What's New?
 - Sitemap - Calendar
Trade Agreements - FTAA Process - Trade Issues 

español - français - português
Search

 BINATIONAL PANEL REVIEW
PURSUANT TO
THE NORTH AMERICAN FREE TRADE AGREEMENT
ARTICLE 1904

 

In the Matter of:

Brass Sheet and Strip
from Canada 
Secretariat File No.

USA-CDA-98-1904-03

DECISION OF THE PANEL

July 16, 1999


Appearances:  David A. Harquist, Jeffrey S. Beckington, and Mary T. Staley, 
Collier, Shannon, Rill & Scott, P.C., on behalf of Hussey Copper, Ltd. et al.
  Ryan Trainer and Carrie A. Simon, Rogers & Wells, on behalf of Wolverine Tube (Canada) Inc.
  Steven J. Powell, Berniece A. Browne, Linda Chang, and Mildred Stewart
Office of the Chief Counsel for Import Administration, on behalf of the Investigating Authority, the United States Department of Commerce.

TABLE OF CONTENTS

  1. INTRODUCTION
  2. BACKGROUND
  3. STANDARD OF REVIEW
  4. OPINION
  1. THE ISSUE OF WHETHER COMMERCE PROPERLY USED A SIMPLE AVERAGE COP IS PROPERLY BEFORE THIS PANEL
  1. Wolverine has not waived its Administrative Remedies
  2. The ITA is not Barred by the Doctrine of Administrative Finality from Correcting the COP
  1. THE CALCULATION OF WOLVERINE'S COST OF PRODUCTION
  1. Did the ITA violate 19 USC §1677(m)(d)?
  2. Did the ITA violate 19 USC § 1677(e)(b) and unlawfully make an adverse inference against Wolverine?
  3. Did the ITA fail to use the COP data submitted by Wolverine, in violation 19 USC 1677(m)(e)?
  4. By Calculating a Simple Average Cost of Production, the ITA did not follow its established Administrative Practice
  1. THE U.S. INDUSTRY'S ALLEGATIONS OF ERROR BY THE ITA IN COMPUTING COST OF PRODUCTION
  1. The U.S. Industry has exhausted its Administrative Remedies
  2. Wolverine's Indirect Selling Expenses in the Home Market
  3. The Toll Sales Handling Charge
  1. DISPOSITION

I. INTRODUCTION

This Panel has been constituted pursuant to Article 1904(2) of the North American Free Trade Agreement. The Panel was appointed to review the final results of the administrative review issued by the U.S. Department of Commerce's International Trade Administration [hereinafter the "ITA"] in Certain Brass Sheet and Strip from Canada. 1

In those final results, which covered U.S. imports of the subject merchandise during 1996, the ITA determined that the weighted average dumping margin was 0.67%. Because the margin was more than de minimis (i.e., more than 0.50%), the ITA also decided not to revoke the antidumping order on the subject merchandise.

On August 14, 1998, Wolverine Tube (Canada) Inc. [hereinafter "Wolverine"], a Canadian producer and exporter of the subject merchandise, and a coalition representing the United States brass industry [hereinafter the "U.S. Industry"] filed separate complaints regarding the final results.

The only issue raised by Wolverine 2 to be decided is whether the ITA, in calculating a dumping margin above the de minimis level, erred in using a simple average annual cost of production ("COP") rather than a weighted average COP. This will be referred to as the "calculation issue". 

Wolverine's calculation issue raises a number of questions which may be subsumed under the  broad heading of whether the ITA, in conducting its administrative review, fulfilled its duty of fairness. Wolverine alleges in this regard that the ITA breached three procedural requirements mandated by the Tariff Act. These issues are:

Did the ITA breach U.S.C. §1677(m)(d) by failing to inform Wolverine of perceived deficiencies in its submitted data, and failing to allow Wolverine to remedy or explain the deficiency?

Did the ITA breach U.S.C. §1677(e)(b) by making an inference adverse to the interests of Wolverine without first finding that Wolverine had failed to cooperate by not acting to the best of its ability to comply with a request for information from the ITA?

Did the ITA breach U.S.C. §1677(m)(e) by declining to consider information submitted by Wolverine that was timely, verified, and could be used without undue difficulty?

The U.S. Industry has raised the following issues related to the calculation issue:

Was it proper for the ITA to have used a "simple" average COP, rather than one weighted by production quantities?

Did Wolverine fail to exhaust its administrative remedies by not having objected to the simple average COP after it was used in the preliminary results?

Is the ITA precluded by the doctrine of administrative finality from now asserting that it had committed an error in using a simple average COP?

The ITA, as will be explained further, has now before this Panel taken the position that its calculation of COP on a simple average basis is not supported by substantial evidence or is not in accordance with law.

The U.S. Industry also raises the following alleged errors by the ITA: 1

Did the ITA improperly fail to add Wolverine's Canadian indirect selling expenses to the COP?

Did the ITA improperly allow Wolverine to decrease its per unit COP for "nontoll" sales by allowing an offset to its costs based on toll sales (which are nonsubject merchandise).

For the reasons more fully set forth below, and on the basis of the administrative record, the applicable law, the written submission of the ITA, Wolverine, and the U.S. Industry, and the Panel hearing held in Washington, D.C. on April 16, 1999, the Panel remands in part, and affirms in part, the ITA's Final Results.

