1. Introduction
The reader might be forgiven for thinking, ‘Oh no, not another AD decision’, or even, ‘Oh no, not another Oil Country Tubular Goods decision’, given how many there have been over the years. But there were a number of distinctive, intriguing features of the (US–OCTG (Korea))Footnote 1 ruling and the behaviour surrounding it. These features speak to important issues such as the proper connection between permitted AD behaviour and sound economic policy; the appropriate level of interaction between politics and dispute settlement proceedings; and even the types of uncertainty introduced when political pressures threaten to overwhelm conventional legal practices.
On its face, the decision simply looked like a resounding defeat for Korea. It had complained about numerous US procedures used to assess AD duties on its industries and it ended up losing all significant legal rulings but one. Further, the decision looked like a defeat for those hoping to push AD policies in a more economically justifiable direction. Often the Panel reasoning focused instead on what was expressly permitted or forbidden by the AD agreement.
The issues surrounding the case are more novel and interesting. There was a rare reversal of a preliminary negative determination. There was extensive political involvement in the United States. The arguments were occurring against a backdrop of diplomatic contention between the United States and Korea, in which the United States had threatened to withdraw from the trade agreement between them.Footnote 2 And there was the rare failure of Korea to appeal the ruling.
Section 2 provides some factual background about the OCTG industry, the GATT and WTO disputes involving it, and its role as the focus of AD efforts. Section 3 details the legal arguments raised in the case and the Panel findings on each. Section 4 describes some of the broader economic concerns raised by the case. Section 5 concludes.
2. Factual Background for OCTG Industry
Oil country tubular goods (OCTG) are steel pipes used for oil and natural gas extraction. OCTG typically covers drill pipe, casing, and tubing whose specifications are stipulated by the American Petroleum Institute. Drill pipe is hollow steel piping that is used on drilling rigs. Casing is usually installed in rigs to prevent water or soil from entering to extraction process. Tubing transports the oil and gas from well to the on ground floor.
Demand for OCTG is driven typically by the levels of activity in oil and gas industries, whereas demand for standard pipe products is driven largely by construction activities. So, OCTG is a specialized and a higher value-added product with complicated specifications, sold at a significant premium over other pipe or tubular products. Thus, exporting governments are often engaged in subsidy policies to promote exportation. In addition, most OCTG exporting countries that have advanced technology in terms of steel manufacturing do not have oil or gas drilling activities in their domestic territories. Due to the lack of domestic markets, antidumping cases involving OCTG typically use constructed values instead of normal values. It explains why OCTG is one of the most frequently targeted products by trade remedy measures that cover a wide range of steel products.Footnote 3 As a consequence, trade measures dealing with OCTG have caused numerous WTO disputes, as shown in Table 1.Footnote 4
WTO Disputes involving OCTG, tube, or pipe
Note: *DS379 involves both AD and CVD complaints. **DS248, 249, 251, 252, 253, 254, 258, 259 are merged in one Panel and Appellate Body proceeding.
As shown in Figure 1, in the US oil and gas industries for 2015–2016, there was a significant decrease in the number of rigs in use due to sudden drops in crude oil prices. Consequently, the dramatically reduced OCTG demand led to antidumping challenges by the domestic producers to mitigate the import competition.
Schematic of a deep well completion
In the underlying case of US–OCTG (Korea), the US Department of Commerce (DOC) imposed AD duties on OCTG imported from not just Korea but also eight other countries, including China, India, Turkey, Vietnam, Ukraine, Taiwan, Philippines, and Thailand.
Oddly enough, the DOC issued a negative preliminary determination only for Korean exporters among the many OCTG exporters, despite Korea exhibiting the largest import share (Figure 2).Footnote 5 But this preliminary result was reversed in the final determination as shown in Table 2. It is, in fact, very rare for the DOC to reverse the negative preliminary determination in the final determination. For 206 cases determined from January 2014 to April 2019, the DOC reversed only three cases, including the current OCTG case.Footnote 6
Rigs count in the us oil and natural gas industries
OCTG imports of US (value in $1,000)
US AD measures concerning OCTG
Notes: * AD rates revised pursuant to the CIT decision. Many of the AD rates were modified later due to ministerial errors and the Court of International Trade proceedings.
** This decision applied the ‘particular market situation’ method that provoked another huge controversy on the legality of the DOC practices.
The negative preliminary determination by the DOC seemed to instigate a very active political lobby by domestic producers despite somewhat weak evidence for dumping allegations.Footnote 7 In the end, the DOC reversed its decision in the final determination, which raised many controversial issues. For example, there was the question of comparing Korean profits across markets. Since Korean producers did not normally sell OCTG in their domestic market, which never had any oil drilling activities, the profits in their home market sales or financial statements did not properly represent ‘the profit to reasonably reflect the merchandise under investigation’. This led the DOC to calculate constructed value (CV) profits. To do so, the DOC considered the profit for Tenaris SA (Tenaris), an Argentinian global producer and seller of OCTG products. But, the Tenaris profit – 26.11% – information was dismissed in the preliminary determination due to a lack of accuracy, since it was based on ‘a research paper prepared by a student at the University of Iowa, School of Management’.Footnote 8 Instead, in the preliminary results, the DOC decided to use home market sales of non-OCTG pipe or the financial statements of the six Korean OCTG producers. The DOC, however, changed its decision in the final determination and adopted the Tenaris financial statements as the best available option for determining CV profit.
