2.1 The WTO Anti-Dumping Agreement

Article 2.2 of the Anti-Dumping Agreement states that if the home price is not reliable because of, for example, a ‘particular market situation’ that does not allow a proper price comparison, the normal value must be, inter alia, a constructed price, i.e. ‘the cost of production in the country of origin plus a reasonable amount for administrative, selling and general costs and for profits’:

Article 2.2.1.1 of the Anti-Dumping Agreement further elaborates on the relevant costs that the investigating authority has to use for constructing the normal value. The first sentence of Article 2.2.1.1 reads:

While Article 2.2 refers to ‘the particular market situation’ as a possible basis for constructing the normal value, the Anti-Dumping Agreement itself does not define that term, leaving this to the national authorities’ discretion. In this regard, the EU anti-dumping rules contain some new elements as discussed below.

2.2 The EU Basic Regulation

The EU Basic Regulation incorporates the substance of Articles 2.2 and 2.2.1.1 above, but it also includes some provisions that have no direct counterpart in the Anti-Dumping Agreement. In particular, Article 2(3) of the Basic Regulation – analogues to Article 2.2 of the Anti-Dumping Agreement – additionally defines, in the second subparagraph, the ‘particular market situation’ as being ‘deemed to exist, inter alia, when prices are artificially low, when there is significant barter trade, or when there are non-commercial processing arrangements’.

Article 2(5) of the Basic Regulation, the first subparagraph contains language greatly resembling that of the first sentence of Article 2.2.1.1 of the Anti-Dumping Agreement, but supplemented with completely new details in the second subparagraph, as follows:

Thus, in constructing the normal value under Article 2(3) of the Basic Regulation, the EU authorities must follow the order of preference prescribed by Article 2(5) in using the data in the records of the investigated party (the producer or exporter) as a prime source, and shifting to other qualified material including ‘information from other representative markets’ as a last resort if the relevant costs are ‘not reasonably reflected’ in the party's records.

As for the background, the EU introduced the ‘WTO-plus’ provisions above – notably, the second subparagraphs of Articles 2(3) and 2(5) of the Basic Regulation respectively – via Council Regulation (EC) No. 1972/2002.Footnote ¹⁰ This instrument officially recognizes Russia as a market economy in anti-dumping procedures, waiving Russian exports to the EU from the special rules for determining the normal value in NMEs,Footnote ¹¹ under which home prices of the target product are substituted for prices in a third market economy country.Footnote ¹² But the same instrument also enables the EU authorities to replace the below market costs of the Russian product's inputs (in practice, mainly natural gas) with the benchmarks ‘from other representative markets’ outside of Russia.Footnote ¹³ In other words, the new provisions were adopted, in principle, to allow the NME methodology vis-à-vis Russia and any other market economy countries with respect to their exported product's input materials acquired under non-market terms.Footnote ¹⁴

Recitals 3 and 4 of Council Regulation (EC) No. 1972/2002 explain the rationales for the introduction of the WTO-plus elements in question with the need to provide further clarification and guidance on related issues. In particular, Recital 3 referring to Article 2(3) states that various ‘market impediments’ constituting a ‘particular market situation’ may not properly reflect supply and demand which results in ‘domestic prices being out of line with world market prices or prices in other representative markets’. With respect to Article 2(5), Recital 4 states that where the records do not reasonably reflect the costs concerned, the relevant data should be obtained from sources not affected by price distortions with a view to adjusting some items of the investigated party's records or establishing the costs of that party.

It is noteworthy that the EU approach of adjusting normal value elements is not unique.Footnote ¹⁵ As a fresh example, section 504 of the US Trade Preferences Extension Act of 2015Footnote ¹⁶ amending national anti-dumping law states that in the presence of a ‘particular market situation’, i.e. when ‘the cost of materials and fabrication or other processing of any kind does not accurately reflect the cost of production in the ordinary course of trade’, the investigating authority may use ‘any other calculation methodology’ in constructing the normal value. It is believed that this amendment was inserted to cope with challenges that could arise from China's potential transition to market economy status, and that ‘any other calculation methodology’ may comprise the use of surrogate country data.Footnote ¹⁷ In a recent anti-dumping administrative review on certain Korean goods containing steel as a key input, the US Department of Commerce, for the first time, relied on this amendment to make an upward adjustment to the reported production costs. For this, it made a cumulative impact assessment of several ‘market distortions’, including, inter alia, alleged distortions in the Chinese steel industry that reportedly flooded the Korean market with cheap exports depressing steel prices there.Footnote ¹⁸ Although such third country distortions taken individually did not constitute the sole basis for the cost adjustments, this case shows that the methodology in question as applied to third countries may impact on the cross-border movement of input materials. Despite the factual differences with the EU practice below, the US perception of a particular market situation as covering the case of production/input cost distortions essentially coincides with the EU position on this issue.

