ruling.

Australia said that while it was disappointed with the outcome, it intended to work closely with Indonesia in order to ensure prompt implementation of the findings. It said the dispute addressed important systemic issues not previously considered, notably with respect to the interpretation of “particular market situation” under Article 2.2 of the Anti-Dumping Agreement, where it welcomed the panel’s findings on this issue.

Russia said it had numerous concerns with the panel ruling which it said undermined the security and predictability of the multilateral trading system. Among other complaints, Russia criticized the panel’s interpretation of the term “particular market situation”, which it said would allow investigators to find dumping where no dumping exists. Thailand thanked the panel for its clear and well-written report and said it would follow this matter closely.

The Australian and Russian reactions support the proposition that Indonesia’s challenge has not entirely “shut down” the use of PMS as a tool to address “market distortion” whether caused by regulatory or industrial policies, or otherwise. The panel report suggests that a PMS can be brought on by something “exceptional” that impacts upon or occurs in a market, as long as the investigating authority can establish why that market difference then denies the ability of the sales on the home market and the export market to permit a proper comparison.

The panel’s insistence on the importance of a “proper comparison” analysis does mean that individual exporters will have a greater role and opportunity to argue their own case. It is likely that investigating authorities will need to make determinations about the ability of home market and export sales to allow a proper comparison on a case-by-case and company-by-company basis. The broad brush approach of finding that there is a distortion in the entire “iron and steel” or “aluminium” industry, due to industrial polices of the exporter’s government (such has been the case in Australia’s treatment of Chinese steel and aluminium product exporters) will not of itself establish an inability to compare home market and export sales for margin determination purposes. The DS529 panel report uses the situation of an individual exporter’s reaction to a changed input price in its pricing in the respective markets as providing a possible example of the inability to compare sales for normal value purposes.

The analytical and evidentiary burden on the part of an investigating authority wishing to rely on PMS is therefore likely to be increased, and the boundaries of any PMS findings that continue to be made could be the subject of future WTO disputes.

New ambiguity about what doesn’t “normally” apply in cost determination

The panel report’s acceptance of Australia’s argument concerning the ability of an investigating authority not to use an exporter’s actual cost record, based on the fact that a “normal” situation might not be presented, opens up some leeway for departing from actual cost records without indicating when that might be appropriate or what might be the result

At the same time it must be noted that Australia has acknowledged that its domestic law requirement that costs must “reasonably reflect competitive market costs…” to be acceptable for normal value purposes is different to and cannot be found in the text of the Anti-Dumping Agreement. The panel seems to think that this consideration could be relevant to normality (the “shall normally” test), provided that the investigating authority properly takes the explicit conditions under Article (of GAAP, and that the records are a reasonable reflection of the costs) into account. Whether there would ever be compelling reasons to find abnormality, rather than normality, if the two conditions are met, and what the investigating authority then does with respect to normal value calculation, are not further clarified in the DS529 report. Nonetheless the door has been left slightly ajar for “exceptional” circumstances.

Cost surrogation (if at all justified) must be sympathetic to the exporter

The panel report provides helpful clarification at the more detailed level about how cost surrogation should be carried out. The comments concerning the level at which cost surrogation should occur will be particularly welcomed by exporters, especially integrated producers. There have been a number of cases in Australia of integrated steel producers having their own costs rejected and surrogated at an intermediate product level (say hot rolled coil or steel billet) when they manufactured the final product from scratch using raw material inputs (such as iron ore, coke and coking coal) sometimes mostly imported from Australia!

Costs “in” the country of origin

In implementing the DS529 report, perhaps the biggest question is how the Australian investigating authority will react and respond to the reaffirmation of the requirement to use the “cost of production in the country of origin” to work out the CNV. Of course, the investigating authority could drop the PMS finding, and the finding that the Indonesian exporters’ costs did not “reasonably reflect the cost of production”, and undertake the dumping margin calculation for those exporters in the usual way. However it is self-evident that cost surrogation using costs in the country of export will rarely create the same punitive effect as replacing, say, an Indonesian woodchip cost with prices of pulp from Brazil to Korea (as in this case), or a Chinese steel cost with the prices of semi-finished steel products in Latin America (as in another Australian case, concerning grinding balls).

