OPINION AND ORDER

STANCEU, Judge.

Plaintiff Union Steel Manufacturing Co., Ltd. ("Union") brought this action to contest a final determination ("Final Results") issued by the International Trade Administration, U.S. Department of Commerce ("Commerce" or the "Department"), in the thirteenth administrative review of an antidumping duty order on imports of certain corrosion-resistant carbon steel flat ("CORE") products from the Republic of Korea ("Korea"). See Certain Corrosion-Resistant Carbon Steel Flat Products from the Republic of Korea: Notice of Final Results of the Thirteenth Admin. Review, 73 Fed.Reg. 14,220 (Mar. 17, 2008) ("Final Results"). Previously, the court denied relief on one of plaintiff's claims, which challenged the Department's construction of section 771(35) of the Tariff Act of 1930, 19 U.S.C. § 1677(35) (2006), to allow application of "zeroing," i.e., the deeming of the sales a respondent makes in the United States at prices above normal value to have individual dumping margins of zero rather than negative margins. Union Steel v. United States, 33 CIT ___, ___, 645 F.Supp.2d 1298, 1300 (2009). At the same time, the court granted defendant's request for a voluntary remand on the issue raised by plaintiff's other claim in this case, which contested the application of Commerce's "model-match" methodology in the thirteenth review. Id. at ___, 645 F.Supp.2d at 1309-10. In this second claim, Union challenged Commerce's model-match methodology, under which Commerce compared Union's U.S. sales of painted CORE products to Union's home market sales, which included not only painted CORE products but also "laminated" CORE products, i.e., CORE products coated with a plastic film made of polyethylene terephthalate ("PET") or polyvinyl chloride ("PVC"). Br. in Supp. of the Mot. of Pl. Union Steel for J. Upon the Agency R. 3 ("Pl.'s Br."). In making that comparison, the model-match methodology relied on the definition of "foreign like product" in section 771(16)(A) of the Tariff Act of 1930, under which merchandise may be compared if it is "identical in physical characteristics." 19 U.S.C. § 1677(16)(A).

Before the court is the redetermination the Department issued in response to the court's remand order ("Remand Redetermination"), in which Commerce decided to leave unchanged its model-match methodology. Final Results of Redetermination Pursuant to Remand (Dec. 28, 2009) ("Remand Redetermination"). Plaintiff raises various arguments in opposition to the Remand Redetermination, which defendant-intervenors urge the court to affirm. See Pl. Union Steel's Comments on Def. United States' Final Results of Redetermination Pursuant to Remand ("Pl.'s Comments"); United States Steel Corp.'s Comments on the Final Redetermination Pursuant to Ct. Remand Issued by the Department of Commerce ("U.S. Steel's Remand Comments"); Nucor Corp.'s Comments on Remand Results. Because the record lacks substantial evidence to support a finding that the physical differences distinguishing the laminated CORE products from the CORE products the Department compared to the laminated CORE products are minor and commercially insignificant, the court sets the Remand Redetermination aside as contrary to law and orders Commerce to issue a second remand redetermination in accordance with this Opinion and Order.

I. BACKGROUND

The background of this case is set forth in the court's previous opinion and is supplemented herein. See Union Steel, 33 CIT at ___, 645 F.Supp.2d at 1300-02.

Rejecting a proposal by Union, Commerce, in the Final Results and again in the Remand Redetermination, declined to adopt a separate model-match type category for CORE products that were coated with a plastic film. Remand Redetermination 5-6. Union had sales of such laminated products in Korea during the period of review (from August 1, 2005 to July 31, 2006) but had no sales of its laminated CORE products in the United States during that period. Pl.'s Br. 3. In support of its proposed type category for laminated products, Union argued that its laminated products underwent a different production process than its painted products, were physically different from its painted products because they were coated with a plastic film, and were costlier than its painted products. Id. at 3-7.

As it had in the Final Results, Commerce, in the Remand Redetermination, grouped the home market sales of products Union had sought to have categorized as laminated products within the type category of "All Other" painted products. Remand Redetermination 1; see Issues & Decisions Mem., A-580-816, ARP 3-08, at 14-15 (Mar. 10, 2008) (Admin.R.Doc. No. 4563), available at http://ia.ita.doc.gov/frn/ summary/KOREA-SOUTH/E8-5298-1.pdf ("Decision Mem."); Final Results, 73 Fed.Reg. at 14,221 (incorporating by reference the Decision Mem.). Rejecting the need to change its model-match methodology, Commerce on remand made no change in the final antidumping duty margin of 4.35% that Commerce assigned to Union in the Final Results, Final Results, 73 Fed.Reg. at 14,221, stating that "[i]n accordance with the Court's instructions, the Department has reviewed and reconsidered information on the record to determine whether on remand to revise its model-match methodology to include a separate category for laminated products as advocated by respondent Union Steel." Remand Redetermination 1. The Remand Redetermination states that "the Department finds that record evidence does not support revising its model-match methodology with respect to laminated products." Id.