Moreover, the DOC limited its AD examination to only two producers, HYSCO and NEXTEEL, while excluding four voluntary respondents. But one of the excluded respondents, ILJIN, was the only supplier of seamless OCTG from Korea, making it distinctly different from the other Korean suppliers that only produced welded OCTG. Such a distinctive market role could presumably have led to lesser AD duties, the reason the producer voluntarily responded. Even US petitioners, in their comments on respondent selection, admitted that ILJIN should be selected as a mandatory respondent.Footnote 9
After the Panel for the WTO dispute settlement procedure was established on 25 March 2015, the US Court of International Trade (CIT) issued the remand order on 2 September 2015.Footnote 10 The CIT directed the DOC to reconsider certain aspects of the CV profit rate calculation as well as the issue of whether the only two selected respondents were representative of the Korean industry since it did not provide sufficient reasoning for declining to select ILJIN as a mandatory respondent.
Pursuant to the CIT decision, on 22 February 2016, the DOC issued the redetermination result that lowered the dumping margins from 15.75% to 6.49% for HYSCO and from 9.89% to 3.98% for NEXTEEL.Footnote 11 But the DOC continued not to select ILJIN as a mandatory respondent. Korea's challenge on these remand decisions by the DOC was rejected by the Panel that ruled the pertinent claims to fall outside its terms of reference.
Although the Panel issued rulings that agreed with Korea regarding only one legal issue, Korea decided not to appeal the Panel ruling.Footnote 12 Also, the Korean government agreed with the US government on the 12-month implementation period – expiring on 12 January 2019 – which seemed an unusually long period to reconsider only the CV profit element.Footnote 13 This arrangement was further modified to extend the expiry date until 12 July 2019.Footnote 14 Such a delay would normally be seen as severely disadvantaging Korea, the complaining country. As a compliance action, the DOC commenced a Section 129 proceeding on 7 November 2018.Footnote 15 On 8 August 2019, the US government referred this case to Article 22.6 arbitration after the Korean government requested the retaliation authorization of $350 million.Footnote 16
3. Key Legal Rulings
While Korea raised about ten legal claims, the Panel ruled in favour of Korea only on CV profit determination. For other claims, the Panel approved a notably broad scope of the investigating authority's discretion. We will focus on two core legal issues: whether the DOC was justified in its decision (i) to use CV profit based on the controversial study result and (ii) to decline the examination of voluntary respondents.
3.1 Viability Test
Korea challenged the DOC decision to use CV instead of accepting third-country export prices. So it contested the legality of the US AD rule that allowed such discretion for the DOC. The ‘viability test’ or ‘minimum quantitative threshold’ set out in 19 USC. §1677b(a)(1)(B)(ii) provides that third-country export sales of the foreign like product can be used for determining normal value if such sales are more than 5% of the quantity or value of sales in the US market.
On the other hand, AD Agreement Article 2.2 allows dumping margin calculations using export prices from an appropriate third country when there are no or low volumes sale of the like product in the ordinary course of trade in the exporting country market. The footnote of Article 2.2 further explains that an export volume shall normally be considered a sufficient quantity for the determination of the normal value if such export constitutes more than 5 per cent of the sales in the importing country.
First, Korea argued that, unlike the ‘viability test’, Article 2.2 does not stipulate an investigating authority to use third-country export sales for the normal value calculation only when such a minimum quantitative threshold is met. Korea claimed that ‘the existence of such a threshold’ itself was an additional requirement not contemplated in the AD Agreement and rendered the ‘viability test’ as such inconsistent with Article 2.2.
The Panel explained that while Article 2.2 permits the authority to have a choice between third-country export prices and the constructed normal value for purposes of normal value determination, Article 2.2 ‘neither expressly limits nor directs’ how to reach that choice. Thus, it ruled that an investigating authority has the complete discretion to choose which method to use based on its own criteria. Moreover, the Panel added that Article 2.2 does not require the authority to explain the basis of the choice between the two methods. As a result, the Panel concluded that the 5% threshold as a criterion for third-country sales in US law is not inconsistent as such with Article 2.2 of the AD Agreement.
Second, Korea claimed that the viability test is inconsistent with Article 2.2 as applied in the OCTG investigation. Based on a rigid application of the viability test, the DOC's questionnaire deprived Korean respondents of even the opportunity to submit third-country sales data or describe their third-country market sales. In other words, the DOC automatically excluded from consideration the Korean respondents’ third-country market sales that did not meet the 5% threshold without any regard to whether the prices of these sales were representative in accordance with Article 2.2.
