1. Introduction
United States – Anti-Dumping Measures on Polyethylene Retail Carrier Bags from Thailand Footnote 1 is yet another in a long line of WTO disputes involving the United States' practice of zeroing when calculating anti-dumping margins.Footnote 2 It is surely one of the least complicated cases ever heard by a WTO Panel. To begin with, only a single violation (model zeroing) was alleged. Moreover, the claim involved an issue that was the subject of numerous prior WTO Appellate Body (AB) and Panel decisions. In its prior decisions, the WTO AB unfailingly had found model zeroing to be inconsistent with the WTO Anti-Dumping Agreement (ADA).Footnote 3 There was every reason to believe it would do so again in this dispute. The Panel's task was further simplified because the respondent did not even contest the claim; in fact, the United States (US) acknowledged that it had used zeroing when calculating dumping margins in this case.
As if the above did not make the dispute's outcome a foregone conclusion, prior to the Panel convening, the parties reached an agreement wherein they agreed: (i) that each would make one written submission only, (ii) that they would forgo Panel meetings; (iii) that the US would not contest Thailand's claim; (iv) that Thailand should not ask the Panel to suggest ways for the US to implement the Panel's recommendations; and (v) that the US should implement the Panel's recommendations using specified provisions of US law.Footnote 4
Given the set of facts above, the intrigue in this dispute is less about the Panel's decision (or the legal basis for its decision) and more about why the parties were unable to resolve the dispute via consultations. Ostensibly, the polyethylene (PET) bags dispute was about zeroing. In fact, at the time of the dispute, the US had already changed its regulations to abolish zeroing in original investigations. The raison d'être for this case was the specific way in which the US had changed its domestic law to eliminate zeroing. In December 2006 – almost two years before Thailand requested consultations – the US ceased zeroing in original investigations. Although the US stopped zeroing in original investigations, it did so only prospectively; that is, the US's new rules apply only to margin calculations in future cases. All existing margins, including the margins in this dispute, were unaffected. Therefore, Thailand was forced to bring the dispute to the WTO because the US was unwilling to apply the WTO rulings on zeroing to a pre-existing case without an adverse WTO decision specifically for that case. And, unlike those of some other countries, which automatically sunset anti-dumping duties after five years (unless a properly documented sunset-review request is received), the US sunset rules could result in the zeroing-distorted margins remaining in effect for many, many years. By way of comparison, in the aftermath of the EU–Bed Linen zeroing disputeFootnote 5, which also involved model zeroing, the EU offered interested parties affected by zeroing in other EU cases the opportunity to request a review. It also refrained from applying model zeroing after the adoption of the AB Report (although it continues to use the targeted dumping exception in some cases).Footnote 6
The sequence of events in this case highlights the basis for the dispute (Table 1). The US initiated an anti-dumping (AD) investigation on imports of PET bags from Thailand on 16 July 2003. The US Department of Commerce (USDOC) made its final determination with respect to the dumping margin on 18 June 2004, as amended on 15 July 2004, and the final AD order was imposed on 9 August 2004.
Table 1. Key dates in PET bags dispute
Then, in response to numerous WTO AB rulings that zeroing was inconsistent with the ADA, the USDOC announced on 27 December 2006 that it would no longer use zeroing in original investigations.Footnote 7 Given the prospective nature of the US policy change, however, the margins in the PET bags case remained unchanged because they had entered into force in 2004 and therefore predated the changed policy.
Thailand challenged the US's continued use of zeroing. On 26 November 2008, Thailand requested consultations with the US. Consultations were held on 28 January 2009 but failed to resolve the dispute. Thailand requested the establishment of a Panel on 9 March 2009. The interim Panel Report was circulated on 11 December 2009. With the Panel Report in hand, the US and Thailand announced an ‘amicable and expeditious resolution of the dispute’ on 10 January 2010; the final Panel Report was circulated on 22 January 2010.
Given that (i) the dispute covers extremely well-trodden ground,Footnote 8 (ii) neither party challenged the Panel to push the envelope on zeroing, and (iii) the Panel's reasoning simply restated previous AB and Panel opinions, we find ourselves with little to say about zeroing beyond what we have previously stated.Footnote 9 Instead, we will comment on what we consider four ancillary aspects of the case. First, what are the implications of pre-dispute agreements for the WTO dispute system? Second, how should the Panels handle undisputed claims? Third, what does the US's prospective implementation mean for the WTO dispute system? Fourth, what are the consequences of the US implementation with respect to original investigations for the unresolved issue of zeroing in administrative reviews?