Upon plaintiff's motion, the court held oral argument on the Remand Redetermination on July 16, 2010. The parties since have made post-argument submissions. Pl. Union Steel's Post-Oral Argument Comments on Commerce's Remand Redetermination ("Pl.'s Post-Oral Argument Comments"); Def.'s Resp. to Pl.'s Post-Oral Argument Comments on Commerce's Remand Redetermination ("Def.'s Post-Oral Argument Resp."); Nucor Corp.'s Post-Oral Argument Resp. Comments ("Nucor's Post-Oral Argument Comments"); U.S. Steel Corp.'s Resp. to Pl. Union Steel's Post-Oral Argument Comments ("U.S. Steel's Post-Oral Argument Resp.").

II. DISCUSSION

The court exercises jurisdiction under section 201 of the Customs Courts Act of 1980, 28 U.S.C. § 1581(c), pursuant to which the court reviews actions commenced under section 516A of the Tariff Act of 1930, 19 U.S.C. § 1516a, including an action contesting the final results of an administrative review that Commerce issues under section 751 of the Tariff Act of 1930, 19 U.S.C. § 1675(a). 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. § 1516a(b)(1)(B)(i). "Substantial evidence is more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion." Consol. Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938).

In the Remand Redetermination, Commerce explained that its model-match type categories, as developed during the investigation and first administrative review, resulted from a determination "that CORE products should be separated into four categories: clad, unpainted, painted, and painted with polyvinylidene fluoride (PVDF)." Remand Redetermination 1. Commerce also explained that during the development of these four categories, certain parties requested that the Department create an additional category for laminated products, that the Department did not do so at the time, and that the Department decided instead to group laminated products with painted products for model-match purposes. Id. at 2. During the review at issue in this case, Commerce rejected Union's proposal to place laminated CORE products in a separate type category instead of combining laminated products with the "other painted" products, i.e., the painted CORE products other than CORE products that are painted with PVDF. Decision Mem. 14-15.

In an administrative review, Commerce is required generally to determine the normal value, the export price or constructed export price, and the dumping margin for each entry of the subject merchandise. 19 U.S.C. § 1675(a)(2). Determining a dumping margin requires Commerce to compare the export price or constructed export price with the normal value, which as a general matter is determined according to the price at which the foreign like product is sold for consumption in the exporting country (the "home market"). 19 U.S.C. § 1677b(a)(1)(B). For this comparison, Commerce, in accordance with the statutory definition of "foreign like product" set forth in 19 U.S.C. § 1677(16)(A), first attempts to match subject merchandise with merchandise "which is identical in physical characteristics with, and was produced in the same country by the same person as, that merchandise." See id. § 1677(16)(A) (emphasis added). Only if "a determination for the purposes of part II of this subtitle," which part is labeled "Imposition of Antidumping Duties," cannot "be satisfactorily made" does Commerce, in accordance with § 1677(16)(B), seek to match the subject merchandise with merchandise produced in the same country and by the same person that is "like that merchandise in component material or materials and in the purposes for which used." Id. § 1677(16)(B) (emphasis added). If the latter determination cannot be satisfactorily made under § 1677(16)(B), Commerce is to consider, under § 1677(16)(C), whether merchandise produced in the same country and by the same person that is "of the same general class or kind as the subject merchandise" and "like that merchandise in the purposes for which used" may reasonably be compared with the subject merchandise. Id. § 1677(16)(C).