The Panel ruled that Article 2.2 does not impose any obligation on an investigating authority to examine whether a respondent's third-country export prices are representative if it chose to use constructed normal value. The Panel therefore concluded that the DOC did not act inconsistently with Article 2.2 in applying the viability test and thus rejecting third-country sales for determining normal value.
Therefore, the Panel concluded that the ‘viability test’ adopted by the DOC is not inconsistent with AD Agreement Article 2.2, either ‘as such’, or ‘as applied’.
3.2 Constructed Value Profit Rate Determination
3.2.1 Duty To Use Actual Data under the Article 2.2.2
When the DOC reversed its own decision in the preliminary determination and decided to use the controversial profit information of Tenaris (26.11%) for CV calculation, Korea argued that the DOC should use at least the actual data, as stipulated in Article 2.2.2 of the AD Agreement. Article 2.2.2 requires price construction to be based on actual data pertaining to production and sales in the ordinary course of trade of the like product. If that is not possible, price construction may be based on: (i) the actual amounts incurred and realized by the exporter in respect of production and sales in the domestic market of the country of origin of the same general category of products; (ii) the weighted average of the actual amounts incurred and realized by other exporters subject to investigation in respect of production and sales of the like product in the domestic market of the country of origin; (iii) any other reasonable method provided that the amount for profit so established shall not exceed the profit normally realized by other exporters or producers on sales of products of the same general category in the domestic market of the country of origin.
Korea argued that the DOC violated Article 2.2.2 since it did not use actual data for Korean exporters’ profits on home market as well as third-country sales of the like product. The United States rebutted that neither of the two Korean respondents had a viable home or third-country market during the period of investigation.
The Panel addressed the following three specific issues in a sequence:
1. whether the DOC was permitted to reject the actual data concerning the Korean respondents’ domestic market sales of the like product during the period of investigation, because these sales were made in ‘low volumes’;
2. if not, whether the DOC had actual profit data concerning the respondents’ domestic market sales of the like product; and
3. whether the investigating authority need not consider the data concerning the losses, even if ‘actual’, when an exporter makes a loss rather than a profit on domestic sales of the like product.
First, the Panel clarified the existing jurisprudence that the chapeau of Article 2.2.2 did not allow an investigating authority to decline the actual data concerning the respondents’ domestic market sales simply because these sales were made in low volumes. The Appellate Body in EC–Tube or Pipe Fittings explained that the absence of any qualifying language related to low volumes in Article 2.2.2 implies no exception for low-volume sales.Footnote 17 The Panel in EC–Salmon (Norway) also confirmed that an investigating authority must include data pertaining to low-volume sales when determining the amounts for profit under the chapeau of Article 2.2.2.Footnote 18 Therefore, the Panel concluded that the chapeau of Article 2.2.2 did not permit the DOC to reject the actual data pertaining to the Korean respondents’ domestic market sales.
This decision raised an interesting point for Article 2.2. As the United States argued, it seems puzzling that, while low-volume sales in the domestic market are rejected as a basis for normal value determination in the chapeau of Article 2.2, data derived from the same sales should be accepted for purposes of CV profit determination under Article 2.2.2. The Panel agreed that, in terms of overall coherence of Article 2, this was ‘somewhat perplexing’.Footnote 19 But the Panel explained that the identification of low-volume sales permits an investigating authority to either construct normal value or use third-country export prices as normal value. That is, the identification of low-volume sales operates as a trigger for an investigating authority to adopt an alternative measure for normal value determination. That, however, does not necessarily exclude the components of such a low-volume sale price from that determination. Moreover, Article 2.2.2 requires that only sales ‘in the ordinary course of trade’ be used as a basis for CV profit determination. Accordingly, ‘what is discarded for normal value determination under Article 2.2 is the price of low-volume sales’, but ‘what is accepted for purposes of normal value construction under Article 2.2.2 is the amount for profit and SG&A on those low-volume sales that are in the ordinary course of trade’.Footnote 20 Therefore, it explained that the exclusion of sales that are not in the ordinary course of trade would, ‘at least in part’, address the overall coherence issue.Footnote 21
Second, the Panel noted that neither the DOC's final determination nor its preliminary determination made any reference to the alleged lack of relevant data as a reason for the CV profit determination. To the contrary, the respondents, HYSCO and NEXTEEL, clearly stated that they had made ‘some sales of the foreign like product in the home market during the POI’. Thus, the Panel dismissed as ex post rationalization the US argument that the DOC could not use the preferred method to determine CV profit because the record did not contain any data pertaining to sales of the like product in the home market.