In the next section, we review the claims and summarize the Panel's findings. We then offer commentary on the ancillary issues associated with this dispute. We conclude with a discussion of how the US implemented the Panel's recommendations.
2. Claims and Panel findings
Zeroing refers to the practice of replacing the actual amounts of dumping that yield negative dumping margins with a value of zero prior to the final calculation of a weighted average margin of dumping for the product under investigation with respect to the exporters under investigation. Zeroing drops transactions that have negative margins from the numerator and hence increases the overall dumping margins and the resulting size of the applied anti-dumping duty.Footnote 10 While there are several forms of zeroing,Footnote 11 the present case concerned the relatively simple form of model zeroing under the weighted-average-to-weighted-average comparison method in an original investigation.
Thailand claimed that the USDOC had used the model zeroing methodology to determine the final dumping margins for individually investigated Thai exporters whose margins of dumping were not based on total facts available, in violation of Article 2.4.2 of the ADA.Footnote 12 In particular, Thailand claimed that the USDOC (i) identified different ‘models’, that is types of products based on the most relevant product characteristics; (ii) calculated weighted average prices in the United States and weighted average normal values in the comparison market on a model-specific basis for the entire period of investigation; (iii) compared the weighted average normal value of each model to the weighted average US price for that same model; (iv) calculated the dumping margin for an exporter by summing up the amount of dumping for each model and then dividing it by the aggregated US price for all models; and (v) set to zero all negative margins on individual models before summing the total amount of dumping for all models.Footnote 13
As mentioned above, following the establishment of the Panel, the parties submitted to the Panel a ‘joint procedural agreement’ wherein the parties agreed to streamlined proceedings and, as a result, the Panel decided not to hold any substantive meetings with the parties.
The US did not contest Thailand's claims. It acknowledged the accuracy of Thailand's description of the USDOC's use of zeroing and recognized that in US–Softwood Lumber V,Footnote 14 the AB found that the use of zeroing with respect to the average-to-average comparison methodology in investigations was inconsistent with Article 2.4.2, by interpreting the terms ‘margins of dumping’ and ‘all comparable export transactions’ in an integrated manner. The US also acknowledged that that reasoning was equally applicable in the present case.Footnote 15
The Panel considered that the issues raised in the present case were very similar to those addressed by the panels in US–Shrimp (Ecuador) and US–Shrimp (Thailand),Footnote 16 and agreed with the approach adopted by these panels.Footnote 17
The Panel then first considered its role under Article 11 of the DSU, in light of the fact that the US did not contest Thailand's claim. Article 11 of the DSU describes the role of panels in DSB proceedings. Notwithstanding the US's decision not to contest Thailand's claim, the Panel explained that it was bound by Article 11 of the DSU to make an ‘objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements, and make such other findings as will assist the DSB in making the recommendations or in giving the rulings provided for in the covered agreements’.Footnote 18
The Panel consequently adopted the approach followed by the panel in US–Shrimp (Ecuador).Footnote 19 The Panel thus reasoned that, in the absence of effective refutation by the defending party, it is still required for the complaining party to present a prima facie case. The Panel could thus rule in favor of Thailand only if it was satisfied that Thailand had made a prima facie case regarding its zeroing claim.Footnote 20
In order to establish that the USDOC zeroed in the present case, Thailand referred to a computer program used by the USDOC to calculate dumping margins. The Panel found in that respect that the computer program did indeed indicate the use of zeroing by the USDOC. The Panel furthermore repeated that the US had acknowledged the accuracy of Thailand's description of the USDOC's use of zeroing. The Panel therefore concluded that it was satisfied that Thailand had demonstrated that the USDOC zeroed in the measure at issue.Footnote 21
The Panel then verified whether Thailand had established that the methodology used by the USDOC was indeed the same in all legally relevant aspects as the methodology reviewed by the Appellate Body in US–Softwood Lumber V. Thailand relied on the description of the methodology used in the USDOC's notice of preliminary determination of sales at less than fair value, as well as the computer program used to determine the dumping margins. According to the Panel, this was sufficient evidence to establish that the USDOC performed the actions as claimed by Thailand in points (i) through (v) above.Footnote 22
The Panel considered that the absence of any denial by the US and the above-mentioned evidence established that Thailand had demonstrated that the methodology applied by the USDOC in calculating the margins of dumping that were not based on total facts available was the same as the methodology that the Appellate Body in US–Softwood Lumber V had found to be inconsistent with Article 2.4.2 of the ADA.Footnote 23
Finally, the Panel verified whether Thailand had established that the methodology applied by the USDOC was inconsistent with Article 2.4.2 of the ADA. The Panel considered in that respect that, while it was not bound by the reasoning in prior Panel or AB reports, adopted reports do consider legitimate expectations among WTO members, and found that following the AB conclusions in earlier reports would not only be appropriate but would also be what would be expected, especially where the issues are the same.