While we agree, in principle, with the very logic of the EU authorities that distorted components of the normal value should be corrected in a market-oriented way, we will see below that this stance fails to meet the current WTO legal standards. Therefore, we will argue in section 5.2 that the relevant provisions of the Anti-Dumping Agreement should be amended to address market distortions more properly in line with the WTO's anti-distortion approach under other comparable legal contexts.

3.1 Energy-intensive goods from Russia

With regard to the Russian energy-intensive goods, including fertilizers and steel products, the EU authorities held that, due to heavy State regulation, the price of natural gas remained ‘abnormally low’ – ‘far below market prices paid in unregulated markets for natural gas’ – being less than one-third of the export price of gas from Russia.Footnote ¹⁹ The EU authorities thus considered the gas costs to be ‘not reasonably reflected’ in the investigated party's records within the meaning of Article 2(5) of the Basic Regulation and adjusted them on the basis of the price of Russian gas sold for export at the German/Czech border (Waidhaus). It is stated that Waidhaus serves as ‘the main hub for Russian gas sales to the EU … the largest market for Russian gas … [having] prices reasonably reflecting costs … [and hence] a representative market’.Footnote ²⁰ Although Russia committed itself to liberalizing gas prices for industrial users in line with ‘normal commercial considerations’ upon its WTO accession in 2012, the EU authorities continuously relied on the Waidhaus price even after 2012.Footnote ²¹

In Acron and Dorogobuzh v. Council, the General Court of the EU, on 7 February 2013, confirmed that the EU authorities were ‘fully entitled’ to adjust the government-controlled gas price in Russia using the data from ‘more representative’ outside markets.Footnote ²² The General Court found that the external price at Waidhaus was ‘reasonable’ for being set by unrestrained market forces as indicated by ‘the volume of gas concerned’ and ‘the number of contracts negotiated’.Footnote ²³ It did not consider alternative benchmarks – such as, for example, gas prices in the US or Canada, or the price of gas entering the EU via different routes – to be more suitable, because the evidence was not compelling on the representativeness and comparability of those prices as juxtaposed with the price at Waidhaus.Footnote ²⁴

As for the EU legislation, the General Court observed that Article 2(5) of the Basic Regulation implementing Article 2.2.1.1 of the Anti-Dumping Agreement had to be interpreted, ‘so far as possible’, in the light of that WTO provision. But since Article 2.2.1.1 did not contain the provision analogous to the second subparagraph of Article 2(5), the General Court acknowledged that the former could not shed light on the proper interpretation of that part of Article 2(5).Footnote ²⁵ The judgment in Acron and Dorogobuzh v. Council was appealed, but the appellate process was later discontinued.Footnote ²⁶

3.2 Biodiesel from Argentina and Indonesia

As for the imports from Argentina and Indonesia, the EU anti-dumping measures targeted biodiesel produced from certain agricultural raw materials, namely Argentine soya beans and soybean oil, and Indonesian crude palm oil respectively. Argentina and Indonesia operate a Differential Export Tax (DET) system, under which the government levies a higher tax on exports of the relevant raw materials than that on the exported final product (i.e. biodiesel).Footnote ²⁷ The EU authorities determined that the DET system had effectively depressed the domestic prices of the raw materials in both countries to ‘an artificially low level’ – by reducing their exports on the one hand, and consequently increasing their domestic supply on the other – causing a downward pricing effect on biodiesel. Due to such a ‘particular market situation’, the EU authorities substituted the actual input costs for some adjusted reference prices for exportation that, in their view, corresponded to the level of international prices.Footnote ²⁸ However, on 15 September 2016, the General Court annulled the Council implementing regulations imposing anti-dumping duties on Argentine and Indonesian biodiesel on the grounds of flawed input cost adjustments.Footnote ²⁹

While the General Court recognized that, unlike Russia's case in Acron and Dorogobuzh v. Council, Argentina and Indonesia's DET systems did not represent direct regulation of prices of the raw materials, it admitted that this fact did not, in itself, preclude the cost adjustment methodology under Article 2(5).Footnote ³⁰ But it held further that not all government actions influencing the prices of raw materials would give rise to the discard of the parties’ records, as this would otherwise render the prime source of information as defined by the first subparagraph of Article 2(5) inutile.Footnote ³¹ Therefore, only the exporting country's measures causing ‘appreciable distortion’ of the prices concerned may trigger, the General Court said, cost adjustment on ‘another reasonable basis’ under the second subparagraph of Article 2(5).Footnote ³² Where the distortion was not ‘an immediate consequence of the State measure’, as it was in Acron and Dorogobuzh v. Council, the investigating authority must explain the operation of the market at issue and substantiate the measure's distorting effect ‘without relying … on mere conjecture’.Footnote ³³