Lastly, the Panel ruled that nothing in the DOC's determinations indicates its decision to reject the Korean respondents’ actual data due to the loss during the POI. The United States has not identified where, in its determination, the DOC relied on this conclusion as a reason for not using the preferred method to calculate CV profit. Therefore, the Panel rejected this justification offered by the United States as ex post rationalization as well.
3.2.1 ‘Same General Category of Products’ under Articles 2.2.2(i) and 2.2.2(iii)
For a profit determination under Article 2.2.2(i) or to calculate a profit cap under Article 2.2.2(iii), an investigating authority must determine the scope of the ‘same general category of products’. Korea argued that the DOC applied the term ‘same general category of products’ in Article 2.2.2 too narrowly to consider non-OCTG products, such as line pipe and standard pipe, as falling within the ‘same general category’ as OCTG. It claimed that, due to this excessively narrow interpretation, the DOC rejected Articles 2.2.2(i) and 2.2.2(iii) as a basis to calculate CV profit. Korea claimed that a like product, based on the DOC's definition of the foreign like product, could be excluded from the same general category of products.
While the Panel noted that there is no precise definition of the term ‘same general category of products’ in Article 2.2.2(i) or (iii), it explained that ‘the scope of the same general category of products must be understood to be broader, not narrower than that of the like product’.Footnote 22 Therefore, if the like product is not limited to pipe products used for down hole applications, the same general category of products, which is broader than the like product, cannot be limited to pipe products used for down hole applications. In other words, the same general category of products cannot exclude pipe products that do not exhibit the same fundamental characteristics for down hole applications and are not used for down hole applications. Yet, the DOC excluded such products from the same general category of products in the underlying investigation. The Panel therefore concluded that the DOC defined the same general category of products more narrowly than it defined the like product by excluding those pipe products not used for down hole applications that fell within the definition of the like product. Thus, the DOC had no proper basis for its conclusions that the methods under Article 2.2.2(i) could not be used, and that the profit cap called for in Article 2.2.2(iii) could not be calculated.
Accordingly, the Panel ruled that the DOC's decision to adopt the CV profit based on Tenaris information is inconsistent with Article 2.2.2. This is the only core legal issue for which the Panel found in favour of Korea.
3.3 ‘Association’ and Construction of Export Price
Korea argued that the DOC's calculation of constructed export prices was inconsistent with Article 2.3 due to the wrongful association determination between NEXTEEL and POSCO. Korea further argued that even where association existed, the DOC had to determine whether the export price was actually unreliable because of such association, but the DOC failed to make such a determination.
Pursuant to Article 2.3, when an investigating authority considers that the export price is unreliable because of association between the exporter and the importer, the export price may be constructed on the basis of the price at which the imported products are first resold to an independent buyer, or if the products are not resold to an independent buyer, or not resold in the condition as imported, on such reasonable basis as the authority may determine.
While the Panel clarified that the appearance of unreliability must be because of association, it explained that the text of Article 2.3 does not require any determination, let alone a determination as to the reliability of the export price. On the other hand, it also explained that Article 2.3 does not allow construction of export price simply because of association. An investigating authority should not ignore evidence suggesting that the export price may be reliable notwithstanding association. Moreover, ‘association’ may be shown by ‘formal legal ties or far less structured and non-binding relationships’.Footnote 23 It ruled that ‘association’ exists ‘where an exporter and the importer or a third party do not act independently of one another’.Footnote 24
The DOC found ‘association’ on the basis of the intermediate factual findings, including:
• POSCO supplied NEXTEEL with ‘virtually all’ of the steel coil, the main input used in the production of OCTG, used by NEXTEEL.
• POSCO had a history of working closely on-site with NEXTEEL departments and providing marketing assistance and other promotional activities for the benefit of NEXTEEL.
• POSCO and NEXTEEL shared technology and market information pertaining to OCTG.
Based on these findings, the DOC concluded that the combination of POSCO's involvement on both the production and sales side creates a unique situation where POSCO is operationally in a position to exercise restraint or direction over NEXTEEL in a manner that affects the pricing, production, and sale of OCTG.
Korea rebutted that the DOC ignored alternative explanations of the evidence as well as other evidence which undermined these findings. For example, the DOC failed to consider that NEXTEEL had alternative sources other than POSCO from which it purchased steel coils. It also argued that the DOC failed to consider evidence regarding marketing and technology collaboration between NEXTEEL and POSCO that undermined its findings
Regarding the determination of ‘association’ between NEXTEEL and POSCO, the Panel found that Korea failed to demonstrate that the evidence relied on by the DOC did not support its intermediate factual findings. The DOC's intermediate findings concerning NEXTEEL's purchases and consumption of steel coils, marketing, and technology collaboration between POSCO and NEXTEEL, and sales in the US market through ‘associated’ companies were found to be sufficient to demonstrate ‘association’ between the exporter and the importer or a third party within the meaning of Article 2.3.