The DS529 panel did not think it worthwhile to re-address the issue of the costs in the country of origin/export, and simply expressed agreement with the Appellate Body report in EU – Biodiesel (Argentina). What has been somewhat at dispute and remains controversial about that report is the Appellate Body’s comments on the use of out of country “data” as part of the determination of the cost of production in the country of origin under Article 2.2. The Australian IA has argued in the Australian domestic context that the EU – Biodiesel (Argentina) decision provides support for its surrogation of foreign cost benchmarks instead of using costs in the country of origin. This is despite the fact that EU – Biodiesel (Argentina) very clearly states that the lack of restriction on the “sources of information” that can be used:

…does not mean that an investigating authority may simply substitute the costs from outside the country of origin for the “cost of production in the country of origin”. Indeed, Article 2.2 of the Anti-Dumping Agreement and Article VI:1(b)(ii) of the GATT 1994 make clear that the determination is of the “cost of production […] in the country of origin”. Thus, whatever the information that it uses, an investigating authority has to ensure that such information is used to arrive at the “cost of production in the country of origin”. Compliance with this obligation may require the investigating authority to adapt the information that it collects. It is in this sense that we understand the Panel to have stated that Article 2.2 of the Anti-Dumping Agreement and Article VI:1(b)(ii) of the GATT 1994 “require that the costs of production established by the authority reflect conditions prevailing in the country of origin”.[15]

This position concerning the “source of information” has also benefited from further clarification in the Appellate Body report in Ukraine – Ammonium Nitrate. There the Appellate Body explained that the source of information could in some circumstances more suitably be a third country, where the exporter-producer concerned actually imported its inputs from that third country. Another scenario for legitimate recourse to alternative sources is where there is an absence of relevant information from the exporter altogether. This explanation makes sense. It highlights that the scenarios where “out of country” data might be relevant to the consideration will be very limited, and requires an investigating authority to take into account the specific production arrangement being considered, and to be guided by the ultimate requirement of using a cost that can be said to be “in” the country of origin.

Make no mistake, in EU – Biodiesel, the Appellate Body found that the EU’s use of soybean export prices from Argentina did not and could not meet the requirement of being a cost in the country of origin (in that case, Argentina), because an export price from Argentina is not a cost in Argentina. Similarly, in Ukraine – Ammonium Nitrate, the Appellate Body found that Ukraine’s use of Russian gas prices at the German border, even though adjusted for the transportation cost back to Russia, did not qualify as costs in the country of origin (in that case, Russia) as required by Article 2.2. A choice by the Australian investigating authority to continue to defend the use of completely foreign prices that have no connection with the country of export at all, and are intended to be as far removed from the country of export as possible, would run counter to the panel and Appellate Body interpretations of the relevant provisions of the Anti-Dumping Agreement.[16] It must now be accepted that such a methodology is inconsistent with the requirements of Article 2.2.

Making the situation even more delicate in Australia is the fact that the investigating authority’s argument that its use of third country costs as costs in the country export is permissible has so far prevailed before the domestic courts. Most recently, in the Changshu Longte case,[17] the Full Federal Court effectively upheld the Australian IA’s claim that “cost in the country of export” is whatever the Minister so determines. That is, as long as the Minister turns her or his mind to the question of determining a cost in that country, then it is completely okay for the Minister to then determine that the price of pulp from Brazil is to be regarded as the cost of pulp in Indonesia.

Rule of law in a less-lawful international trade environment

DS529 tells us that the Anti-Dumping Agreement is not designed to adjust prices and costs so as to equalize those prices and costs to the level of higher third country or “world” prices and costs. Fellow WTO trade partners need to be dealt with on their own economic merits when it comes to anti-dumping protection.