The Panel referred to the Panel Report in US–Shrimp (Ecuador) with regard to the AB's reasoning in US–Softwood Lumber V concerning the compatibility of zeroing with Article 2.4.2 of the ADA.Footnote 24 The AB in US–Softwood Lumber V had found that it was clear from the texts of Article VI:1 of the GATT 1994 and Article 2.1 of the ADA that dumping is assessed in relation to the product as a whole as defined by an investigating authority; it cannot be found to exist only for a type, model, or category of that product. The AB in US–Softwood Lumber V on that basis concluded that the treatment of comparisons for which the weighted average normal value is less than the weighted average export price as ‘non-dumped’ comparisons was not in accordance with the requirements of Article 2.4.2 of the ADA.
The Panel also noted that the panel in US–Shrimp (Ecuador) found that there was a consistent line of AB reports holding that zeroing in the context of the weighted average-to-average methodology in original investigations was inconsistent with Article 2.4.2 of the Anti-dumping Agreement.Footnote 25
The Panel therefore concluded that, given that the issues raised in the present case were identical in all material respects to those addressed by the AB in US–Softwood Lumber V, Thailand had established a prima facie case that the use of zeroing by the USDOC was inconsistent with Article 2.4.2 of the ADA, because the USDOC did not calculate those dumping margins on the basis of the product as a whole, taking into account all comparable export transactions.Footnote 26
3. Analysis of Panel's decision
Bown and Prusa (2010) document that, between the first zeroing dispute of 1998 and early 2010, nearly 20 WTO disputes have involved zeroing. Consequently, there is not only a plethora of WTO AB rulings on the issue but also a substantial body of legal and economic analyses of zeroing, including Janow and Staiger (2003), Grossman and Sykes (2006), Bown and Sykes (2008), Prusa and Vermulst (2009, 2011), Crowley and Howse (2010), and Hoekman and Wauters (2011).
The key phrase is contained in Article 2.4.2 of the ADA:
the existence of margins of dumping during the investigation phase shall normally be established on the basis of a comparison of a weighted average normal value with a weighted average of prices of all comparable export transactions or by a comparison of normal value and export prices on a transaction-to-transaction basis.Footnote 27
The case does not add much to the existing WTO rulings on zeroing. In Prusa and Vermulst (2009) we stated that:
the Appellate Body was arguably correct in prohibiting the use of zeroing under the main methods of Article 2.4.2 of the ADA as well as in various reviews.Footnote 28
and in Prusa and Vermulst (2011) we wrote:
We agree with the legal justification logic behind the AB's decisions – given that this case largely follows the script developed in prior cases, this should not be surprising. Further, as we have argued in the past, we also believe that standard economic statistical inference methods demand that all transactions be included in a ‘fair’ comparison.Footnote 29
Instead of belaboring our views on zeroing, we instead discuss certain related issues in this case.
3.1 Pre-dispute agreements and the WTO dispute system
Once the Panel was established, the US and Thailand submitted a ‘joint procedural agreement’. Three aspects of the agreement merit comment.