II. BACKGROUND

On January 12, 1987, the ITA published an antidumping order on brass sheet and strip from Canada. On March 3, 1997, the ITA initiated its tenth annual review of the underlying dumping order. 2 The period under review was January 1 through December 31, 1996. Wolverine was the sole manufacturer or exporter involved in the review. Wolverine had requested the review and had requested revocation of the antidumping order.

On March 7, 1997, the ITA transmitted its first questionnaire to Wolverine. Wolverine responded on April 28, 1997, including sales and cost of production ("COP") data. It also distinguished between product made from "roll" and "as cast" material. For quarters when certain product was not produced, Wolverine reported the production quantity as "zero", although it did report costs for those quarters. Wolverine also reported Canadian sales for the "window" periods (the last quarter of 1995 and the first quarter of 1997).

On July 15, 1997, the ITA issued a supplemental questionnaire to Wolverine. 3 Question 15 instructed Wolverine to:

Delete source as an element of the product control numbering system and submit a single weighted average cost for each unique product as represented by a specific matching control number. Adjust the cost of unit production.

The ITA never specifically stated that it required Wolverine to report an average annual COP as opposed to quarterly COP data. Wolverine apparently did not interpret Question 15 as requiring the calculation of average annual COP and continued to provide quarterly data in response to Commerce's request.

Questions 5 and 15 of the supplemental questionnaire also asked Wolverine to either eliminate the roll/as-cast distinction or re-classify its product categories using physical characteristics of the finished merchandise. Wolverine re-classified its products according to brass purity and relative grain size: the net effect of which was the same as the previous roll/as-cast distinction.

During December 1997, the ITA conducted sales and cost verifications at Wolverine's facilities. As part of the verification, the ITA verified the quarterly costs for one product that was not produced in some quarters and for which Wolverine had reported "zero" production in those quarters. 

The ITA published its preliminary results of the review on February 9, 1998. 4 The ITA preliminarily determined that Wolverine's margin of dumping for the 1996 period was 0.42% and therefore was de minimis. The ITA, based, in part, on this de minimis dumping margin, tentatively determined to revoke the antidumping order. 

Shortly afterwards, the ITA released its analysis memorandum and computer calculations underlying the Preliminary Results. In the Preliminary Results, the ITA had stated that it disagreed with Wolverine's distinction between relative grain sizes. However a review of the computer program language showed that Commerce failed to eliminate this distinction from its calculations. The ITA had not merged or "collapsed" the COPs for ultra fine grain and other products. With the "uncollapsed" data, the ITA had averaged together the separately reported COPs for each of the product codes. In calculating this average, the ITA used a simple average annual COP for the four quarters of data. 

In April 1998, Wolverine and the U.S. Industry provided written submissions to the ITA on the Preliminary Results and a public hearing was held shortly thereafter. Both Wolverine and the U.S. Industry noted that Commerce in its calculations had, contrary to its stated intention, failed to ignore the ultra fine grain distinction. 

On June 9, 1998 the ITA released its analysis memorandum and computer calculations underlying the Final Results. These documents reveal that the ITA had now eliminated the distinction by "collapsing" the COPs for ultra fine grain and other products. However, in the Final Results, the ITA did not weight average the COPs of the collapsed data; it again used a simple average. 

The final results of the tenth review were published on June 17, 1998. 5 The ITA determined that Wolverine's dumping margin was 0.67% - above de minimis - and, accordingly, did not revoke the order with respect to imports of the subject merchandise from Wolverine. 

On June 18, 1998 Wolverine submitted a request that the ITA correct what it described as a clerical error by the ITA. One week later, the U.S. Industry responded to Wolverine's request. On July 14, 1998, the ITA issued a memorandum (dated June 24, 1998) responding to Wolverine's request for a correction. In that memorandum, the ITA purported to explain what it had done.

Wolverine and the U.S. Industry submitted separate Complaints before this Panel. On August 28, 1998, the ITA submitted its Notice of Appearance in which it admitted no errors had been committed in the Final Results. Nearly two months later, the ITA amended the "Admission" portion of that Notice of Appearance as follows:

With respect to Count Two of the Complaint submitted by Wolverine Tube (Canada), Inc., the Department erred in utilizing a simple average cost of production, rather than a weighted average cost of production, in calculating its Final Results of Administrative Review. 6

The U.S. Industry moved for leave to add the following allegation to its Complaint: That the ITA committed an error in "admitting" that it had improperly relied on a "simple", rather than "weighted", average in calculating Wolverine's normal value. On January 28, 1999, the Panel denied the U.S. Industry's motion to amend. 

Wolverine filed a Notice of Motion For Remand on November 2, 1998. Wolverine sought an immediate remand, without a hearing, of the contested determination back to the ITA for correction of the admitted error and for consideration of its request for revocation. Wolverine also asserted that, once the ITA's simple averaging error is corrected, its margin would be 0.42% - below the de minimis level. In its Order of January 28, 1999, the Panel denied Wolverine's motion for an immediate remand and set a briefing and hearing schedule.

Briefs were submitted and oral arguments were heard at the public hearing on April 16, 1999.

III. STANDARD OF REVIEW

Both Wolverine and the U.S. Industry, as well as the ITA, agree that the applicable standard of review is specified by NAFTA Articles 1904(2)–(3) and Annex 1911 of the NAFTA. Chapter 19 review panels are directed by Article 1904(3) to apply:

the standard of review set out in Annex 1911 and the general legal principles that a court of the importing Party otherwise would apply to a review of a determination of the competent investigating authority.