Australia’s simplistic and broad-brush “artificially low cost” conclusions – about Indonesia’s paper industry, and China’s steel and aluminum industries, and Russia’s ammonium nitrate industry – cannot be maintained based on their current assumptions and need to be reviewed. One would think that the outcome of that review would be to ensure that Australia abides by the rule of law and, if not satisfied with the situation that has been reached, to then seek new international agreements, and new understandings, using the traditional means of negotiation and the trading of rights and obligations to achieve trade policy objectives.

We expect that there will be renewed consideration of the Subsidies and Countervailing Measures Agreement, and how it might be called upon to identify and counteract government policies that equate to and cause “financial contributions” as defined in that Agreement. Indeed, the recent WTO compliance panel report in US – Carbon Steel (India)[18] is a timely reminder of the availability of the public-body based subsidy measure as a valid countervailing tool for investigating authorities, where the circumstances permit. The Joint Statement of the Trilateral Meeting of the Trade Ministers of Japan, the United States and the European Union of 14 January 2020 has refreshed the trio’s call for stronger anti-subsidy rules,[19] and encourages greater engagement between WTO Members on these important trade remedy rules in 2020 and beyond.

Moulis Legal’s market-leading international trade team advises national and multinational clients in their cross-border and WTO affairs. This includes customs-related issues, trade rules investigations, cross border commercial transactions and disputes, business establishment and foreign investment.

For more information about this newsletter please contact Charles Zhan, Partner, on +61 2 6163 1000 or

This memo presents an overview and commentary of the subject matter. It is not provided in the context of a solicitor-client relationship and no duty of care is assumed or accepted. It does not constitute legal advice.

© Moulis Legal 2020


[1]       Panel Report, Australia – Anti-Dumping Measures on A4 Copy Paper, WTO Doc. WT/DS529/R (27 January 2020).

[2]       Marrakesh Agreement Establishing the World Trade Organization, Annex 1A – Agreement on Implementation of Article VI of the GATT.

[3]       Anti-Dumping Agreement, Art. 2.2.

[4]       Ibid, Art. 2.2.1.

[5]       Ibid, Art. 2.2.

[6]       GATT Panel Report, European Economic Communities – Imposition of Anti-Dumping Duties on Imports of Cotton Yarn from Brazil, GATT Doc ADP/137 (30 October 1995).

[7]       Australia – A4 Copy Paper, at [7.22].

[8]       Ibid, at [7.22].

[9]       Ibid, at [7.87].

[10]      Ibid, at [7.90].

[11]      Appellate Body Report, European Union, Anti-Dumping Measures on Biodiesel from Argentina, WTO Doc. WT/DS473/AB/R (6 October 2016).

[12]      Appellate Body Report, Ukraine, Anti-Dumping Measures on Ammonium Nitrate, WTO Doc WT/DS493/AB/R (12 September 2019).

[13]      Australia – A4 Copy Paper, at [7.117]

[14]      Ibid, at [7.163] to [7.164].

[15]      EU – Biodiesel (Argentina), at [6.73]

[16]      The Australian IA, in the grinding balls anti-dumping investigation mentioned, said:

Based on the depth of the market, and the geographic distance from China minimising the potential distortions of GOC influenced billet prices impacting on the Latin American billet export prices, the Commission considers that the Latin American export billet prices in FOB terms represent the best available information for competitive market costs of steel billets.

See Anti-Dumping Commission Report No 316 – Alleged dumping of grinding balls exported from the People’s Republic of China, at page 26.

[17]      Changshu Longte Grinding Ball Co., Ltd v Parliamentary Secretary to the Minister for Industry, Innovation and Science [2019] FCAFC 122.

[18]      Panel Report, United States — Countervailing Measures on Certain Hot-Rolled Carbon Steel Flat Products from India – Recourse to Article 21.5 of the DSU by India, WTO Doc. WT/DS436/RW (15 November 2019).