With these considerations in mind, the General Court concluded that the EU authorities had failed to establish the existence of the DET-related appreciable distortion, as they had not examined how the difference between the export tax on the raw materials and that on the targeted final product – rather than the export tax on the raw materials alone – was affecting the price of those raw materials in the home market. In other words, the EU authorities should have considered the extent to which the price effect on the raw materials arising from the DET system per se would differ from that of a fiscal system without such tax differentiation.Footnote ³⁴ At the time of this writing, the General Court's findings were under appeal.Footnote ³⁵

3.3 Practical problems with input cost adjustments

The EU practice specifically suggests that the records of the parties under investigation are to be disregarded where there is a ‘particular market situation’ in the exporting country, which is evidenced by unreasonable (‘artificially low’) input costs attributable to a government intervention. The costs are to be considered unreasonable only in the context of an ‘appreciable distortion’, the concept that arguably raises practical difficulties in identifying the appreciable degree of a distortion in each particular case. For instance, the General Court held that the investigating authority should have compared the price effects under the DET and non-DET systems, but it created uncertainty as to the right threshold for the effects’ differential that would be able to justify market-based cost adjustments.

Under the EU methodology, the substitution of actual but distorted input costs for ‘reasonable’ market proxies increases the normal value of the subject product, because the prices set by supply and demand tend to be higher than those fixed or otherwise influenced by the State. Where the inputs account for a sufficiently large share in the total production cost of the investigated end product,Footnote ³⁶ the extent of such a normal value increase may be substantial. A higher constructed normal value resulting from the surrogate costs inflates the dumping margin, and hence the amount of the related anti-dumping duties. For example, the final dumping margins of Argentine biodiesel established with input cost adjustments were about five to six times as high as the provisional dumping margins, calculated with the actual home prices.Footnote ³⁷ Such high rates virtually closed the EU market – the then largest foreign destination for Argentina's biodiesel – causing the affected imports from this country to considerably drop in the year of imposition and to almost cease in the following years.Footnote ³⁸ Thus, it is not surprising that the relevant anti-dumping measures were challenged in the WTO.

4.1 Alleged WTO violation by the basic regulation ‘as such’

In considering the ‘as such’ allegations, the Appellate Body holistically reviewed the legal text of the EU provisions at issue, the legislative history with particular reference to Recitals 3 and 4 of Council Regulation (EC) No. 1972/2002, as well as the practice of the EU authorities and the General Court regarding the anti-dumping methodology in question.

Argentina argued that the second subparagraph of Article 2(5) of the Basic Regulation violated Article 2.2.1.1 of the Anti-Dumping Agreement by requiring the investigating authority to determine that the records of the party concerned did not reasonably reflect the production costs where such costs were considered to be artificially low as a result of a distortion, while Article 2.2.1.1 did not allow refusal of those costs by reason of their distortion.Footnote ⁴¹ But the Appellate Body disagreed, concluding that the second subparagraph in fact dealt with what should be done after the investigating authority had found, under the first subparagraph, that the costs in records were ‘not reasonably reflected’.Footnote ⁴²

Argentina contended further that the second subparagraph in question contradicted Article 2.2 of the Anti-Dumping Agreement by (i) requiring WTO-inconsistent action, (ii) or otherwise providing for the possibility of such an action.Footnote ⁴³ With respect to the first line of argument, the Appellate Body observed that the second subparagraph indeed did not rule out that EU investigators could use out-of-country information on the production costs without adapting it to the conditions of the country of origin. But the mere existence of this possibility did not mean, the Appellate Body said, that this provision mandated construction of the normal value on the basis of such an out-of-country proxy.Footnote ⁴⁴ In particular, the second subparagraph's phrase ‘any other reasonable basis, including information from other representative markets’ (emphasis added) suggested that such information was merely an illustration of ‘any other reasonable basis’.Footnote ⁴⁵ As for the second line of argument, the Appellate Body admitted that the second subparagraph was ‘capable of being applied’ in breach of Article 2.2 of the Anti-Dumping Agreement, but this was not sufficient for establishing a prima facie case of an ‘as such’ violation. The Appellate Body specifically found that Argentina had failed to prove that the provision in question restricted the EU authorities’ discretion to construct the cost of production in a WTO-consistent way.Footnote ⁴⁶

4.2 Alleged WTO violation by the basic regulation ‘as applied’

On the ‘as applied’ front, Argentina asserted that the EU anti-dumping procedures had failed to construct the normal value on the basis of the records kept by the Argentine exporter or producer, contrary to Article 2.2.1.1 of the Anti-Dumping Agreement.Footnote ⁴⁷ But the EU countered that the recorded production costs distorted by the DET system could not be considered as being ‘reasonable’ so as to meet the ‘reasonably reflect’ requirement of Article 2.2.1.1. In support, the EU submitted that the reference in that provision to ‘the costs associated with the production and sale’ (emphasis added) suggested a broader connection between the costs and production/sale covering the costs that would exist under ‘normal’ market conditions.Footnote ⁴⁸

