U.S. Steel Group v. United States

24 Ct. Int'l Trade 757, 2000 CIT 99

• Court: United States Court of International Trade
• Decided: August 15, 2000
• Docket: Court 97-06-01015
• Status: Published

Opinion

Opinion

Restani, Judge:

This matter is before the court on plaintiffs’ USCIT Rule 56.2 motion for judgment on the administrative record. Plaintiffs, domestic steel companies, challenge the final determination in Certain Corrosion-Resistant Carbon Steel Flat Products and Certain Cut-to-Length Carbon Steel Plate from Canada, 62 Fed. Reg. 18,448 (Dep’t Commerce 1997) (final results of antidumping duty admin, review) [hereinafter “Final Results”]. At issue therein was the second review period of August 1, 1994 through July 31, 1995.

Plaintiffs request application of adverse facts available pursuant to 19 U.S.C. § 1677e(b) (1994) on the basis that Algoma Steel, Inc. failed to provide cost information requested by the United States Department of Commerce (“Commerce” or “the Department”). Alternatively, plaintiffs request a remand for a new review because the information accepted by Commerce was unreasonably distorted.

Jurisdiction and Standard of Review

The court has jurisdiction under 28 U.S.C. § 1581(c) (1994). In reviewing final determinations in antidumping duty determinations, the court will hold unlawful those agency determinations which are unsupported by substantial evidence on the record, or otherwise not in accordance with law. 19 U.S.C. § 1516a(b)(l)(B)(i) (1994).

Background

During the administrative review, Commerce requested that Algoma respond to the cost of production (“COP”) portion of section D of Commerce’s questionnaire. Antidumping Questionnaire (Sept. 14, 1995), at 1, PR. Doc. 9, Pls.’ App., Tab 4, at 1. Section D requested Algoma (1) to report COP figures based on the actual costs incurred by Algoma during the period of review (“POR”) as recorded under its normal accounting system; and (2) to calculate the reported COP figures on a weighted-average basis using model-specific production quantity as the weighting factor. Id. at D-l to D-2, Pis.’ App., Tab 4, at 2-3. If Algoma produced the merchandise under review at more than one facility, it was to report COP based on the weighted-average of costs incurred at all facilities. Id. at D-2, Pis.’ App., Tab 4, at 3. Algoma explained in its responses to Commerce’s original and supplemental questionnaires, that it was not reporting COP based on the weighted-average costs incurred at each of its two rolling mills. Algoma’s Response to Section B of Questionnaire (Nov. 22, 1995), at B-59 to B-60, PR. Doc. 43, Pis.’ App., Tab 5, at 8-9; Algo-ma’s Response to Sections A, B and C of Supplemental Questionnaire (Jan. 19, 1996), at 34, PR. Doc. 53, Pis.’ App., Tab 7, at 3.

Algoma produced all plate sold during the POR at its facility in Sault Ste. Marie, where most of the slab was rolled into plate on the 166” Plate Mill (“plate mill”). Response to Section B, at B-59, Pis.’ App., Tab 5, at 7. For approximately 20 percent of the total Canadian and U.S. sales reported, however, slab was rolled into plate on Algoma’s 106” Wide Strip Mill (“strip mill”). Id. According to Algoma, it was not in a position to report actual rolling costs for the subject merchandise at each mill because (1) its cost accounting system computed one average rolling cost for all products rolled on the plate mill and one average rolling cost for all products on the strip mill; (2) less than five percent of the sales of products rolled on the strip mill during the POR would be considered plate based upon Commerce’s width/gauge definition for subject merchandise; and (3) Algoma had no records that would permit direct calculation of costs incurred at the strip mill that related only to plate defined by Commerce as subject merchandise. Id., at B-59 to B-60, Pis.’ App., Tab 5, at 7-8.