First, there is the general issue of whether a Panel should be bound by such agreements. Within the WTO context, it appears relatively well established that the parties to a dispute basically determine the confines of the dispute and that those confines should be respected by the Panel. This is perhaps most clear in the principle of non ultra petita, Footnote 30 as a result of which a Panel should never rule beyond that which has been requested by the complainant in a WTO dispute. The storm of protest in the DSB in the wake of the 21.5 Panel Report in Australia–Automotive Leather provides a good illustration of the strong feelings the violation of this principle arouses in many WTO members. While such limitations may not necessarily be good for WTO dispute settlement from a systemic point of view, they appear to be a necessary consequence of the WTO being very much a member-driven organization. More particularly in the case at hand, the pre-agreement appeared largely proceduralFootnote 31 and motivated by considerations of process economy. As the outcome of the process was virtually a given in light of the many previous Panel and AB reports dealing with the same issue, it supposedly made sense to limit the procedural steps to the absolute minimum for the benefit of both parties, the Panel and the Secretariat.
Second, the pre-agreement in this case was quite extensive. The parties were not simply restraining themselves to a single submission or limiting what issues they would ask the Panel to review; rather, the agreement essentially amounted to a pre-decision resolution. The US agreed not to contest Thailand's claim, Thailand agreed not to ask the Panel to suggest how the US should implement the Panel's recommendations, and the US agreed to implement the Panel's recommendations using specified provisions of US law. As the EU rightly pointed out in its third party submission,Footnote 32 the procedural agreement in fact not only resolved certain procedural issues, it also partially represented a resolution of the dispute between the parties, without, however, any reference being made to Articles 3.6 or 12.7 of the DSU.Footnote 33 The EU considered that:
the ability of parties to a dispute to agree on certain matters and to then have such agreement translated into findings and recommendations of a panel which are eventually adopted by the DSB is not unlimited. The manner of proceeding chosen by the Parties cannot affect the rights of WTO Members which are not parties to the Agreement on Procedures; nor can the approach chosen by the Parties seek to obtain findings having equal weight in practice vis-à-vis other WTO Members as a ‘conventional’ panel report. (Footnote omitted)
This seems a valid concern, although it was to some extent mitigated by the Panel satisfying itself that Thailand had properly discharged its burden of proof.Footnote 34
Third, since it appears that the US and Thailand understood how they would resolve the case, and since the US and Thailand could agree on numerous issues, why could they not resolve the dispute via consultations?
The explanation is rooted in the way the US implemented the prior WTO decisions on zeroing. The US opted to apply the new (no zeroing) rules for dumping calculations only in original investigations done after December 2006. The US refuses to concede that dumping margins calculated under its old procedures must be changed unless there is an explicit WTO dispute decision with respect to each instance. Thus, the US considers that WTO rules require it to calculate post-2006 initial dumping margins without zeroing, but it will only recalculate previously determined margins with an adverse WTO ruling in hand.Footnote 35 There is no mechanism within US procedures for exporters to have their margins recalculated under the new rules without a US court order or a WTO decision telling the USDOC to do so. This limitation explains why the US could agree to recalculate the margins in the pre-agreement in the case of an adverse Panel report but could not resolve the case via consultations. The Panel's decision was needed to overcome the prospective nature of the US's implementation.
3.2 Undisputed claims
Even though the US had not contested Thailand's claim, the Panel nevertheless considered it necessary, in accordance with the approach of previous panels,Footnote 36 to satisfy itself – albeit in a summary fashion – whether Thailand had discharged its burden of proof in light of the Panel's obligation under Article 11 DSU to ‘make an objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements and make such other findings as will assist the DSB in making the recommendations or in giving the rulings provided for in the covered agreements’. It therefore followed the analysis we have outlined in Section 2 above.
The analysis covers some seven of the 11 pages of the report and most of it is taken up by quotations from previous AB and Panel reports. Thus, even though one could argue the necessity of analyzing an undisputed claim, particularly in a simple dispute such as this one, the Panel effectively would appear to have followed a ‘minimum effort, maximum result’ approach, which at least puts some meat on the bone while avoiding possible suspicions about shady deals.
3.3 Implications of the US prospective implementation for the WTO dispute system
As of March 2011, the US had approximately 200 AD orders in place where the margins were calculated with the model zeroing method.Footnote 37 In theory, this could mean that we could see 200 WTO disputes following a similar script as this one: disputes with pre-agreement resolutions; Panel after Panel reconfirming the AB's view that model zeroing is inconsistent with the ADA, all as a result of the US's requirement that a WTO decision be in place before it implements the rulings regarding zeroing on pre-existing margins. The specter of hundreds of such ‘copycat’ cases for the WTO dispute system is alarming, if not silly.
