1. What was the Case about?
This dispute-settlement case involved a complaint from the EC against the practice of ‘zeroing’ by the US Department of Commerce (DOC). The US used ‘zeroing’ in an earlier antidumping review case where the EC had been accused of dumping into the US market. In this dispute case, the Panel judged that ‘zeroing’, which involves throwing out a series of price observations to calculate the dumping margin, is inconsistent with the current Antidumping Law under Article 2.4.2 dealing with the determination of the dumping margin. This decision was consistent with earlier Panel rulings involving previous zeroing cases. However, the Panel also ruled that the US legislation, as such, was not WTO-inconsistent since the practice of ‘zeroing’ is not a mandatory practice of the US DOC.
Prusa and Vermulst, in their discussion of this dispute case, argue that while they approve of the Appellate Body's decision, they doubt the grounds on which the decision was based. More precisely, they point out that Article 2.4.2 of the Antidumping Law does not explicitly prohibit the use of ‘zeroing’. Therefore, for the Appellate Body to dismiss the practice of zeroing on Article 2.4.2, in their view, is not correct. Prusa and Vermulst are of the opinion that while the Appellate Body made the right decision, they should have based their decision on another Article of the Antidumping Law, notably Article 9.3, which deals with transaction-specific duty liabilities. In principle, under US Antidumping Law, duties imposed should not exceed the dumping margin. The practice of zeroing therefore not only affects the dumping margin but also has an effect on the duties imposed.
2. Lifting the ‘zeroing’ mystery
The analysis by which Prusa and Vermulst come to the above conclusion is a very exhaustive one. It has a strong legal component with plenty of references to previous zeroing cases. There is also an elaborate numerical treatment of the ‘zeroing’ issue in the paper. Using simple numerical examples, the paper explains the differences between ‘simple zeroing’ and ‘model zeroing’ and between ‘Average-to-transaction zeroing’ and ‘Transaction-to-transaction zeroing’. Thanks to these efforts, the paper really lifts the ‘zeroing’ mystery in the sense that it makes this highly technical matter accessible to the more nontechnical reader and makes the economic implications of ‘zeroing’ very clear.
Because of the exhaustive nature of the authors' analysis, it is difficult for any discussant to add many things. My contribution will therefore be limited to additional clarification. One downside of the exhaustiveness in the Prusa and Vermulst text is that it is lengthy and at times tedious. Therefore, in the discussion below, I summarize some of the main elements involved, offering a quick insight into the debate.
Simply put, ‘zeroing’ is a correction applied to the calculation of the dumping margin performed by the importing country. This correction involves the elimination of transactions where prices show negative dumping. Dropping observations when calculating the dumping margin tends to inflate overall positive dumping. Since the WTO only sanctions positive dumping, it is easy to verify that dropping the negative ones will inflate the dumping margin and ultimately result in higher import duties. When I started to read about this dispute-settlement case, I was inclined to think that it should be easy for an exporting country to win such a case. It should be straightforward to show that zeroing tends to inflate dumping margins, which makes exporters more vulnerable to high duties. But reading the discussion by Prusa and Vermulst, I soon realized that while for most economists it seems quite obvious that zeroing is unfair, from a legal point of view it is not so easy to show!
3. Zeroing: why is it still there?
Throughout the paper by Prusa and Vermulst, as I read it, the legal grounds for prohibiting zeroing are shaky. Or put differently, there is nothing in the current Antidumping Law that explicitly rules out the practice of zeroing.
The current Antidumping Agreement refers to dumping margins of ‘products’, not of ‘models’ or types. The complainant in this dispute case, the EC, argued that given these definitions, the US should have involved all models or types in the calculations of dumping and not just the ones with positive dumping margins but also those with negative ones. From an economic point of view, this makes sense. Throwing away data is not done unless in very exceptional circumstances with obvious ‘outliers’. The paper draws a useful analogy to illustrate this point. When an academic researcher is faced with outliers, there are several alternatives that are much preferred to eliminating observations. Inspired by the practices of economists, Prusa and Vermulst towards the end of their paper formulate an alternative to ‘zeroing’ that would seem to be preferable in the case of exceptional circumstances, when there are transactions that show unusually low prices resulting in negative dumping margins. The paper concludes that the challenge for WTO members in the future will be how to deal with outliers in a way that is WTO-consistent.
4. Alternative suggestion
Personally, I would be inclined to think that there is another alternative route that the WTO could take to resolve the high number of dispute cases involving zeroing, and that is to change the current Antidumping Law to explicitly ban the practice of zeroing. Such a change of the law could be complemented with a clause where only in exceptional circumstances would an importing country be allowed to disregard particular transactions in their calculation of the dumping margin. Important in such a clause would be that the burden of proof of the ‘exceptional character of the excluded transactions’ would be put on the side of the importing country.
A change in the law involving a ban of zeroing would most definitely reduce the number of disputes surrounding it.