The Appellate Body noted that Article 2.2.1.1 instructed the investigating authority to resort to the domestic records as ‘the preferred source for cost of production data’ when they simultaneously (i) comply with ‘the generally accepted accounting principles’ and (ii) ‘reasonably reflect’ the relevant costs.Footnote ⁴⁹ Satisfaction of the first condition does not necessarily secure conformity to the second condition. For example, this may be the case of companies’ financial statements that reflect the production costs of both the product under investigation and other products.Footnote ⁵⁰ The Appellate Body concluded that, contrary to the EU view, the terms ‘reasonably reflect’ under the second condition require the recording (reflection) of the costs to be reasonable, not the costs:

In addition, the Appellate Body confirmed that the second condition above refers to the actual costs incurred by the exporter or producer so that the investigators must compare the costs reflected in the records with the costs actually borne by that party, rather than consider which costs ‘would pertain’ to the production and sale of the product ‘in normal circumstances’ free from price distortions.Footnote ⁵² This implies that once the costs are recorded in an accurate and reliable manner, the investigators must accept them without considering alternative counterfactuals, namely ‘hypothetical costs’ that could have been incurred under imaginary more ‘reasonable’ circumstances.Footnote ⁵³ Applying these parameters to the case at hand, the Appellate Body eventually found that the EU's allegations about distorted domestic prices of soybeans in Argentina did not provide a sufficient basis under Article 2.2.1.1 for disregarding the relevant records and costs.Footnote ⁵⁴

With respect to Article 2.2 of the Anti-Dumping Agreement authorizing the construction of the normal value, the Appellate Body observed that this provision applied ‘harmoniously’ with the obligation in Article 2.2.1.1. While the constructed normal value includes the internal production cost, the Appellate Body made it clear that neither Article 2.2 nor Article 2.2.1.1 prevented investigators from utilizing external information if it was appropriate for calculating that cost. Accordingly, the investigating authority may use in-country and out-of-country evidence other than the records of the parties where (i) there is a need to analyze or verify the data in such records; (ii) the obligation to rely on those records does not apply;Footnote ⁵⁵ or (iii) the relevant information from the targeted exporter or producer is not available. But in the light of the language of Article 2.2, the information from other sources must be adapted to establish the cost of production in the country of origin.Footnote ⁵⁶

Argentina argued that the selected surrogate price reflecting international prices could not qualify as a domestic price, but the EU countered that the surrogate price was derived from a price at the border which could ‘by definition’ be both international and domestic prices.Footnote ⁵⁷ While the Appellate Body admitted the possibility of characterizing a border price as both international and domestic prices, it concluded that the surrogate price in question did not reflect the cost of soybeans in Argentina, because the EU used that price without necessary adaptation ‘precisely because it did not represent the cost of soybeans in Argentina’ to remove ‘the perceived [cost] distortion’ there.Footnote ⁵⁸

Moreover, Argentina appealed the panel's finding under Article 2.4 of the Anti-Dumping AgreementFootnote ⁵⁹ – the provision that requires the investigating authority to make a ‘fair comparison’ between the export price and the normal value, and make due allowance for ‘differences which affect price comparability’. The Appellate Body had reservations regarding the panel's proposition that differences arising from the methodology used for establishing the normal value could not be challenged under Article 2.4 as ‘differences affecting price comparability’.Footnote ⁶⁰ But since the Appellate Body had already confirmed the breach of Articles 2.2 and 2.2.1.1, it decided not to rule further on the Article 2.4 issue.Footnote ⁶¹

Finally, the Appellate Body agreed with the panel that the EU violated Article 9.3 of the Anti-Dumping Agreement by imposing anti-dumping duties in excess of the dumping margin that should have been established under Article 2.Footnote ⁶²

4.3 Assessment

In summary, the EU prevailed on its ‘as such’ claims, as the Appellate Body had found that the mere existence of the potential for the second subparagraph of Article 2(5) of the Basic Regulation to violate WTO anti-dumping law did not mean that the second subparagraph mandated WTO-inconsistent actions. However, the EU lost in the ‘as applied’ part for its failure to adapt surrogate prices to the conditions of Argentina as the country of origin.

Overall, this case largely indicates defeat of the EU, as it casts doubt on the very motive for the measures at issue. Specifically, the appellate decision disapproved of the dumping cost adjustments triggered solely by input market distortions per se, i.e. by the exact factor that the EU authorities have sought to counter through the methodology in question. Even keeping the current domestic rules at issue intact, the EU has to revise the challenged anti-dumping measures accordingly and discontinue controversial input cost adjustments in the future in order to avoid new trade frictions.