These provisions therefore require that an Chapter 19 panel apply the standard of review and "general legal principles" which a federal court in the United States would otherwise apply in reviewing an ITA antidumping determination. 7

Annex 1911 defines the standard of review to be applied in a Panel review as "the standard set forth in section 516A(b)(1)(B) of the Tariff Act of 1930, as amended." Section 516A(b)(1)(B), in turn, defines that standard of review as:

The court shall hold unlawful any determination, finding, or conclusion found...to be unsupported by substantial evidence on the record, or otherwise not in accordance with law. 19 U.S.C.A. §1516a(b)(1)(B).

Accordingly, the standard of review for the instant proceeding includes the "substantial evidence" test as set out in section 516A(b)(1)(B) of the Tariff Act of 1930, as amended (19 U.S.C. §1516a(b)(1)(B)). 

The Panel must, therefore, affirm the ITA's Final Results "unless we conclude that the ITA determination is not supported by substantial evidence or is otherwise not in accordance with law." PPG Industries, Inc. v. United States, 978 F.2d 1232, 1236 (Fed. Cir. 1992).

The U.S. Supreme Court has interpreted "substantial evidence" as follows: 

Substantial evidence is more than a mere scintilla, and must do more than create a suspicion of the existence of the fact to be established. "It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion,"...and it must be enough to justify, if the trial were to a jury, a refusal to direct a verdict when the conclusion sought to be drawn from is one of fact for the jury. 8

The Court of Appeals for the Federal Circuit has applied the same interpretation of "substantial evidence" in reviewing antidumping determinations. 9

Furthermore, substantial evidence constitutes "something less than the weight of the evidence." Consolo v. Federal Maritime Commission, 383 U.S. at 619-20 (1966). "The possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency's finding from being supported by substantial evidence." Matsushita Electric Industrial Co., Ltd. v. United States, 750 F.2d at 933 (Fed. Cir. 1984) (quoting Consolo v. Federal Maritime Commission, 383 U.S. at 619-20); and PPG Industries, Inc. v. United States, 978 F.2d at 1237 (Fed. Cir. 1992). 10

It is "not within the Court's domain either to weigh the adequate quality or quantity of the evidence for sufficiency or to reject a finding on grounds of a differing interpretation of the record." Koyo Seiko Co. Ltd. v. United States, 810 F. Supp. 1287, 1289 (CIT 1993) quoting Timken Co. v. United States, 699 F.Supp. 300, 306 (1988) aff'd, 894 F.2d 385 (Fed. Cir. 1990); and Can-Am Corp. v. United States, 664 F.Supp. 1444, 1450 (CIT 1987).

Panel review of an anti-dumping determination is to be conducted "upon the administrative record." Article 1904(2). 11

Therefore, the Panel is not to review the agency determination de novo. Cabot Corp. v. United States, 694 F. Supp. 949, 952-53 (CIT 1988); Ceramica Regiomontana, S.A. v. United States, 636 F. Supp. 961, 966 (CIT 1986) aff'd. 810 F.2d 1137 (Fed. Cir. 1987); Luciano Pisoni Fabbrica Accessori v. United States, 640 F. Supp. 255, 256 (CIT 1986).

The requirement that a review be "on the record" means that a Panel's review must be limited to only "information presented to or obtained by [the ITA]...during the course of an administrative proceeding...." 19 U.S.C. §1516a(b)(2)(A)(i). Consideration of information which was not presented to, or obtained by, the ITA during the course of an administrative review would be beyond the jurisdiction of this Panel. As will later become clear, this requirement is of some real significance to the disposition of this matter.

Neither the Court of Appeals for the Federal Circuit or the Court of International Trade ("CIT") "may ... substitute its judgment for that of the [agency] when the choice is 'between two fairly conflicting views, even though the court would justifiably have made a different choice had the matter been before it de novo.'" Tehnoimportexport, UCF America Inc. v. United States, 783 F. Sup. 1401, 1404 (CIT 1992) quoting Universal Camera Co. v. NLRB, 340 U.S. 474, 488 (1951) and American Spring Wire Corp. v. United States, 590 F. Supp. 1273, 1276 (CIT 1984) aff'd sub nom., Armco, Inc. v. United States, 760 F.2d 249 (Fed. Cir. 1985). Accordingly, this Panel is similarly constrained. This deference to the agency has its limits. As the CIT has held:

[T]he substantial evidence standard requires courts generally to defer to the methods and findings of an agency's investigation .... [T]he Court must not permit an agency in the exercise of that discretion to ignore or frustrate the intent of Congress as expressed in substantive legislation that the agency is charged with administering....Were the scope of the discretion accorded to the agency unlimited, there would be no point in the (statutorily mandated) judicial review here undertaken.

Armco, Inc. v. United States, 733 F.Supp. 1514, 1519 (CIT 1990)(citations omitted); Cabot Corp. v. United States, 694 F.Supp. 949, 953 (CIT 1988)(and cases cited therein). 

The other element of the standard of review (whether the determination is "in accordance with law") 12 applies to questions of statutory interpretation by the agency. Section 516A(b)(1)CB) of the Tariff Act of 1930, as amended, 19 U.S.C.A. §l516a(b)(1)(B).