While we believe there will be additional WTO disputes over zeroing, we feel that the zeroing issue is coming to an end and that it is highly unlikely we will see a large number of additional disputes involving zeroing. There are several reasons for our view.
First, the expense, time, and effort to pursue a WTO dispute settlement case mean a firm (government) would likely initiate a WTO dispute only if the revised calculations were to produce de minimis margins. Under the US's retrospective duty collection system, the recalculation effort generates significant benefits only if the margin falls to de minimis levels. Otherwise, deposits will still be collected and a significant portion of the ‘burden’ of US AD duties is unchanged.Footnote 38
A couple of examples will clarify our view. First, suppose that with zeroing an exporter has a dumping margin of 60%, and suppose that without zeroing its margin will fall to 50%. In this case, the exporter has little reason to push for a WTO dispute, as the revised margin is unlikely to generate any economic benefits. Second, suppose that with zeroing the exporter has a dumping margin of 15%, and without zeroing its margin falls to 8%. Even in this case, the benefit of the dispute is attenuated. Under the US system, the 8% is the deposit rate; the actual dumping rate will be determined during the administrative review. Consequently, the specter of a higher final dumping margin will deter many potential buyers. This is especially true if, as currently is the practice, the US uses the zeroing methodology in the administrative review.
How many US AD cases might have de minimis margins without zeroing? Bown and Prusa (2010) suggest that zeroing typically raises the margin as much as 5 to 10 percentage points.Footnote 39 It seems reasonable to think, then, that only cases with margins below 5% or 10% would be worth bringing to the WTO. Using data from Bown (2010), we estimate that approximately 10% (25%) of all currently imposed US AD margins are less than 5% (10%).Footnote 40 This suggests that the US could find itself involved in another 20–50 ‘copycat’ WTO disputes.
Second, this estimate likely overstates the number of possible additional disputes that might proceed similarly to this case. As of mid-2011, the US has stopped zeroing only in original investigations, not in administrative reviews.
Under US AD law, the USDOC will, upon request, periodically conduct ‘administrative reviews’ of an AD order for purposes of (i) calculating the exact dumping margin for each exporter on entries covered by that review, and (ii) recalculating the duty deposit rate for future entries. Reviews are carried out indefinitely until the AD order is either completely or partially revoked. Typically, either domestic firms or foreign exporters will request a review within the first few years after the order was imposed.
The December 2006 policy change affected zeroing only in original investigations. While there have been numerous WTO AB decisions finding zeroing in administrative reviews inconsistent with the ADA, the US has not yet revised its domestic statutes to comply with the rulings. This is a highly significant consideration. If there had been an administrative review in the PET bags case, this dispute would have proceeded in a very different way. Given current US policy, it is unlikely that the US would have conceded the violation and approved such an expansive pre-dispute agreement. Yet, even if Thailand and the US could have reached some type of joint pre-agreement, the US could not have agreed to implement the Panel's recommendations using specified provisions of US law, since US law (with respect to duty calculations in administrative reviews) still allows for zeroing; Thailand would have been unlikely to agree not to ask the Panel to suggest ways for the US to implement the Panel's recommendations. Consequently, much of what makes this case so straightforward (the expansive pre-dispute agreement) would not have occurred.
Even if we do not observe this type of ‘copycat’ disputes, the possibility that we could see additional zeroing disputes to some extent highlights the limitations of the current WTO dispute system, particularly in cases involving highly regulated anti-dumping systems such as that of the US. On the other hand, most other WTO members using anti-dumping laws have typically simply copied and pasted the text of the ADA into their domestic legislation and rely largely on internal guidelines to implement the framework legislation. While this clearly makes their application of the law less transparent and more arbitrary, it has the advantage that it tends to facilitate implementation of adverse AB/Panel reports. Thus, in many cases only the internal guidelines will need to be changed, without necessitating a lengthy – and often politically sensitive – process of legislative change and/or extensive consultations with stakeholders.Footnote 41
Clearly, the prospective nature of WTO remedies limits their effectiveness, as the US zeroing saga abundantly shows. However, the WTO membership seems overwhelmingly unwilling to change this.