The appellate decision in EU–Biodiesel has far-reaching implications going beyond the scope of this dispute. The legal findings here will have some ‘precedential’ effect within the WTO judicial system and possibly lead to similar conclusions in the ongoing WTO disputes on the same EU methodology. The appellate interpretations in this case will also make it more difficult for other WTO members operating the EU-like regime to justify market-based cost adjustments. Further, following the expiry of certain WTO accession provisions envisaging NME assumption of China on 11 December 2016, Viet Nam on 31 December 2018, and Tajikistan on 2 March 2028,Footnote ⁶³ other relevant jurisdictions will hardly be able to ‘counterbalance’ the potential treatment of these countries as a market economy with the now invalidated cost adjustment methodology.Footnote ⁶⁴

The most important part of this case was obviously the clarification of Article 2.2.1.1 of the Anti-Dumping Agreement regarding the second condition for the domestic records to reasonably reflect the relevant costs. The Appellate Body found that reasonableness was a requisite feature of a company's records rather than the costs, with the constructed normal value having to reflect the actual costs concerned. This finding is correct for several reasons.

First, the rest of Article 2.2.1.1 essentially calls for proper reflection (and treatment) of cost allocations, certain non-recurring costs and costs affected by start-up operations. As a closest legal context for the second condition in question, this part of Article 2.2.1.1 can implicitly reaffirm the appellate interpretation that the reasonableness under that condition implies accuracy in recording of the costs.Footnote ⁶⁵

Second, some contextual support can also be found under Article 2.2.2 of the Anti-Dumping Agreement, which states that administrative, selling, and general costs and profits must be based on actual values in the country of origin, with no reasonability against certain market benchmarks required.Footnote ⁶⁶ Indeed, it would be arguably weird to impose such requirement for these integral elements of the constructed normal value, while exempting from it the production (input) costs that make up the rest of the same normal value.

Third, in the pre-WTO discussions on anti-dumping rules, it was widely considered that anti-dumping measures could not generally apply to imported end products by reason of dumped or below cost inputs (i.e. ‘input dumping’). Although the drafters failed to inscribe this proposition into the legal text, some commentators argue that such a historical background refutes the view shared by the EU authorities that the cost under Article 2.2.1.1 must be reasonable.Footnote ⁶⁷

Fourth, the Appellate Body's finding in question is in line with the previous WTO jurisprudence on Article 2.2.1.1 that concerned other anti-dumping calculation techniques. For instance, the US–Softwood Lumber V case considered the legality of the US ‘by-product offset’ methodology related to the normal value, under which a target company could receive an offset to its costs of production if it could show that the manufacturing of the product under consideration generated a by-product that had commercial value. There, the WTO adjudicators did not find any textual basis in Article 2.2.1.1 that would require the use of a ‘market value’ of by-products. Nor did they believe that Article 2.2.1.1 precluded the US investigating authority from using the ‘actual cost of the input – as it appeared in [the target company's] records – as the benchmark for valuing the by-product revenue offset’.Footnote ⁶⁸

As we saw before, the General Court also ruled against the EU anti-dumping measures on Argentine biodiesel.Footnote ⁶⁹ But it did so invoking the absence of an evident ‘appreciable distortion’, the market phenomenon that, according to the General Court's logic, would certainly arise from a government's strict price-setting policy but not necessarily from indirect public measures. The same reasoning may, in principle, be raised in the WTO proceedings as well. In particular, a disputing party can argue that, in relation to heavy price regulation by an exporting State (hence, an ‘appreciable distortion’), the investigating authority is entitled to use surrogate prices, despite the available valid records of producers, because Article 2.2.1.1 requires it to rely on those records only ‘normally’ rather than always. Indeed, this argument may be ‘encouraged’ by the Appellate Body's remark in EU–Biodiesel that for ‘resolving this dispute’ it did not need to consider ‘whether there [were] other circumstances [different from the two prescribed conditions to be met by the relevant records] in which the obligation … ‘normally’ to base the calculation of costs on the records kept by the exporter or producer … would not apply’.Footnote ⁷⁰ While the word ‘normally’ in Article 2.2.1.1 implies the existence of a ‘norm’ for investigators to use those records,Footnote ⁷¹ WTO panels in previous cases had suggested that deviation from that norm was possible if the records either (i) did not comply with generally accepted accounting principles or (ii) did not reasonably reflect the production and sale costs.Footnote ⁷² Accordingly, an appreciable distortion as such will hardly provide legitimate grounds for rejecting the records that meet both conditions. But even if the ‘encouraging’ appellate comment above would still suggest some room for the departure from that norm because of the appreciable distortion, the Appellate Body in the present case made it clear that under Article 2.2 of the Anti-Dumping Agreement, which applies ‘harmoniously’ with Article 2.2.1.1, the relevant data from other (market-based) sources must be adapted to the actual costs in the country of origin. To conclude, the per se existence of price distortions in the exporting country – regardless whether ‘appreciable’ or not – will likely fail, under WTO anti-dumping law, to justify recourse to surrogate values that were not properly adjusted to the prevailing actual conditions in that country's market.