In determining whether the ITA's interpretation of the statute is "in accordance with law", the Panel is to afford deference to the agency's reasonable interpretation of the statute which it administers. "The Supreme Court has instructed that the courts must defer to an agency's interpretation of the statute an agency has been charged with administering provided its interpretation is a reasonable one." PPG Industries, Inc. v. United States, 928 F.2d 1568, 1571, rehearing denied and rehearing en banc declined (Fed. Cir. 1991). 13 Also, Zenith Radio Corp. v. United States, 437 U.S. 443, 450-51 (1978); Georgetown Steel Corp. v. United States, 801 F.2d 1308, 1318 (Fed. Cir. 1986); American Lamb Co. v. United States, 785 F.2d 994, 1001 (Fed. Cir. 1986); Consumer Product Division, SCM Corp. v. Silver Reed America. Inc., 753 F.2d 1033, 1039 (Fed. Cir. 1985); and Smith Corona Group v. United States, 713 F.2d 1568, 1571 (Fed. Cir. 1983) cert. denied 465 U.S. 1022 (1984).14 This deference extends to the administering authority's interpretation of its own regulations as well. 15

In accordance with this principle of administrative law, the ITA has been granted great discretion in administering the anti-dumping duty laws. "Given these circumstances, appellant's burden on appeal is a difficult one, for it must convince us that the interpretation ... adopted by the ITA is effectively precluded by the statute." PPG Industries, Inc. v. United States, 928 F.2d at 1571, rehearing denied, and rehearing en banc declined (Fed. Cir. 1991) 16

Nonetheless, this discretion and deference is not unfettered. "The traditional deference courts pay to agency interpretation is not to be applied to alter the clearly expressed intent of Congress." Saudi Iron and Steel Co. (Hadeed) v. United States, 675 F. Supp. 1362, 1365 (CIT 1987).

In the instant review, the standard of review is not affected by the fact that the ITA itself, in addition to Wolverine, is requesting a remand. The ITA would like to change the final margin calculation program to use a fully weighted average in calculating the cost of production for Wolverine, in accordance with its standard practice. 

IV. OPINION

A. THE ISSUE OF WHETHER COMMERCE PROPERLY USED A SIMPLE AVERAGE COP IS PROPERLY BEFORE THIS PANEL

The U.S. Industry has raised two procedural objections to Wolverine's and the ITA's argument that the simple average calculation constitutes grounds for remand. The first is that Wolverine failed to exhaust its administrative remedies regarding this issue during the ITA administrative review. 17 The second is that the Department itself, at this point, is precluded by the doctrine of administrative finality from revisiting the COP calculation issue and seeking to revise it. 18

(1) Wolverine Has Not Waived its Administrative Remedies 

The U.S. Industry asserts that the methodology employed by Commerce in its Preliminary  Results was identical to that used in the Final Results. It argues that Wolverine, not having raised any objection to that methodology before the agency, is therefore precluded from raising it here before this Panel for the first time. 

Wolverine presents several points in response: It argues that it had been given no notice by the ITA Preliminary Results as to why a simple average COP was being calculated. The ITA had only announced that it had deleted the grain size characteristic (which it then failed to do in the Preliminary Results' computer program). Based on the prior 1995 review, which had used a weighted average COP, Wolverine claims it had no reason to expect a different approach once the grain size characteristic was deleted. Accordingly, Wolverine specifically noted, in its comments on the Preliminary, only that the ITA program had not deleted the grain size characteristic. Wolverine also claims that it called for a weighted average by having given the ITA computer command language which provided for a weighted average COP.

Wolverine further claims that the ITA never mentioned until after its Final Results that the ITA had concerns with Wolverine's data and, therefore, had to resort to a simple average, thereby depriving Wolverine of an opportunity to respond. Wolverine argues that the exhaustion doctrine does not apply where the agency has not disclosed to the parties what it intended to do. It also argues that where another party made a similar argument before the agency, the exhaustion requirements have been met. It claims that the U.S. Industry, in its comments on the Preliminary Results to calculate, had urged the ITA to use a weighted average COP calculation, thus preserving the issue for Panel review. Lastly, it claims that the exhaustion defense can be waived by the agency, which the ITA has done so in this case by now agreeing with Wolverine that it erred using a simple average COP.

The ITA supports Wolverine's position. 19 It disputes the U.S. industry's claim that the COP averaging methodology was the same as in the preliminary's. It agrees that the issue had been raised by the U.S. Industry before the agency, as Wolverine did in its reply brief to the agency, yet the ITA still did not address the issue in its Final Results. The ITA asserts as well that it, as the agency, is not bound by the exhaustion doctrine and can raise this issue here.

The Panel agrees that the issue is properly before it for review. It is true, as the U.S. Industry claims, that Wolverine was aware at the time of the Preliminary that a simple average had been used by the ITA. Wolverine seems to admit as much. 20 Wolverine claims that it did not object because a simple average of the "uncollapsed" products would produce the same de minimis margin as a weighted average of the same products. Implicit in this is that Wolverine would have objected had the simple average been used with the "collapsed" products. However, Wolverine was on notice that the ITA fully intended to drop the grain size characteristic, even though it failed to implement that change in its preliminary computer calculations. Wolverine knew that a simple average COP calculation of the "collapsed" products would not be correct and would distort the costs. 21

Nonetheless, we believe that Wolverine's submission to the ITA, before the Final Results, of computer language calculating a weighted average has preserved the issue. More importantly, Wolverine also told the ITA in its Reply Brief, "If the Department nevertheless determines in the final results to follow the course taken in the Preliminary Results...certain modifications to the program will be needed...to calculate a weighted-average cost of production." 22 While Wolverine now states that it had not really "objected" to the Preliminary Results (although the ITA asserts that it had), the Panel believes that its statement in the Reply put the ITA on notice and preserved the issue.

