For foreign exporters, one of the most favourable changes to United States law resulting from the Uruguay Round is the requirement that the anti-dumping calculations in original investigations be based on comparison of weighted average home market prices to weighted average United States prices. Under prior United States law, the Commerce Department calculated the anti-dumping margin by comparing individual United States prices to weighted average home market prices (which resulted in much higher margins). Under current United States law, the Commerce Department is supposed to base its calculation in an original investigation on a comparison of weighted average prices.

Although the United States statute allows for departure from an average-to-average comparison (and resort to individual-to-average comparisons) upon a finding that a respondent has engaged in 'targeted dumping', to date the Commerce Department has been reluctant to apply this exception.

Normally, the Commerce Department will calculate weighted average prices by model (ie. control number or CONNUM) and by level of trade (to the extent that there are different levels of trades) for the entire investigation period. Although the Commerce Department also has the discretion to create further `averaging groupings', it has done so relatively infrequently.

Note, however, that the benefit of average-to-average comparison is offset by the Commerce Department's practice of 'zeroing'. 'Zeroing' refers to the Commerce Department's practice of changing negative dumping margins (i.e. no dumping) for individual products to zero when calculating the overall anti-dumping margin for the company. This practice is best illustrated with an example.

Assume that a foreign company subject to a United States anti-dumping case sells five product models to the United States, A, B, C, D, and E and further assume that the Commerce Department's anti-dumping calculation for individual models yields the following results.

Thus, depending on how the negative margins are treated, the dumping margin can vary substantially. If the negative margins are allowed to offset the positive margins, the total value of dumping is only 915 and the margin is only 4.93%. But if the negative margins are set to zero, the effect of the two transaction that are dumped is exaggerated - the value is 1,510 and the margin rises to 8.14%.

A WTO panel has ruled that the EU's practice of zeroing violates the WTO Anti-Dumping Agreement. In EC - Anti-dumping duties on imports of cotton-type bed linen from India, the Appellate Body ruled that zeroing was inconsistent with the language of Article 2.4.2 of the Anti-Dumping Agreement, at least in the context of original investigations. Similarly, in Softwood Lumber from Canada, the WTO Dispute Settlement Body ruled that the United States practice of zeroing in original investigations was also inconsistent with the language of Article 2.4.2 of the Anti-Dumping Agreement. However, as of this writing, the Commerce Department has not changed its practice to eliminate zeroing as part of its investigation methodology.

In the meantime, foreign companies should be aware of this issue in their cases, and build a record from which their country can challenge this United States practice. Since zeroing can have a significant impact on dumping margins, and often will be the difference between dumping margins and legally de minimis margins, it is very much in the interest of foreign companies to pursue this issue.

Source: Business Guide to Trade Remedies in the United States: Anti-dumping, countervailing and safeguards legislation practices and procedures