Finally, it should be noted that the key issues in EU–Biodiesel concerning the cost adjustment methodology are also raised in the remaining four ongoing disputes. Among other common points, the complainants there specifically question the EU methodology under Article 2.4 of the Anti-Dumping Agreement. In essence, they argue that the EU authorities failed to make a fair comparison between a constructed normal value that neglected actual production/input costs and the export price that reflected those actual costs.Footnote ⁷³ As previously stated, the Appellate Body in EU–Biodiesel practiced judicial economy on this issue, although it had doubts about alleged non-application of Article 2.4 to the price differences arising from the methodology of determining the normal value. On the basis of the consultation/panel requests in the pending cases,Footnote ⁷⁴ one may expect that future rulings will probably show how such appellate reservations concerning Article 2.4 could work in practice, and will additionally examine the legality of adjustment of the other normal value elements, such as administrative, selling, and general costs and profits.

5.1 EU anti-dumping reforms

In April 2013, the European Commission adopted a proposal to modernize the EU trade defence system with a view to adapting it to the changing economic environment. The proposal calls for better transparency and predictability, faster investigations, easier access for small and medium enterprises, and discontinuation of the lesser duty principle in certain circumstances.Footnote ⁷⁶ In parallel, in order to cope with the consequences of the expiry of the aforementioned WTO accession NME provisions, the Commission has additionally insisted on rule changes that would address NME situations on the basis of objective country-neutral criteria, i.e. without official listing of NME countries as in the current practice.Footnote ⁷⁷

Following the agreement reached in a ‘trilogue’ meeting between the European Parliament, the Council, and the Commission on 3 October 2017, the European Parliament on 15 November 2017 approved proposed amendments to the Basic Regulation that will enter into force, once all remaining EU internal procedures have been completed.Footnote ⁷⁸ These amendmentsFootnote ⁷⁹ insert additional provisions in Article 2 of the Basic Regulation to lay down new rules for constructing the normal value in the case of certain State-created market distortions. Such a new approach will apply to imports from any qualifying country and replace the EU practice of using the so-called analogue country methodologyFootnote ⁸⁰ vis-à-vis pre-defined NMEs as regards WTO members. In relation to our topic, the amending legislation states:Footnote ⁸¹

Accordingly, in the presence of ‘significant distortions’ in the exporting country, the normal value ‘shall be constructed exclusively on the basis of costs of production and sale reflecting undistorted prices or benchmarks’.Footnote ⁸² As specified further, ‘significant distortions’ occur ‘when reported prices or costs, including the costs of raw materials and energy, are not the result of free market forces because they are affected by substantial government intervention’.Footnote ⁸³ As for possible indicators of such distortions, the investigating authority must consider ‘the potential impact’ of certain factors like the exporting country's market being served predominantly by government-controlled enterprises; State presence in firms that affects prices or costs; public policies or measures favouring domestic suppliers or influencing free market forces; inadequate bankruptcy, corporate, or property laws; distorted wage costs; or access to financing by institutions that implement public policy objectives or depend on the State.Footnote ⁸⁴

In constructing the normal value, the investigating authority may use relevant costs from a representative third country with a similar economic development level as the exporting country under investigation, undistorted international prices or benchmarks, or undistorted domestic costs. If there are multiple representative countries with the available data, preference will be given to a country ‘with an adequate level of social and environmental protection’.Footnote ⁸⁵ This, in turn, will require an examination of whether such countries comply with core conventions of the International Labour Organization and relevant multilateral environmental conventions.Footnote ⁸⁶ The constructed normal value must also include ‘an undistorted and reasonable amount for administrative, selling, and general costs and for profits’.Footnote ⁸⁷ The Commission will publish and regularly update country- or sector-specific reports on significant distortions abroad. These reports will be utilized in related anti-dumping investigations, subject to comments of interested parties.Footnote ⁸⁸

In sum, the amendments above add, for the first time, some labour and environmental dimensions to the EU anti-dumping process, clarify the scope of data for constructing the normal value, and confirm that State-distorted costs will not altogether be considered in calculations of the normal value.Footnote ⁸⁹ When ultimately put into force, such a clear-cut discard of distorted costs in the exporting country – even if properly reflected in domestic records – will most likely fail to meet the current WTO anti-dumping standards as we can learn from the appellate ruling in EU–Biodiesel. Not surprisingly, 11 WTO members – including Russia, China, Bahrain, Kazakhstan, Egypt, and Colombia, among others – have already expressed their concerns about the EU draft amendments in the WTO Committee on Anti-Dumping Practices, pointing to ambiguity of the concept of ‘significant distortions’ and inconsistency of the associated methodology with the Anti-Dumping Agreement.Footnote ⁹⁰ To shield the new methodology from possible WTO complaints in the future, the EU should arguably initiate WTO-level reforms as discussed in the next section.