While the U.S. Industry may be correct that the only relevant difference between the preliminary and final computer methodology was the product "collapsing" in the final, the fact remains that Wolverine proposed a weighted average before the final results. That the ITA failed to follow Wolverine's approach is beside the point.

The exhaustion of administrative remedies doctrine ("exhaustion doctrine") is a fundamental principle of U.S. administrative law. See, e.g., Schwartz, Timing of Judicial Review: A Survey of Recent Cases, 8 Admin. L.J. 261 (Summer 1994) ("Exhaustion of administrative remedies and its twin doctrine of primary jurisdiction are vital both to the operation of the administrative process and the effective judicial review of agency action"). 

In accordance with the doctrine, "[w]here Congress specifically mandates, exhaustion is required." 23 Moreover, there is a specific statutory obligation for the CIT to require the exhaustion of administrative remedies. "[I]n any civil action not specified in this section, the CIT shall, where appropriate, require the exhaustion of administrative remedies." 28 U.S.C.§2637(d). 

Where Congress is silent, exhaustion is still "the general rule" because it "serves the twin purposes of protecting administrative agency authority and promoting judicial efficiency." McCarthy v. Madigan, 503 U.S. 146 (1992).

One rationale, of course, for the exhaustion doctrine is that an agency should have notice of an issue while it is still deliberating the matter so that it can resolve it at that point. Myers v. Bethlehem Shipbuilding Corp., 33 U.S. 41, 50-51 (1938); Hormel v. Helvering, 312 U. S. 552, 557 (1940). In Weinberger v. Salfi, 422 U.S. 749, 765 (1975), the Supreme Court explained that:

Exhaustion is generally required as a matter of preventing premature interference with agency processes, so that the agency may function efficiently and so that it may have an opportunity to correct its own errors, to afford the parties and the courts the benefit of its experience and expertise, and to compile a record which is adequate for judicial review.

See, also, Davis, K. & Pierce, R., II Administrative Law Treatise, §15.2 at 309 (3d ed. 1994).

The U.S. Court of Appeals for the Federal Circuit, whose decisions this Panel is bound to recognize and follow, has stated that "judicial review is inappropriate unless and until the person seeking to challenge the action has utilized the prescribed administrative procedure for raising the point." Sharp Corp. v. United States, 837 F.2d 1058, 1062 (Fed. Cir. 1988). 

The CIT has often refused to consider arguments on issues which the parties failed to bring to the ITA's attention during underlying antidumping or countervailing duty investigation. See, e.g., Borden, Inc. v. United States, 4 F. Supp.2d 1221, 1232 (CIT 1998)(plaintiff not permitted to challenge certain calculations to offset the difference in commissions between the U.S. and foreign markets when it had not challenged the methodology during the antidumping investigation); Saarstahl, A.G. v. United States, 949 F. Supp. 863, 866 (CIT 1996) (alleged use of improper period for allocating the benefit of nonrecurring subsidies); Aramide Maatschappij V.o.F. v. U.S., 901 F. Supp. 353 (CIT 1995) (partly precluded from raising argument about exclusion of certain sales from the profit calculation included in constructed value because sales had been excluded in preliminary determination and had not been objected to); Budd Co., Wheel and Brake Division v. United States, 15 C.I.T. 446, 773 F.Supp. 1549, 1555-1556 (1991) (alleged flaws in methodology used for compensating for Brazilian hyperinflation in calculating cost factors, with court noting that plaintiff "declined to raise its contentions at the administrative level as an apparent tactical decision"); Rhone-Poulenc, Inc. v. United States, 13 C.I.T. 218, 226-227, 710 F. Supp. 341 (1989), aff'd 899 F.2d 1185 (Fed. Cir. 1990) (plaintiff precluded from raising new argument that Commerce should have accounted for exchange and interest rate fluctuations since the entry of the original antidumping order).

It is thus clear that this Panel may not consider an argument that the CIT would be constrained from hearing by virtue of the fact that a party had failed to exhaust its administrative remedies with respect to that argument.

While Wolverine could certainly have been more expansive in its "objection" to the simple average, its submission to the ITA was sufficient to preserve the issue.

Even if Wolverine had made no such submission between the Preliminary and Final Results, the fact that the ITA failed at every opportunity to reveal to the parties the basis for simple averaging (viz., that it had problems with Wolverine's data) provides another basis to allow this issue to be heard. It is not fair to penalize a party for not exhausting its remedies where the party "could not attack a policy they could not be aware existed." Bowen v. New York, 476 U.S. 467, 476 (1986)(quoting Heckler v. New York, 578 F. Supp. 1109, 1118 (E.D.N.Y. 1984)). 