It appeared to Algoma that it had two options to calculate rolling costs for the plate rolled at the strip mill: either (1) assign the average cost of the strip mill to the small fraction (less than five percent) of products produced there that constituted subject merchandise; or (2) assign the average rolling cost of the plate mill to all plate. Response to Sections A, B, and C, at 34, Pis.’ App., Tab 7, at 3. It appears that the first option would not have been an appropriate choice, because less than five percent of the products rolled on the strip mill during the POR consisted of subject merchandise. Thus, an attempt to allocate costs of the strip mill to the small fraction of the subject merchandise produced on that mill would have been a relatively speculative exercise because virtually all of the cost of the mill relates to non-subject merchandise sheet products. See id. The second option appeared to be a good substitute because it was a conservative cost approach because, during the POR, the cost of producing plate on the strip mill was substantially less than the cost of producing plate on the plate mill. Id. 1

Algoma chose the second option. It reported estimated weighted-average rolling costs based upon the actual rolling costs incurred at the plate mill. To allocate these costs to specific products, Algoma developed a “productivity matrix” (or production factors) based upon the length of time it took to produce a product of a specific width and thickness on each mill. Response to Section B, at B-57, Pis.’ App., Tab 5, at 5. For each product (i.e., “CONNUM”), Algoma weight-averaged the productivity factor for the plate mill with the productivity factor for the strip mill to derive a composite productivity factor. Id. at B-58, Pis.’ App., Tab 5, at 6. Algoma then applied these composite productivity factors to the average cost of production on the plate mill to derive product specific costs for all CONNUMs. Id. at B-56 to B-59, Pis.’ App., Tab 5, at 4-7.

At verification, Commerce examined the issue of the two mills in great detail, including Algoma’s analysis of plate mill versus strip mill rolling costs. Verification of Algoma’s Cost Response (Aug. 12, 1996), at 10-13, P.R. Doc. 112, Def.’s App., Ex. 1, at 10-13. At verification, Algoma explained that, although it did not track width and gauge for costing purposes in the normal course of business, it did have sensors that can track the length of time that a slab product spends on the mill and that slabs were time stamped for both the plate mill and the strip mill. Id. at 11, Def.’s App., Ex. 1, at 11. After the slabs were time stamped, the data was entered into a mill performance data base, from which Algoma selected the weight and time data for slabs produced during the POR and those rolled to plate gauges and sorted the slabs by CONNUMs. Id. Commerce verifiers examined a summary of the mill performance data base for both mills, which showed the percentages of the plate mill production and of the strip mill production that were captured by the data base. The verifiers were able to tie the volume and value amounts to process cost sheets for both mills. Id. at 11-12, Def.’s App., Ex. 1, at 11-12.

Based upon Algoma’s responses and the results of the verification, Commerce accepted Algoma’s reported costs, stating in pertinent part:

Algoma’s cost reporting methodology is reasonable, considering (1) we verified its cost accounting system, (2) Algoma’s verified inability to determine specific rolling costs based upon the gauge of the material being manufactured at either facility, (3) the conservative methodology adopted by Algoma and verified by the Department, and (4) respondent’s compliance with Department instructions on cost reporting methodology in this review.

Final Results, 62 Fed. Reg. at 18,451.

Discussion

Commerce’s decision not to apply facts available to Algoma for the first administrative review period based on this exact reporting methodology was sustained in Bethlehem Steel Corp. v. United States, No. 96-05-01313, 2000 WL 726931, at *2-5 (Ct. Int’l Trade June 2, 2000). The reasoning of that decision on this point is adopted here. Whatever one’s view of Commerce’s decision to accept Algoma’s methodology, Commerce did accept it here and accepted it previously. Further, there is no allegation that Algoma deceived Commerce or somehow tricked Commerce into accepting a faulty methodology. Thus, Algoma cannot be penalized under 19 U.S.C. § 1677e(b) by the use of adverse facts available for failing to comply to the best of its ability. Algoma gave Commerce exactly what was requested after Commerce’s final decision on what it would accept.

The next issue is whether remand for a new review is required because the methodology was distortive. First, the fact that Commerce accepted a different methodology (essentially an expanded productivity matrix) in a subsequent review is irrelevant. Many methodologies may be acceptable. The only real basis for objecting to this methodology hinges on its effect on the difference in merchandise (“DIFMER”) 2 adjustment.

Commerce was aware that accepting some high costs (as indicated, normally adverse to the respondent) might cause DIFMER adjustments more favorable to respondents, but it reasonably concluded that COP allocation issues were paramount. 3 Final Results, 62 Fed. Reg. at 18,451. The parties disagree as to the number of sales comparisons affected by a possible DIFMER distortion and the magnitude of the potential distortion. The court concludes, however, that no possible factual scenario in this record could render Commerce’s choice unreasonable or not supported by the record.