3.4 Consequences for zeroing in administrative reviews
As mentioned above, one key aspect of this case was that USDOC had only calculated margins in the original investigation. There had not been an administrative review in which the margins were recalculated.
This case underlines the tension in current US policy with respect to zeroing in original investigations and in administrative reviews. It is quite possible that, had an administrative review been done in this case, the exporters would have the same (or nearly the same) margins as they did in the original investigation. Yet, under current US law, the Thai exporters were entitled to have margins calculated without zeroing only because the original investigation margins were still in effect.
The US's different treatment of zeroing in original investigations and administrative reviews appears to be coming to an end. In December 2010, the USDOC proposed a rule change that would eliminate zeroing in administrative reviews.Footnote 42 While the rule change has not gone into effect, it appears that the US will bring its administrative review methods into harmony with original investigation methods.Footnote 43
This rule change will eliminate the tension in USDOC practice. In addition, the rule change should reduce, if not eliminate, WTO disputes involving the zeroing method in reviews. The new rule will mean that exporters seeking to have their dumping margins calculated without zeroing can do so without resorting to a WTO complaint. Once the USDOC stops zeroing in administrative reviews, exporters will find it less expensive and timelier simply to request a review rather than seek a remedy via the WTO dispute process through their government (which, after all, would ultimately result in the US performing calculations similar to those it would do in an administrative review).
4. Concluding comments
The Panel's decision in the PET bags dispute does not have significant legal or economic interest. In its decision the Panel simply restated longstanding (and numerous) Appellate Body views on the inconsistency of zeroing with Article 2.4.2 of the ADA.
The interesting aspects of the case are (i) why it was needed in light of the pre-agreement; (ii) how it highlights the need for harmony in duty calculation methods in original investigations and administrative reviews, and (iii) the role of pre-agreements for WTO disputes.
As a postscript, we also note that, in the pre-agreement, the US agreed to ‘implement the Panel's recommendations using specified provisions of US law’.Footnote 44 What did this mean? While Thailand assumed that this would mean conducting a weighted-average-to-weighted-average comparison with no zeroing, the US petitioning industry did not interpret the language in the same way. Rather, the US industry requested USDOC to recalculate the margins using the ‘targeted dumping’ methodology, which they asserted would allow the US to continue to zero:
The Department continued to use the average-to-average methodology but allowed ‘negative margins’ on non-dumped products to offset positive margins on dumped products. It is equally permissible, however, to bring a measure into compliance by changing the comparison methodology. In Softwood Lumber, for example, the Department – rather than revise the average-to-average methodology to allow offsets – brought that determination into compliance by switching to a transaction-to-transaction methodology. Although the WTO Panel Report held that offsets were required under the average-to-average methodology, it made no similar finding with respect to the average-to-transaction methodology. The Department, therefore, may bring its determination into conformity with the Panel Report by recalculating margins pursuant to the average-to-transaction methodology.Footnote 45 (Footnote omitted)
The Thai government opposed the change in calculation method and urged the USDOC ‘to reject that request because it would be inconsistent with the agreement between our governments for an expedited resolution of the underlying WTO dispute’.Footnote 46
Ultimately, the USDOC rejected the Petitioners' request because they had not made a targeted dumping allegation in any of their prior submissions. Consequently, the margins were calculated using the weighted average to weighted average method with no zeroing. The revised margins are given in Table 2.
Table 2. Duties without zeroing
Thus, the largest Thai exporters (TPBI) had their margins changed to zero and, given that the recalculation technically involved original investigation margins, the dumping order was revoked for TPBI. This case therefore supports our conjecture that cases where the removal of zeroing results in de minimis margins are the most likely to be challenged at the WTO.
However, it remains to be seen whether the USDOC in future investigations or reviews will now resort to the targeted dumping exception of Article 2.4.2 in cases where such targeted dumping is duly alleged by the petitioners. While the sweeping condemnation of zeroing in several Appellate Body reports indicates that targeted dumping is unlikely to find favor with the AB, similar usage of the targeted dumping method by other major AD users such as the EUFootnote 47 may make it politically and/or strategically more difficult to challenge.