Moreover, the Panel finds that the U. S. Industry's submission to the Department was also sufficient to put the Department on notice and preserve the issue. The U.S. Industry did state that "the different costs reported by Wolverine should be collapsed to calculate a single weighted-average cost or products classified under each unique product control number." 1 This placed the issue in play and the Department failed to respond or explain why it did not address the point. 2

Where one party to an agency proceeding raises a point before an agency, another party to the proceeding - who did not raise the issue - may argue the same point on appeal. In National Resources Defense Council v. EPA, 824 F.2d 1146 (D.C. Cir. 1987)(Bork, J.), in an unanimous en banc decision, the D.C. Circuit permitted NRDC, which had not even participated in the agency proceeding, to appeal an issue raised by other parties. "Courts have waived exhaustion if the agency 'has had an opportunity to consider the identical issue [presented to the court]....but which were raised by other parties'...The EPA, therefore, had notice of his issue and could, or should have, taken it into account in reaching a final decision...." 824 F.2d at 1151 (citing Buckeye Cablevision, Inc. v. United States, 438 F.2d 948, 951 (6th Cir. 1971) and Office of Communication of the United Church of Christ v. FCC, 465 F.2d 519, 523 (D.C. Cir. 1972). We find here that the ITA was on notice and should have addressed the weighted average issue raised before it by the U.S. Industry.

Moreover, regardless of whether either party in fact raised the weighted average issue before the ITA, we find that the ITA may, in fact, waive the exhaustion issue. E.g., Mathews v. Diaz, 426 U.S. 67, 76-77 (1976), Mathews v. Elridge, 424 U.S. 319 (1976); Dugan v. Ramsay, 727 F.2d. 192 (1st Cir. 1983). "The power of an agency to waive exhaustion of any administrative remedy except a remedy that Congress has made jurisdictional seems well established. It also makes a great deal of sense." Davis, K. & Pierce, R., II Administrative Law Treatise, §15.6 at 335 (3d ed. 1994). The ITA has now waived the exhaustion rule by adopting, in essence, Wolverine's position that using the simple average was in error and that it failed to provide a proper reason for selecting that methodology. 3

For all of the above reasons, we hold that the issue of whether the ITA reasonably calculated a simple average COP has not been waived and is properly before this Panel for review.

(2) The ITA is Not Barred by the Doctrine of Administrative Finality from Correcting the COP

The U.S. Industry argues that the ITA is barred by the doctrine of administrative finality from "revisiting" on remand the COP calculation issue. The premise for the U.S. Industry's position is that the COP methodology used by the ITA is supported by substantial evidence and in accordance with law. It argues that the ITA should not be permitted to simply choose a different methodology on remand. As Judge Restani of the CIT stated in Badger-Powhatan v. United States, 10 CIT 241, 633 F. Supp. 1364, 1369 (1986): "There is no indication anywhere in the statutory scheme that the agency, for policy or similar reasons, may simply change a final determination after an antidumping order is issued....If the agency could amend determinations endlessly, it would be difficult to answer the question as to when a final determination could ever be made."(Emphasis in original). 

The U.S. Industry relies on Borden, Inc. v. U.S., 4 F. Supp. 2d 1221, 1242 (CIT 1998). However, Judge Restani also held there that "An agency cannot disturb the finality of its determinations...,except where a mistake has been shown." Here we determine infra, that a mistake by the ITA has been shown: the agency's use of the simple average is not in accordance with law. Moreover, in Borden, the court went on to say:

Here, neither the opposing parties nor the government made any specific representation that Commerce erred in its original CEP calculation; Commerce merely seeks an additional review of the calculation to rethink and revise. On remand, Commerce may not revisit the question of the CEP calculation.

Borden, 4 F.Supp.2d at 1242. In the instant review, of course, both Wolverine and the ITA itself have specifically stated that the ITA erred. There is no remand "to rethink and revise", only to correct an identified error.

In Borden Judge Restani cited her earlier decision in Badger-Powhatan. 4 In Badger-Powhatan she had held that:

The statute contains no exceptions to the general rule that these adjudicatory-type decision, which are relied upon may later be changed for policy reasons....That this principle may give way when errors are committed does not give the agency authority to upset final decisions where no errors have occurred. Therefore, in order to determine if remand is appropriate, the court must rule on plaintiff's motion for judgment on the agency record.

633 F. Supp. at 1369 (Emphasis added). Therefore, we hold that where one interested person as well as the ITA assert that an error was committed by the agency below, and we concur, then the agency is not necessarily barred by the doctrine of administrative finality from correcting that error. 

We now turn to the merits of Wolverine's calculation issue argument.

B. THE CALCULATION OF WOLVERINE'S COST OF PRODUCTION

Wolverine's calculation issue raises a number of questions which can be subsumed under the broad heading of whether the ITA, in conducting its administrative review, fulfilled its duty of fairness. 

Wolverine alleges that the ITA breached three procedural requirements mandated by the Tariff Act:

Did the ITA breach 19 U.S.C. §1677(m)(d) by failing to inform Wolverine of perceived deficiencies in its submitted data, and failing to allow Wolverine to remedy or explain the deficiency?

Did the ITA breach 19 U.S.C. §1677(e)(b) by making an inference adverse adverse to the interests of Wolverine without first finding that Wolverine had failed to co-operate by not acting to the best of its ability to comply with a request for information from the ITA?

Did the ITA breach 19 U.S.C. §1677(m)(e) by declining to consider information submitted by Wolverine that was timely, verified, and could be used without undue difficulty?