3.4 Disclosure of Essential Information and Communication
The DOC only revealed in its final determination that it had accepted the Tenaris profit data for purposes of CV profit calculation. The DOC was similarly coy about political arguments that had been made for rethinking the preliminary determination. On 17 June 2014, one day before the deadline to submit case briefs, the DOC posted to the record a letter signed by 57 US Senators dated 15 May 2014. On 23 June 2014, the same day as the deadline to submit rebuttal briefs, the DOC posted to the record a letter signed by 155 Members of the US House of Representatives dated 10 June 2014. These letters, as well as other letters from US lawmakers, local government leaders, and industry representatives and memoranda to the file describing phone calls and meetings with US lawmakers and industry representatives, revealed concerns regarding the DOC's preliminary determination.Footnote 25
Korea argued that the DOC did not afford Korean respondents an opportunity to present evidence in defence of their interests under Article 6.2 because they had no notification from the DOC that the Tenaris financial statements submitted by US Steel were properly on the record. In addition, the DOC failed to provide the Korean respondents with any opportunity to prepare presentations regarding such data, inconsistent with Article 6.4. Finally, the DOC's decision to accept and rely on the Tenaris financial statements constitutes an ‘essential fact’ which was not disclosed in sufficient time before the final determination, inconsistent with Article 6.9.
The Panel recognized that, ‘while Article 6.2 imposes a general duty on an investigating authority to ensure that interested parties have a full opportunity throughout an anti-dumping investigation for the defence of their interests, it does not give specific guidance on the type of procedural steps an investigating authority should take in ensuring the rights of interested parties’. The Panel was not persuaded that because USDOC did not disclose its acceptance of the Tenaris financial statements on the record until the final determination, the Korean respondents were prevented from launching a ‘full-scale argument’.
The Panel also explained that Article 6.4 requires an investigating authority to provide interested parties timely opportunities to see all non-confidential information that is relevant to the presentation of their cases and that is used by that authority in the anti-dumping investigation, and to prepare presentations on the basis of this information. The Panel, however, ruled that nothing in Article 6.4 suggest an investigating authority to ‘inform’ interested parties of procedural decisions to accept or use certain information in the anti-dumping investigation.
Concerning Article 6.9, the Panel observed that this provision does not require the investigating authority to disclose its decisions or conclusions, but rather requires ‘disclosure of essential facts “under consideration”, which form the basis for the decision whether to apply definitive measures’.Footnote 26 Since the essential facts which must be disclosed are ‘under consideration’ at the time of their disclosure, it implies that the investigating authority has not yet reached conclusions regarding its reliance on them at that stage. On the other hand, Article 6.9 requires an investigating authority to disclose only the essential facts but not its reasoning or its conclusions.
Given the above reasoning, the Panel ruled that the DOC's reliance on the Tenaris profit data to calculate profit constitutes its ‘conclusion’ to use that data in calculating CV profit. Thus, it concluded that such reliance was not an essential fact within the meaning of Article 6.9, and the DOC was not required to disclose it. It elaborated that the acceptance of the Tenaris profit data on the record would not mean the DOC to necessarily consider that data as the basis for the CV profit determination. Acceptance on the record by an authority did not necessarily mean that the contents of that submission were considered ‘essential facts under consideration’. Even if the DOC had disclosed its acceptance of the Tenaris profit data on the record, that disclosure in itself would not have identified the Tenaris data as an ‘essential fact’.
Concerning the Korea's allegations in connection with the letters and other communications, the Panel ruled that they fell outside the terms of reference since they were entirely new claims, based on different facts and legal arguments. On these bases, the Panel declined all the Korean claims concerning Article 6 of the AD Agreement.
3.5 Limited Examination of Respondents
Korea claimed that the DOC acted inconsistently with Article 6.10 by limiting its examination to only two exporters, NEXTEEL and HYSCO, and by failing to individually examine the voluntary respondents, particularly Husteel, SeAH, and ILJIN. The DOC rebutted that there were a large number of exporters or producers involved in the underlying investigation. Moreover, the DOC maintained that it lacked the resources to examine all such exporters or producers, especially considering numerous concurrent anti-dumping proceedings conducted by the relevant office.
It should be noted that the CIT remanded the DOC decision not to include ILJIN since welded OCTGs produced by NEXTEEL and HYSCO were not ‘representative’ of seamless OCTGs.Footnote 27
Article 6.10 provides the authorities ‘shall, as a rule, determine an individual margin of dumping for each known exporter or producer concerned of the product under investigation’. Only in cases where the number of exporters or producers is so large as to make such a determination impracticable, may they limit their examination to a reasonable number of interested parties. In addition, Article 6.10.2 further provides, in cases where the authorities limit their examination, that they shall nevertheless determine an individual margin of dumping for any exporter or producer not initially selected and < A.who submits the necessary information in time for that information to be considered during the course of the investigation.