First, while Commerce attempts to use the most directly related costs of production as reported by respondents, see 19 U.S.C. § 1677b(f)(l)(A), sometimes allocations are required. As recognized in Bethlehem whenever Commerce:

relies on a respondent’s other, existing data to ascertain the cost of production, a petitioner may argue that they distort the DIFMER. But the law does not require rebanee on actual costs, and the record indicates that the [Department] made a reasonably accurate assessment of the costs in this case, thereby minimizing any arguable distortion.

Bethlehem, 2000 WL 726931, at *5. 4

Second, Commerce’s view that in this case the DIFMER issue likely could affect an extremely small portion of the sales comparison is supported. 5 Commerce reasonably decided not to require a different cost allocation methodology based on the possibility of a DIFMER distortion for a few sales. The court finds the remainder of plaintiffs’ arguments are without merit.

Accordingly, Commerce’s determination is sustained.

1 Higher costs are usually adverse to the respondent. Substantial below cost sales may result in use of cost-based constructed value instead of actual price and a high constructed value will result in a larger antidumping duty margin. See 19 U.S.C. § 1677b(b) (1994) and infra, note 3.

2 it is well recognized that, in calculating margins, it is not always possible to compare the product sold in the United States to an identical product sold in the home market. If there is no identical product in the home market, the statute directs the Department to base its margin calculation on the next most similar product. 19 U.S.C.A. § 1677(16) (West Supp. 1999). The statute recognizes, however, that an adjustment to price is necessary to account for the fact that the price of the home market product and the price of the U.S. market product will reflect the different costs associated with their different physical characteristics. 19 U.S.C. § 1677b(a)(6)(C)(ii), referring to 19 U.S.C.A. § 1677(16)(B) or (C). The DIFMER adjustment is used to eliminate this cost difference and to permit a fair comparison of the two prices.

3 In this case the DIFMER issue relates only to rolling costs and not to all costs, so that the DIFMER distortion would have to be quite significant to affect the outcome. On the other hand, the COP calculation is central to any antidumping review. Higher cost numbers tend to lead to higher normal values, and thus higher antidumping margins. Sales at prices below cost in the home market are subject to being eliminated from the calculation of normal value. See 19 U.S.C. § 1677b(b)(l). Higher costs thus tend to remove low priced sales, increasing normal value and increasing antidumping margins. Further, where there are no sales above cost for a given home market product, U.S. sales will be compared to constructed value, not home market prices. See id. Sesee 19 U.S.C. § 1677b(a)(4). And constructed value itself is largely composed of a respondent’s costs, so higher costs again will tend to increase dumping margins. See 19 U.S.C. § 1677b(e).

4 Section 1677b(f)(l)(A) reflects Commerce’s long established preference for using a respondent’s most directly related reported costs. See 19 U.S.C. § 1677b(f)(l)(A). Commerce, however, is not required to use costs reflected in respondent’s records which are distortive. See Thai Pineapple Public Co. v. United States, 187 F.3d 1362, 1366 (Fed. Cir. 1999), cert, denied, 120 S.Ct. 1830 (2000) (stating that agency may accept records kept according to generally accepted accounting principles or reject records which would distort company’s true costs).

5 In this case, G of Algoma’s [] sales of subject merchandise in the United States were matched to non-identical products sold in the home market. Final Analysis Memorandum (Apr. 3,1997), at 1, C.R. Doc. 82, Pis.’ App., Tab 16, at 1. In all of these Ü non-identical matches, the U.S. product was classified within a CONNUM that was produced only on the plate mill. See Comparison of U.S. Products Matched to Non-Identical Home Market Sales, Def. Int.’s App., Ex. 4. Furthermore, [] of these U.S. transactions were matched to home market products within CONNUMs that were produced only on the plate mill. Id. (Plaintiffs contend that the G sales implicate strip mill costs based on petitioners’ method of allocation of costs, but Commerce is not barred from testing the hypothetical potential for DIFMER distortion based on full CONNUM information.) For DIFMER on the D transactions, therefore, Commerce could conclude strip mill costs are irrelevant. Thus, Commerce also could conclude that if Algoma were to use a methodology that allocated strip mill costs to the product categories that were produced on that mill, no costs would be allocated to the CONNUMs involved in these Q transactions, because the strip mill did not produce any products that are classified in those CONNUMs. See Final Analysis Memorandum, at 1-2, Pis.’ App., Tab 16, at 1-2.