As a preliminary matter, it is useful to note, as set out in Borden, Inc. v. U.S., 4 F. Supp. 2d 1221 (CIT 1998), that the Tariff Act was amended in 1994 to comply with the United States' obligations under the Uruguay Round GATT Agreements. The general object was to foster more fairness and transparency in administrative proceedings, a goal already articulated in the applicable jurisprudence (e.g., National Steel Corp. v. United States, 870 F. Supp. 1130). 

Behind these statute-based issues lie related questions which are dealt with separately. One important matter is whether the ITA is required to explain the methodology it uses to calculate cost of production, something that was not done specifically in this case. In addition, there is the related or parallel question of whether the ITA is required to explain and notify parties when it decides to depart from long-standing agency practice. More specifically, the question arises as to whether the ITA is required to give notice to parties, and explain its decision, when it decides to depart from its longstanding practice of calculating weighted averages in determining cost of production. 

The Relevant Facts

In order to decide these issues, it is helpful to review the more relevant facts. 

In the 1996 review, Commerce issued its standard questionnaire seeking information, among other things, necessary to calculate costs of production.

Pursuant to instructions, Wolverine, in responding, assigned codes to various products by control numbers ("CONNUMs") for purposes of comparison. For each product code Wolverine provided quarterly material costs per unit, a weighted average fabrication cost, and the quantity of each CONNUM which was produced in each quarter.

Significantly, there were quarters in which a given CONNUM was not produced. In this instance, Wolverine nonetheless calculated, and supplied, a quarterly Cost of Production. This figure was arrived at by using the materials cost encountered in producing other products (a standard cost), and the annual fabrication cost for that CONNUM. The production quantity was indicated as "zero".

The U.S. Industry asserts, and it seems not to be disputed, that the quarterly materials cost for all CONNUMS was the same for each quarter, and that the fabrication costs were the same for each CONNUM for all quarters. Thus a constructed COP calculation was possible in spite of the fact that some CONNUMS were not actually produced in some quarters.

Pursuant to the ITA request, Wolverine also reported Canadian sales for the "window" periods, the last quarter of 1995 and the first quarter of 1997.

The ITA, in a second questionnaire, asked Wolverine to:

Delete source as an element of the product control numbering system and submit a single weighted average cost for each unique product as represented by a specific matching control number. Adjust the unit cost of production accordingly.

The ITA did not otherwise request that Wolverine supplement its responses to the original questionnaire as to its cost of production calculations. Subsequently, ITA verified the submitted data, and, in its verification report 1 reported no discrepancies in the data which Wolverine had submitted.

Go to the continuation of this Panel Decision


1 Final Results of Antidumping Duty Administrative Review and Notice of Intent Not to Revoke in Part, 64 Fed. Reg. 33,037 (June 17, 1998).

2 Wolverine initially alleged two errors by the ITA. During the Panel proceedings, it later abandoned one issue: whether the ITA erred in excluding the grain size and purity of its brass products from the product comparison criteria.

1 The U.S. Industry also raised, and briefed, an argument that the ITA erred in making an adjustment to normal value for Wolverine's "quantity adder". The U.S. Industry dropped this issue at the Panel hearing and, accordingly, it is no longer part of this review and will not be addressed by the Panel.

2 Initiation of Antidumping and Countervailing Duty Administrative Reviews and Request for Revocation in Part, 62 Fed. Reg. 9413, 9414 (1997).

3 Letter from Tom Futtner to Carrie Simon, 1-3 (July 15, 1997)(document does not appear in the  administrative record given to this Panel; attached as Exhibit 1 to Wolverine's Reply  Brief).

4 Brass Sheet and Strip from Canada, Preliminary Results of Antidumping Duty Administrative  Review and Notice of Intent to Revoke Order, 63 Fed. Reg. 6519 (1998).

5 Brass Sheet and Strip from Canada: Final Results of Antidumping Duty Administrative Review  and Notice of Intent Not to Revoke Order in Part, 63 Fed. Reg. 33,037 (1998).

6 Letter to James R. Holbein from Linda S. Chang, Esq., 5 (October 16, 1998) (emphasis in  original).

7 Annex 1911 defines such "general legal principles" as, for example, "standing, due process, rules  of statutory construction, mootness, and exhaustion of legal remedies."

8 NLRB v. Columbian Enameling & Stamping Co., 306 U.S. 292, 300 (1939) quoting from  Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 (1938). See, also, the Supreme  Court's decisions in Consolo v. Federal Maritime Commission, 383 U.S. 607, 619-20  (1966); Universal Camera Corp. v. NLRB, 340 U.S. 474, 477 (1951); and Consolidated  Edison Co. v. NLRB, 305 U.S. 197, 229 (1938).

9 E.g., Matsushita Electric Industrial Co., Ltd. v. United States, 750 F.2d 927, 933 (Fed. Cir. 1984)  and Atlantic Sugar, Ltd. v. United States, 744 F.2d 1556, 1562 (Fed. Cir. 1984).