The Panel explained that institutional resource constraints are equally valid as a basis for a limited examination under Article 6.10 as they are for not examining voluntary respondents under Article 6.10.2. Moreover, it ruled that an investigating authority has no obligation to provide a separate explanation of why it is practicable to examine only the selected number of exporters. The DOC's conclusion of undue burden to individually examine the voluntary respondents was based on complexities unique to the underlying investigation, time constraints, and resource constraints faced by the USDOC. The Panel ruled that the DOC's explanation of undue burden to individually examine the voluntary respondents was reasoned and adequate, and sufficient to satisfy its obligations under Article 6.10.2. Therefore, the Panel declined Korea's claim that DOC acted inconsistently with Article 6.10.2.
In this ruling, the Panel seemed to miss the key issue of the CIT remand decisions. It raised an interesting puzzle: why the complainant did not appeal the Panel ruling. In fact, the complainant decided not to bring an appeal due to the problem in the Appellate Body procedure that has caused a significant delay.Footnote 28 The business communities wanted a quicker resolution of the situation, instead of being too much delay due to the handicapped Appellate Body procedure. They thought that winning for CV profit issues was practically enough to rectify the situation. But their expectation turned out to be wrong when the DOC introduced the ‘particular market situation’ method to raise the AD rates in the administrative review.
4. Economic Analysis
The economic shortcomings of AD regimes have been described at length. We therefore focus our analysis on the broader political-economic context in which the regime exists, rather than delving into the implications of particular measures supported by the Panel. This seems particularly relevant given the unusual intrusions of politics into the case.
4.1 Economic Considerations at Play
We discuss three big issues:
1. We offer two distinct ways of thinking about anti-dumping. The choice ultimately determines whether one approaches this case in a narrow fashion – as offering new guidance on how countries should pursue anti-dumping cases – or in a broader political context.
2. We argue that there is at least compelling circumstantial evidence that politics played a significant role in bringing about a highly unusual outcome.
3. The intrusion of politics into what was supposed to be an apolitical legal proceeding raises serious questions about implications for the Dispute Settlement Mechanism. We consider some of them.
4.2 How Should We Think about the Less than Fair Value Determination?
At one level, the US–OCTG (Korea) decision is the latest in a long line of jurisprudence that has helped to distinguish permissible from impermissible approaches to anti-dumping policy. At the core of the discussion has been the question of how countries may appropriately estimate less-than-fair-value (LTFV) margins. As those margins, when combined with affirmative injury determinations, lead directly to levels of protection, the debate can also be interpreted as asking what levels of protection are permissible under WTO rules.
In the US–OCTG (Korea) decision, the debate over permissible approaches revolves around issues such as the determination of constructed value (CV) profits. The decision to include or exclude certain companies’ profits, as noted above, had a significant effect on the AD duties that the DOC ultimately put forward.
If one abstracts from the specific questions at issue here – of which all but one were decided in favour of the United States – the tone is exceedingly familiar. It brings to mind the string of decisions on zeroing, for example. In such disputes, narrow interpretations of permissible behaviour lead to smaller (or no) margins and duties, while permissive interpretations lead to larger margins and duties.
On what basis can a Panel decide whether the narrow or permissive interpretation is more appropriate? Of course, if there is plain language in the AD agreement that explicitly addresses the question at hand, then the decision is easy. But what of the more common case where it is necessary to infer meaning from language that seems related to the question at hand, but vague? Or what principles should guide if there are clauses that appear in conflict?
This question has been central to a critique of the WTO DSM that now seems to threaten the system itself. In the context of a different DSM decision – a softwood lumber case that approved of a US zeroing approach – USTR Robert Lighthizer made clear his sentiments about the entire line of jurisprudence:
The WTO rules do not prohibit ‘zeroing’…The United States never agreed to any such rule in the WTO negotiations, and never would. WTO Appellate Body reports to the contrary are wrong, and reflect overreaching by that body. The United States commends this Panel for … having the courage to stand up to the undue pressure that the Appellate Body has been putting on Panels for many years.Footnote 29
This official commentary on a separate case is invoked only to help justify the following meta-approach that we offer as a way to think about these cases in general. We suggest that there are two distinct ways of thinking about the proper approach to LTFV determination:
1. The ‘Best Estimator’ approach (BE);
2. The ‘Bound Protection’ approach (BP).
The BE approach should be familiar to any student of econometrics. When trying to estimate a variable Y with an estimator 
$\hat{Y}$, we can consider various properties of the estimator, such as whether it is biased, or the narrowness of the confidence interval it will generate. In this vein, we can think of Y as the LTFV margin. We imagine there is a ‘true’ value of the LTFV margin that corresponds to the intent of the AD agreement. If the text of the agreement were sufficiently detailed – a complete contract – we could simply follow the directions for each configuration of facts. For example, what does one do if a key profit estimation for a constructed value stems only from a University of Iowa student paper? Does one include it or not?