In addition to the changes adopted by the European Parliament on 15 November 2017, the EU anti-dumping reforming process may also introduce some other elements involving the production costs of imports under investigation. According to the European Commission, the legislative amendments above do not replace the aforementioned modernization proposal of April 2013, which itself does not cover the issue of modifying the way of calculating dumping margins and is currently discussed in parallel trilogue talks.Footnote ⁹¹

Initially, deliberations on the 2013 proposal were deadlocked for nearly three years mainly because of the disagreement of major stakeholders on the lesser duty issue.Footnote ⁹² The EU authorities have systematically applied the lesser duty rule, imposing an anti-dumping duty below the dumping margin when it was deemed adequate to remove the injury to the domestic industry. WTO law recommends, but does not require, this practice.Footnote ⁹³ The impasse over the 2013 proposal was due to the fact that while the lesser duty benefits general consumers and import-dependent industries, it may significantly lower the level of protection for import-competing industries. In this regard, the Commission gives an example of the EU anti-dumping duties that were imposed on Chinese hot rolled coils at the rate of 19% corresponding to the injury margin, while the actual dumping margin was 102%. Further, making a comparison with the US that does not apply the lesser duty rule, the Commission also warns that the huge rate difference of about 112–245 percentage points between the US and EU anti-dumping duties on steel products from China could divert trade to the EU market, ‘putting additional pressure on EU industry and workers’.Footnote ⁹⁴

While France, Germany, and some other EU members wanted complete removal of the lesser duty rule, nearly half of the EU members opposed this.Footnote ⁹⁵ In December 2016, they reached a provisional compromise to lift this rule in the case of distorted raw material prices, provided that the raw materials (including energy) account for more than 27% of the cost of production in total and over 7% taken individually.Footnote ⁹⁶ If this proposal is eventually implemented, it will allow imposition of anti-dumping measures in the amount of higher than the lesser duty rates but still within the dumping margins in compliance with the Anti-Dumping Agreement. To some extent, this will enable the EU to counter related upstream market distortions abroad, even without resorting to the cost adjustment methodology that is now delegitimized by the WTO appellate decision. Such a lesser duty-related approach, which is arguably WTO-consistent, could also be followed by other relevant jurisdictions having the EU-like anti-dumping system.

5.2 WTO anti-dumping reforms

The EU and like-minded counties may also explore other options, such as pushing for legislative reforms in the WTO system that would retain the controversial part of the input cost adjustment practice. Such an initiative can be defended by the fact that the reasonableness standard under Article 2.2.1.1 of the Anti-Dumping Agreement contrasts sharply with other comparable legal regimes addressing similar market distortions.

In particular, the second Ad Note to Article VI:1 of the GATT, read together with Article 2.7 of the Anti-Dumping Agreement, allows importing countries to disregard home prices of the investigated product from certain NMEs in determining the normal value.Footnote ⁹⁷ Many WTO members, including the EU, have relied on this clause to substitute the prices in NMEs for prices in a selected market economy country. This technique is known as the analogue (or surrogate) country methodology. Since WTO anti-dumping law already permits such deviation from a standard way of calculating normal values vis-à-vis NMEs because of inappropriate government interventions there, one may rightly ask why this cannot apply, on the same rationale, to a market economy's upstream industries that supply investigated downstream producers with raw materials under non-market conditions created by State interference. Should inputs represent a significant proportion of a downstream product's price,Footnote ⁹⁸ there is not much difference whether the NME calculation approach applies to the entire normal value of the end product or only the input-related part of that normal value.

Furthermore, the resort to external market benchmarks is also permitted in the context of countervailing measures, i.e. special duties that operate in a similar way as anti-dumping duties but target subsidized imports to offset injurious effects of foreign subsidies. In US–Softwood Lumber IV, the US investigating authority found the imports of Canadian softwood lumber to be subsidized through ‘stumpage’ programs that conferred the right to harvest standing timber – the input used for making softwood lumber – on private companies on the payment of low fees. At issue was the pass-through of such an input subsidy to the softwood lumber exported to the US.Footnote ⁹⁹ To calculate the amount of the subsidy and hence the countervailing duty, the US authority replaced State-distorted domestic prices of timber in Canada, which is a market economy, with adjusted out-of-country prices. In this dispute and thereafter, the Appellate Body confirmed the importing member's right to rely on out-of-country proxies where private prices in the subsidizing country are distorted because of the predominant role of the government in the relevant market.Footnote ¹⁰⁰ Thus, it can be argued that approval of the use of surrogate market prices for inputs in countervailing duty investigations but not in the anti-dumping context would undermine the consistency in the WTO's law and policy of counteracting unfair price advantages in trade.