The U.S. Court of International Trade employs this definition as well. E.g., Koyo Seiko  Co., Ltd, v. United States, 810 F.Supp. 1287, 1289 (CIT 1993); Tianjin Machinery Import  & Export Corp. v. United States, 806 F.Supp. 1008, 1013 (CIT 1992); Tehnoimportexport,  UCF America Inc. v. United States, 783 F.Supp. 1401, 1404 (CIT 1992); Minebea Co.,  Ltd. v. United States, 782 F.Supp. 117, 119 (CIT 1992) aff'd. 984 F.2d 1178 (Fed. Cir.  1993); and Armco. Inc. v. United States, 733 F.Supp. 1514, 1518 (CIT 1990).

10 Also, Zenith Electronics Corp. v. United States, 812 F. Supp. 228, 231 (CIT 1993); Minebea Co.  Ltd. v. United States, 782 F. Supp. at 119 (CIT 1992); Torrington Co. v. United States,  745 F. Supp. 718, 723 (CIT 1990) aff'd. 938 F.2d 1276 (Fed. Cir. 1991); and American  Spring Wire Co. v. United States, 590 F Supp. 1273, 1276 (CIT 1984) aff'd sub nom.  Armco, Inc. v. United States, 760 F.2d 249 (Fed. Cir. 1985).

11 Section 516A(b)(1)(B) of the Tariff Act of 1930 as amended, 19 U.S.C. §1516a(b)(1)(B), similarly  limits the Panel's review to information placed on the record during the administrative proceeding.

12 NAFTA Article 1904(2) states that the "law" to be considered shall consist of "relevant statutes,  legislative history, regulations, administrative practice and  judicial precedents to the extent that a court of the importing Party would rely on such  materials." Decisions of the United States Supreme Court and the United States Court of  Appeals for the Federal Circuit are binding on this panel.

13 Citing Chevron. U.S.A.. Inc. v. Natural Resources Defense Council, 467 U.S. 837, 844 (1984);  Udall v. Tallman, 380 U.S. 1, 16, rehearing denied, 380 U.S. 989 (1965); K Mart v.  Cartier, Inc., 486 U.S. 281, 291 (1988); and United States v. Riverside Bayview Homes,  Inc., 474 U.S. 121, 131 (1985).

14 The Court of International Trade has often applied this principle. See, e.g., Tianjin Machinery  Import & Export Corp, v. United States, 806 F.Supp. 1008, 1013 (CIT 1992); PPG  Industries, Inc. v. United States, 712 F. Supp. 195, 197-98 (CIT 1989) aff'd. 978 F.2d  1232 (Fed. Cir. 1992); and Cabot Corp. v. United States, 694 F.Supp. 949, 953 (CIT  1988).

15 "Since Commerce administers the trade laws and its implementing regulations, it is entitled to  deference in its reasonable interpretations of those laws and regulations." PPG  Industries, Inc. v. United States, 712 F.Supp. at 198 (CIT 1989) aff'd 978 F.2d 1232 (Fed.  Cir. 1992).

16 Prior to the passage of the Trade Agreements Act of 1979, the Treasury Department, which  administered the antidumping law, also enjoyed such discretion. United States v. Zenith  Radio Corp., 562 F.2d 1209, 1216 (C.C.P.A. 1977) aff'd 437 U.S. 443 (1978).

17 Case Brief of the U.S. Industry at 26 et seq. 

18 Id. at 32-33.

19 Rule 57(2) Response Brief of the Administering Authority at 41-45.

20 "The only modification that the Department made to Wolverine's COP data was that it converted  the Company's submitted quarterly COPs (for the 'uncollapsed' products) into simple (or  arithmatic) [sic] average annual COPs." Canadian Complainant's Brief at 10. See, also,  the Department's Preliminary Disclosure Documents which clearly identified that a simple  average had been made. See Memorandum from Paul Stolz to the File (Undated) (Non.  Pub. Admin. Rec. 36/01/35).

21 Canadian Complainant's Brief, 11 n.38.

22 Wolverine Reply Brief, at 36 (Admin. Rec. Non-Pub 34/39/30).

23 McCarthy v. Madigan, 503 U.S. 146 (1992).

1 Administrative Case Brief, CD 25 at 10.

2 The U.S. Industry claims that it never intended to urge the Department to collapse the data using  the numbers submitted by Wolverine. U.S. Industry's Reply to the U.S. Government's and  Wolverine's Reply Briefs, 16 n. 5. That may well be, but the fact remains that the basic  issue of using a weighted average was raised.

In addition, the CIT has also noted that "[a] party may be excused from failure to raise an  argument before the administrative agency as long as the agency in fact considered the  issue." Holmes Products Corp. v. United States, 16 C.I.T. 1101, 1992 Ct. Intl. Trade  LEXIS 258, *7-8 (citing cases).

3 Letter from Linda Chang, Esq. to James Holbein, NAFTA Secretariat, U.S. Section, 2 (October  16, 1998).

4 The court also cited, in both cases, Supreme Court precedent supporting the point that in federal  regulatory licensing cases, adjudicatory-type decisions, which are relied upon, cannot be  changed for policy reasons. Citing American Trucking Assoc. v. Frisco Transportation Co., 358 U.S. 133 (1958). Also, United States v. Seatrain Lines, 329 U.S. 424 (1947)  and Hirschey v. Federal Energy Regulatory Commission, 701 F.2d 215 (D.C.Cir. 1983).

1 ITA Sales Verification Outline/Report, 3 (undated)(Admin. Rec. Non-Pub. 32/66/24) and Cost  Verification Report, 5-6 (undated)(Admin. Rec. 32/54/23).