1. What was the Case about?
This dispute-settlement case involved a complaint from the EC against the practice of ‘zeroing’ by the US Department of Commerce (DOC). The US used ‘zeroing’ in an earlier antidumping review case where the EC had been accused of dumping into the US market. In this dispute case, the Panel judged that ‘zeroing’, which involves throwing out a series of price observations to calculate the dumping margin, is inconsistent with the current Antidumping Law under Article 2.4.2 dealing with the determination of the dumping margin. This decision was consistent with earlier Panel rulings involving previous zeroing cases. However, the Panel also ruled that the US legislation, as such, was not WTO-inconsistent since the practice of ‘zeroing’ is not a mandatory practice of the US DOC.
Prusa and Vermulst, in their discussion of this dispute case, argue that while they approve of the Appellate Body's decision, they doubt the grounds on which the decision was based. More precisely, they point out that Article 2.4.2 of the Antidumping Law does not explicitly prohibit the use of ‘zeroing’. Therefore, for the Appellate Body to dismiss the practice of zeroing on Article 2.4.2, in their view, is not correct. Prusa and Vermulst are of the opinion that while the Appellate Body made the right decision, they should have based their decision on another Article of the Antidumping Law, notably Article 9.3, which deals with transaction-specific duty liabilities. In principle, under US Antidumping Law, duties imposed should not exceed the dumping margin. The practice of zeroing therefore not only affects the dumping margin but also has an effect on the duties imposed.
2. Lifting the ‘zeroing’ mystery
The analysis by which Prusa and Vermulst come to the above conclusion is a very exhaustive one. It has a strong legal component with plenty of references to previous zeroing cases. There is also an elaborate numerical treatment of the ‘zeroing’ issue in the paper. Using simple numerical examples, the paper explains the differences between ‘simple zeroing’ and ‘model zeroing’ and between ‘Average-to-transaction zeroing’ and ‘Transaction-to-transaction zeroing’. Thanks to these efforts, the paper really lifts the ‘zeroing’ mystery in the sense that it makes this highly technical matter accessible to the more nontechnical reader and makes the economic implications of ‘zeroing’ very clear.
Because of the exhaustive nature of the authors' analysis, it is difficult for any discussant to add many things. My contribution will therefore be limited to additional clarification. One downside of the exhaustiveness in the Prusa and Vermulst text is that it is lengthy and at times tedious. Therefore, in the discussion below, I summarize some of the main elements involved, offering a quick insight into the debate.
Simply put, ‘zeroing’ is a correction applied to the calculation of the dumping margin performed by the importing country. This correction involves the elimination of transactions where prices show negative dumping. Dropping observations when calculating the dumping margin tends to inflate overall positive dumping. Since the WTO only sanctions positive dumping, it is easy to verify that dropping the negative ones will inflate the dumping margin and ultimately result in higher import duties. When I started to read about this dispute-settlement case, I was inclined to think that it should be easy for an exporting country to win such a case. It should be straightforward to show that zeroing tends to inflate dumping margins, which makes exporters more vulnerable to high duties. But reading the discussion by Prusa and Vermulst, I soon realized that while for most economists it seems quite obvious that zeroing is unfair, from a legal point of view it is not so easy to show!
3. Zeroing: why is it still there?
Throughout the paper by Prusa and Vermulst, as I read it, the legal grounds for prohibiting zeroing are shaky. Or put differently, there is nothing in the current Antidumping Law that explicitly rules out the practice of zeroing.
The current Antidumping Agreement refers to dumping margins of ‘products’, not of ‘models’ or types. The complainant in this dispute case, the EC, argued that given these definitions, the US should have involved all models or types in the calculations of dumping and not just the ones with positive dumping margins but also those with negative ones. From an economic point of view, this makes sense. Throwing away data is not done unless in very exceptional circumstances with obvious ‘outliers’. The paper draws a useful analogy to illustrate this point. When an academic researcher is faced with outliers, there are several alternatives that are much preferred to eliminating observations. Inspired by the practices of economists, Prusa and Vermulst towards the end of their paper formulate an alternative to ‘zeroing’ that would seem to be preferable in the case of exceptional circumstances, when there are transactions that show unusually low prices resulting in negative dumping margins. The paper concludes that the challenge for WTO members in the future will be how to deal with outliers in a way that is WTO-consistent.
4. Alternative suggestion
Personally, I would be inclined to think that there is another alternative route that the WTO could take to resolve the high number of dispute cases involving zeroing, and that is to change the current Antidumping Law to explicitly ban the practice of zeroing. Such a change of the law could be complemented with a clause where only in exceptional circumstances would an importing country be allowed to disregard particular transactions in their calculation of the dumping margin. Important in such a clause would be that the burden of proof of the ‘exceptional character of the excluded transactions’ would be put on the side of the importing country.
A change in the law involving a ban of zeroing would most definitely reduce the number of disputes surrounding it.