Of course, the AD agreement represents an incomplete contract and has nothing like that sort of detail. So instead we look for clues as to which 
$\hat{Y}$is likely to deviate least from the true Y that the parties to the agreement would have specified, if only they had delved deeply enough. As guides to discovering this BE, we look at related language in the agreement and subsequent interpretations of that language. These are all attempts to minimize error in our search for the true value of Y.
The BP approach, in contrast, does not presume that there exists a true value of Y. This, in turn, pretty much kills the idea of a BE 
$\hat{Y}$. If the AD agreement did not offer specifics on particular estimation methodologies, it was not because the contracting parties were being coy, or did not have time to spell out all their thinking on the matter. Instead, it could be that there was no agreement on such matters. Instead, the specified elements of the agreement were meant to produce a general correspondence between a set of observables in instance i {xi 1, xi 2,…, xin} and 
$\bar{Y}_i$, where 
$\bar{Y}_i$ represents the highest permissible level of protection that corresponds to the instance (which generally describes industry conditions at a given time). This corresponds to the much more straightforward idea of a tariff binding that one conventionally finds in market access agreements.
These two approaches can lead to tensions when they coexist. Someone subscribing to the first approach could ‘interpret’ the AD agreement to require use of available data, for example, because that use will lead to a more precise estimate. Someone subscribing to the second approach could object, saying that if the AD agreement did not explicitly require the use of such data, then this new additional restriction unjustifiably limits the extent of protection the United States was allowed under the AD agreement.
In the BP approach, for example, there is a general sense that a particular history of performance (prices, capacity utilization, profits, imports) in one part of the US steel industry should allow a certain level of tariff protection, 
$\bar{Y}_{}$. The DOC is free to adjust its procedures to get the highest Y ′ it can construct, so long as 
${Y}^{\prime} \le \bar{Y}$.
In the BE approach, the DOC can only adopt sanctioned estimation procedures, of the sort that presumably lead to tight, unbiased estimates of the true Y. It matters far less what the ultimate level of protection might be, so long as the procedures were deemed legitimate.
One obvious appeal of the BE approach over the BP approach is that it has the consistency and predictability that we expect in legal procedures. As interpretations accumulate over time, the range of acceptable estimators diminishes and there is a tighter correspondence between observed facts and the ultimate LTFV estimate. In contrast, under the BP approach, there is the fundamental difficulty that the underlying correspondence between observables and bound protection is, at best, implicit in the agreement. Without textual guidance, the limits of permissibility may shift over time and may reflect the relative power of the two parties to the dispute.
This description of the two approaches is certainly not offered as a normative endorsement of the BP approach. There are strong economic and legal reasons to prefer a more predictable approach that is relatively immune to external influence. Instead, the BP approach is presented as a means of interpretation for a stance that has come to play an important role and may serve as a better positive description.
When applied to the US–OCTG (Korea) case at hand, the BP idea may help explain some of the particular oddities of the US approach, in which there was a negative preliminary determination, then an active lobbying campaign, then the reversal of the decision in the final determination.
4.3 The Behavioural Oddities Surrounding this Case and Potential Interpretations
An unconventional interpretation of a case is only merited if unconventional behaviour was observed. In the case of US–OCTG (Korea), it was. Two unusual developments stand out. First, there was the rarity of seeing a negative LTFV preliminary determination followed by a positive final determination. As noted above, this was observed in less than 1% of cases preceding the current one. Second, there was the oddity of a Panel decision in which the Korean government lost on all but one major point, yet the government declined to appeal. For each oddity, there seems to be at least circumstantial evidence of political influence at play.
In the wake of the preliminary LTFV determination, there was a notable lobbying campaign.Footnote 30 As far as the lack of appeal is concerned, it is worth noting the broader political context surrounding the US–Korean relationship. That relationship had already seen political pressure trump legal guarantees when the United States threatened withdrawal from the KORUS free trade agreement unless the Republic of Korea agreed to renegotiate it. This renegotiation process commenced in July 2017 and was signed in September 2018. It thus straddled the November 2017 release of the Panel report in US–OCTG (Korea).
None of this is to suggest that the Panel itself was subject to political influence, only that the unusual behaviour surrounding the case and the interpretation of its significance are hard to understand without the broader political context.
4.4 The Implications of a More Political Dispute Settlement Mechanism
Perhaps the central virtue of any dispute settlement mechanism is its objectivity. It is so fundamental that its importance usually requires no discussion. The attempt at dispassionate justice is the entire reason that so much effort is spent basing decisions on close textual analysis or established precedent.
It would be naïve to pretend, however, that the global trading system has been immune from political pressures. In the early days of the GATT, the European Coal and Steel Community hardly seemed to meet the requirements of Article XXIV, yet the political demands of the day – and the strong sentiments of Contracting Parties – dictated its acceptance. Also, in the 1950s, when the US Congress applied import quotas on agricultural products, it received a waiver for the measures, which otherwise seemed to contravene GATT prohibitions.Footnote 31 Then there were sensitive political issues that contracting parties decided were best addressed outside of the dispute settlement mechanism, such as issues surrounding large passenger aircraft, at least for a time. The ability to adapt to political exigencies helped the system survive.
