Koyo Seiko Co., Ltd. And Koyo Corporation of U.S.A. v. The United States, and the Timken Company
Docket: 94-1363
Court: Court of Appeals for the Federal Circuit; September 20, 1995; Federal Appellate Court
The United States Court of Appeals for the Federal Circuit reviewed the September 21, 1993 decision of the United States Court of International Trade regarding Koyo Seiko Co. Ltd. and Koyo Corporation of U.S.A., which challenged the International Trade Administration's (ITA) calculation of final dumping margins for imported tapered roller bearings (TRBs). The appellate court found that the lower court improperly failed to defer to the ITA's reasonable interpretation of statutory provisions concerning antidumping duties. Under 19 U.S.C. § 1673, Commerce is required to impose duties on imports sold at less than fair value, calculated as the difference between foreign market value (FMV) and U.S. price (USP). If FMV cannot be established through market prices, constructed value may be used instead. The case centers on whether the Court of International Trade correctly assessed Commerce's methodology for determining what constitutes "such or similar" merchandise for comparison under 19 U.S.C. § 1677b. The investigation into TRBs began in 1986, covering a variety of uses and involved Koyo’s manufacturing and export operations from Japan to the U.S.
Commerce established six criteria to match U.S. tapered roller bearings (TRBs) with those from home markets: outside diameter, inside diameter, width, physical characteristics (including rows of rollers and heat treatment), basic load rating (BLR), and the Y factor (thrust load effect). The fourth criterion is qualitative, leading to the remaining five being termed the "five criteria." During the less-than-fair-value investigation, Commerce compared U.S. TRB models to home-market models, excluding those with deviations exceeding ten percent in any of the five criteria from being considered "such or similar" merchandise. The most similar home-market TRB was identified based on the smallest deviation among the remaining models. This approach is known as the "greatest-single-deviation" methodology.
In March 1989, Commerce began the first administrative review of TRBs from Japan, initially applying the greatest-single-deviation methodology but later switching to a "sum-of-the-deviations" methodology. This new approach compared TRBs across all five criteria without a ten percent cap on deviations, selecting the model with the lowest total deviation as the best match. Koyo challenged this methodology, arguing that omitting the ten percent cap resulted in comparisons of dissimilar products with distinct applications and pricing. Koyo asserted that significant deviations in critical criteria indicated a lack of commercial similarity. Commerce rejected these arguments and determined a dumping margin of 36.2 percent ad valorem for Koyo TRBs, proceeding with the sum-of-the-deviations methodology.
Koyo initiated litigation in the Court of International Trade to challenge the final results of the 1987-1988 administrative review conducted by Commerce. Koyo contended that Commerce should have continued using the greatest-single-deviation methodology from the initial less-than-fair-value investigation and argued for a ten percent cap on permissible deviations in comparing TRB models. The court acknowledged that both methodologies could yield comparable results but agreed with Koyo's assertion that a ten percent cap was necessary to limit permissible deviations. The court affirmed Commerce's change in methodologies but mandated the cap to prevent comparisons between products that were not commercially similar. Consequently, the court partially granted Koyo’s motion, remanding the case for Commerce to apply the ten percent cap alongside the sum-of-the-deviations methodology. On remand, Commerce imposed a dumping margin of 30.19 percent ad valorem and the Court of International Trade subsequently affirmed this calculation, dismissing the case. The government appealed the requirement for Commerce to apply the ten percent cap.
The standard of review for antidumping duty determinations involves affirming the International Trade Administration's (ITA) decisions unless unsupported by substantial evidence or not in accordance with law. The current case focuses solely on whether Commerce's application of the sum-of-the-deviations methodology without the ten percent cap contravenes statutory requirements. The relevant statute does not specify the methodology for matching U.S. products with similar home-market products, suggesting that Congress has not prescribed a uniform approach for different product types, including TRBs.
In instances where Congress has not explicitly addressed a legal question, courts do not impose their own interpretations but instead assess whether an agency's interpretation is a permissible construction of the statute, as established in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc. Koyo argues that Chevron is inapplicable because the antidumping law does not contain gaps for the Commerce Department to fill, asserting that the methodology used for matching U.S. and home-market tapered roller bearings (TRBs) failed to meet the "such or similar" threshold required by 19 U.S.C. § 1677b. Koyo contends that the court should interpret the statute without deference to agency interpretation, framing the issue as a pure question of law.
Conversely, the government argues that Chevron applies because the statute does not specify how to define "similar" merchandise, suggesting that Congress has granted Commerce discretion in policy decisions regarding product comparisons. The government asserts that this implicit delegation of authority warrants deference to Commerce's reasonable methodologies under Chevron.
The court aligns with the government, recognizing that Congress has implicitly authorized Commerce to determine the model-match methodology necessary for identifying "such or similar" merchandise. It notes that in cases of statutory ambiguity, the court's role is to respect the agency's legitimate policy choices rather than to judge the merits of competing interpretations. The court rejects Koyo's position that the issue is purely legal, affirming that Chevron deference applies to agency determinations made under an implicit delegation of authority, provided the agency's interpretation is a permissible construction of the statute.
Under the Chevron framework, an agency's interpretation of a statute does not need to be the only reasonable one or the one a court would have chosen. The International Trade Administration (ITA) is recognized as the primary authority on antidumping law and its decisions receive significant deference. The focus is on whether Commerce's model-match methodology—lacking a ten percent cap—is reasonable. It is deemed reasonable because no single criterion among the five is shown to be decisive for matching U.S. and home-market tapered roller bearings (TRBs). By not applying the cap, Commerce effectively treats all criteria equally, which is considered reasonable.
Commerce employs additional safeguards, such as excluding comparisons where the variable cost difference exceeds twenty percent and selecting the most similar match when deviations are equal. Despite Koyo's argument that this methodology results in comparisons of commercially dissimilar TRBs, the court agrees with Commerce that it is not necessary for the home-market models to be technically substitutable or used for identical purposes. The statute requires only that the matched products be "such or similar," which Koyo fails to demonstrate as impermissible under the statute.
Consequently, the Court of International Trade's modification of Commerce's methodology was deemed improper, as it should have deferred to Commerce's interpretation under Chevron. The judgment requiring Commerce to impose a ten percent cap is reversed, while the remainder is affirmed, with the case remanded for further proceedings consistent with this opinion.
Each party involved in the case will bear its own costs. The judgment is partially affirmed and partially reversed, with the case remanded for further proceedings. The Timken Company is noted as a defendant but is not part of this appeal. Subsequent to the proceedings in the Court of International Trade and the briefing in this court, relevant statutes—19 U.S.C. Sections 1673, 1677, and 1677b—were amended; however, these amendments do not affect the case's outcome or analysis. The summary of amendments includes:
- Section 1673 now uses "normal value" instead of "foreign market value" and "export price (or constructed export price)" instead of "United States price."
- Section 1677b replaces "foreign market value" with "normal value" and "imported merchandise" with "subject merchandise."
The term "such or similar merchandise" is redefined to include merchandise that is identical in physical characteristics, produced by the same manufacturer, or comparable in material and value, now referred to as "foreign like product" after the revisions. Additionally, a matching methodology example illustrates that the first TRB is selected based on smaller deviations compared to the second TRB. References to specific regulations governing final determinations and administrative reviews of antidumping orders are also cited.