The inconsistent treatment of price distortions under the aforementioned contexts, on the one hand, and under the cost adjustment provisions of the Anti-Dumping Agreement, on the other, obviously reveals a certain WTO ‘regulatory anomaly’. Since the Appellate Body's interpretation of the reasonableness standard under Article 2.2.1.1 was right, this anomaly is attributable to the ill-written (or outdated) rules. In this regard, it is quite remarkable that the panel in EU–Biodiesel compared Article 2.2.1.1 with the second Ad Note to Article VI:1 of the GATT to infer that the former did not textually permit discard of domestic prices as opposed to the latter under which the ‘drafters considered explicit derogations to be needed’ to allow the analogue country methodology.Footnote ¹⁰¹

Therefore, the EU–Biodiesel case highlights the need for reforming the WTO anti-dumping disciplines dealing with NME-like situations. To remove such abnormality in the rules, the Anti-Dumping Agreement should be amended to definitely permit market-oriented adjustments to input costs as well as other components of the normal value affected by market distortions. But this amendment should be implemented together with a legal guarantee of parallel adjustments to be made to the export price to reflect those distortions. Indeed, while it is reasonable to assume that the same cost distortion may be seated in both the normal value and the export price, the methodology at issue is one-sided in authorizing cost adjustment towards the normal value only. In the light of the Appellate Body's judicial economy on Article 2.4, it remains to be seen if that provision demanding a fair price comparison could effectively ensure appropriate adjustability of the distorted cost and other relevant components within both prices in question.Footnote ¹⁰² For greater certainty, however, the existing rules should be reformulated in a way that would explicitly ensure this.

The proposed rule changes would mean that the investigating authority resorting to undistorted surrogate costs for establishing the normal value would be required to appropriately modify the actual export price if it is proved that the latter was also affected by the distortions concerned. This would prevent ‘artificial’ inflations of dumping margins when it is determined that the export price actually would have been higher if it had reflected a corresponding input's market prices. But for this, the investigating authority must examine whether and to what extent the input cost distortions have passed through to both the normal value and the export price of the end product under consideration. By analogy with the pass-through principle developed under the WTO subsidy regime, where the end product is made from an input that was obtained at arm's length – in essence, on market terms that can be assumed to exist, for instance, in transactions between unrelated entities – the investigating authority must establish, rather than just presume, that a non-market reduction in the input cost has flowed through to the end product's home and export prices.Footnote ¹⁰³

By way of illustration, let us consider the example of company x in country A whose chemical product is targeted by an anti-dumping investigation in country B. If A subsidizes the electricity supply to manufacturers, B may prove the pass-through of the electricity subsidy to the chemical production and thus make an upward adjustment to the normal value by adding the subsidy amount to the distorted (reduced) electricity cost. If so, B also has to check if the actual export price should be raised accordingly. But this does not mean that the home and export prices of the product in question will automatically be increased by the same amount, as the extent of the electricity subsidy's pass-through to both prices may differ. For instance, when company x in our scenario sells its product, at arm's length, to domestic company y which then exports it to B, it is conceivable that the price paid by company y for that product absorbs the electricity subsidy at least partially. When the dumping margin is determined for company x (the producer), the extent of the upward cost correction within the normal value mentioned above will be larger than that (if any) within the export price – the price at which company y (the exporter) sells the chemical product to B.Footnote ¹⁰⁴ Moreover, depending on circumstances and the pass-through result, a similar outcome may come up when company x itself exports to B through intermediate country C (the country of export), with the corresponding export price from an exporter in C being compared with the constructed home price in A (the country of origin).Footnote ¹⁰⁵

We can conclude that parallel adjustments to both the normal value and the export price will positively reflect the interests of a country producing below market cost products for export, because they preclude artificial dumping inflations. At the same time, at least two illustrated examples with indirect exports by company x above demonstrate that such adjustments will not diminish the anti-distortion policy of an importing country, as the latter will still be able to offset unjustified State-influenced low costs abroad by making a higher upward revision of the normal value compared to the export price.

Opponents may argue that the WTO subsidy rule book is more suitable to capture input-related State interference. But the subsidy disciplines have their own constraints in addressing certain market distortions similar to the types investigated by the EU authorities above. With respect to the Russia-like factual situation, even if the energy sector (e.g. natural gas, oil or electricity) in the exporting country is proved to be subsidized, countervailing duties cannot, in principle, be levied on downstream products from that country due to typically general availability, and hence non-specificity, of the input (energy) subsidy.Footnote ¹⁰⁶ As for the Argentina/Indonesia-like circumstances, the panel in US–Export Restraints denied the proposition that export restrictions (including public charges on exports) constitute indirect subsidies, namely government's entrustment of, or direction to, a private body to provide subject goods domestically.Footnote ¹⁰⁷ In the light of such anti-subsidy limits, the reformed anti-dumping track would probably be the only efficient way to level the playing field in these and perhaps other factual contexts.

We are mindful that the rule-making process in the WTO is far from being smooth given the consensus-based way of reaching decisions, the large number of stakeholders, outstanding issues on the negotiating table and the current anti-trade sentiment in some parts of the world. But we hope that the introduction of clear-cut rules prescribing parallel cost adjustments on both sides of price comparisons will secure better chance of reaching a compromise within WTO membership on the idea of applying the NME approach to cost distortions in anti-dumping cases.