But there is a qualitative difference between crafting a system so as to accommodate political concerns and seeing influence exerted in the midst of a dispute. It is the difference between ex ante and ex post reinterpretation of the rules.
To place such a shift in context, there is one further contrast to be drawn, between ex post reinterpretation and ex post renegotiation. The latter has been a longstanding de facto feature of the dispute settlement system. It occurs whenever retaliation is invoked in a dispute – the reciprocal withdrawal of concessions effectively functions as a renegotiation.
The granting of the agriculture waiver, as an example, represented an agreement between countries about what the rules should be. It was worked out over a period of years.Footnote 32 Instances of retaliation as a dispute outcome are one known, potential outcome of an agreement.
The exertion of influence in the midst of a dispute is a more serious matter. It is difficult to predict and calls into question the meaning of agreements themselves. Herein lie the two principle costs, should political influence become an important determinant of dispute outcomes. They are tightly related. The first cost is the loss of predictability in the system. The second cost is the loss of faith in the system.
Predictability is important to economic actors, who must make investment and resource allocation decisions based on expectations. There is recent empirical work explicitly demonstrating the cost of uncertainty in the trading system.Footnote 33 To describe the effects accurately, though, this may be a useful moment, however, to recall the classic distinction between risk and uncertainty.Footnote 34 In the formulation of Frank Knight, risk occurred when outcomes are uncertain, but the uncertainty is quantifiable. In the context of dispute settlement, this might be, for example, when there is an 80% chance the complaining country will win a dispute, and then, upon the complaining country winning, a 50–50 chance that the losing country chooses to maintain its practices and accept retaliation. In such a case, companies and consumers do not know what outcome will prevail, but they can place odds on the possible outcomes. The alternative, in Knight's approach, is uncertainty. This obtains when it is not possible to assign probabilities to the possible outcomes. By moving into a world of aberrant, politically driven interference with dispute settlement cases, potential outcomes become much more difficult to quantify and it becomes much more difficult for economic agents to plan. This undermines a major benefit of the rules-based trading system.
The second, closely correlated cost could be a loss of faith in the trading system. In the absence of a rules-based trading system, the presumed default system is a ‘law of the jungle’, in which countries exercise what political power and threats they can. This need not mean a reversion to Nash or prohibitive tariffs; threats can work to sustain lower tariff levels.Footnote 35 But it would likely mean discarding some of the key features of the WTO system, such as MFN barriers and the additional conformity to rules that comes through a general desire to be in good standing within the system. This scenario – trade's version of an apocalyptic outcome – becomes significantly more likely if dispute settlement outcomes are seen as subject to political influence. If that is the case, then the practical distinction between a rules-based system and ‘law of the jungle’ diminishes. That means, in turn, that there is a diminishing incentive to work to support the rules-based system.
To conclude the point, it is worth noting that political influence on dispute settlement outcomes is neither binary, nor novel, nor evidence of corruption. Countries can agree to settle on certain issues while conforming to prior understandings – or accepting Panel interpretations – on others. There have certainly been past instances in which the threat of international relations repercussions affected countries’ behaviour in trade disputes. And the influence we are describing concerns moves such as curious failures to challenge Panel rulings, rather than pressure on PanelPanellists or, in other cases, appellate body members.
Nonetheless, the behaviour in US–OCTG does seem to herald a concerning move away from the predictable application of trade rules.
5. Conclusion
As a straight legal case, US–OCTG marked a broad victory for the United States over the Republic of Korea. Though Korea had complained about a broad range of US behaviours in the way it calculated its antidumping duties, the Panel found in Korea's favour only on the issue of objectionable US constructed value calculations. In the other areas, the Panel seemed to follow the general reasoning that antidumping practices could be economically dubious but still permissible, so long as they were not explicitly prohibited by the Antidumping Agreement.
The more remarkable elements of the case were to be found in unusual the behaviour surrounding it. There was the rare instance of a negative preliminary determination being reversed by the US Department of Commerce in its final decision. There were blatant political interventions. There was the oddity of Korea declining to appeal such an adverse Panel report and granting the United States a long implementation period.
The evidence of politics infiltrating the dispute settlement process is entirely circumstantial, but the dispute came amidst a time of tense diplomatic dealings between the United States and Korea, particularly with the threatened elimination and subsequent renegotiation of the trade deal between them (KORUS). While the GATT and WTO systems have never been immune from international politics, this case seems to denote a heightened level of susceptibility to pressure. To some extent, the global trading system's ability to accommodate shifting political pressures is one of its strengths. Beyond a certain point, however, the introduction of political pressures into dispute settlement can introduce a new and damaging type of uncertainty into the system. The US–OCTG case can serve to illustrate this threat, as it ended up effectively validating AD behaviours that might have seemed open to challenge.