On remand, in again rejecting Union's proposal and thereby treating Union's laminated CORE products as "other painted" CORE products, the Department applied the "like product" definition of subparagraph (A) of § 1677(16). See Remand Redetermination 3-6. Although the Remand Redetermination does not state this point clearly, it includes text accompanied by a case citation with a parenthetical, as follows:

Id. at 3 (emphasis added). Defendant's counsel, Ms. Claudia Burke, confirmed at oral argument that the Remand Redetermination applied § 1677(16)(A), as indicated by the Department's not allowing an adjustment to normal value to account for the difference in the variable cost of manufacturing the merchandise ("difmer adjustment") when comparing Union's subject painted CORE products with the home market laminated CORE products. Oral Tr. 49-50 (July 16, 2010).1 Based on the above-quoted parenthetical and the clarification provided by defendant's counsel, the court analyzes the issue presented by the Remand Redetermination according to the § 1677(16)(A) definition of foreign like product. Under that definition, the merchandise sold in the comparison market (here, the home market, Korea) must be "identical in physical characteristics with" the subject merchandise. 19 U.S.C. § 1677(16)(A).

In the case the Department cited, Pesquera Mares Australes Ltda. v. United States, the U.S. Court of Appeals for the Federal Circuit ("Court of Appeals") affirmed the Court of International Trade's decision sustaining the Department's matching of premium grade salmon with comparison market sales of both premium grade and super-premium grade salmon. 266 F.3d 1372, 1385 (Fed.Cir.2001). Affording the Department's statutory construction deference according to the Supreme Court's analysis in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), the Court of Appeals upheld as reasonable the Department's construing the statutory term "identical in physical characteristics" such that merchandise should be considered identical despite the existence of minor, commercially insignificant differences in physical characteristics. Id. at 1383-84 ("Commerce has concluded that merchandise should be considered to be identical despite the existence of minor differences in physical characteristics, if those minor differences are not commercially significant."). The appellate court then sustained, under the substantial evidence standard, the Department's finding "that super-premium was not a commercially recognized separate grade of salmon for purposes of the Japanese salmon import market." Id. at 1384.

In upholding the Department's statutory construction, the opinion in Pesquera provides guidance on the meaning of the statutory term "identical in physical characteristics." Concluding that dictionary definitions of the word "identical" established two distinct common usages, the Court of Appeals saw the choice as between construing the term to mean having the exact same identity or, alternatively, construing the term to mean having such a near similarity or resemblance as to be essentially equal or interchangeable or having such close resemblance and such minor differences as to be essentially the same. Id. at 1382-83. Concluding that a construction of "identical" according to the latter category of definitions was reasonable, the Court of Appeals upheld Commerce's construction of the statutory term "identical in physical characteristics" to mean "that merchandise should be considered to be identical despite the existence of minor differences in physical characteristics, if those minor differences are not commercially significant." Id. at 1384.

The Remand Redetermination appears to rely on the same statutory construction of the term "identical in physical characteristics" that the Court of Appeals affirmed in Pesquera. See Remand Redetermination 4. To the extent that the Remand Redetermination does so, the court defers to the Department's construction. The court proceeds to consider the question of whether the Remand Redetermination lawfully compared laminated and non-laminated, painted CORE products as "identical in physical characteristics" within the meaning of that term as used in 19 U.S.C. § 1677(16)(A) and expounded in Pesquera. Such comparisons are lawful only if substantial evidence on the record in this case can support a factual determination that the physical characteristics distinguishing laminated CORE products from the "other painted," i.e., non-laminated, CORE products to which Commerce compared laminated CORE products are minor and not commercially significant. For the reasons discussed below, the court concludes that substantial evidence does not exist on the record to support such a determination.

A. The Remand Redetermination Relies on Factual Findings that Have Little, If Any, Probative Weight on Whether the Physical Differences Are Minor and Not Commercially Significant

At the outset of its analysis, the Remand Redetermination states that "[i]n the twelfth administrative review," i.e., the review previous to the one at issue in this case, "the Department engaged in a detailed analysis of the sales and cost information submitted by the parties on the record of that review to support its continued reliance on the model-match methodology classifying laminated CORE within the painted category." Remand Redetermination 4 (citations omitted). With respect to the review at issue here, Commerce states that "[o]n remand in the thirteenth administrative review, we have now performed a similar analysis on Union Steel's sales, cost and customer data from the record of the thirteenth administrative review as we did in the twelfth administrative review" and concludes that "the Department's findings are consistent for both reviews." Id. Alluding to the current and the previous review, the Remand Redetermination states four findings of fact, which it characterizes as common to both reviews, id. at 4-5, and then bases the following determination on those four findings: "Based on the above findings, the Department determines that there is nothing new on the factual record of the thirteenth administrative review that would cause us to reach a different conclusion about the model-match methodology from that made in the twelfth administrative review." Id. at 5. Those four findings, even if presumed to be supported by substantial record evidence, are only minimally probative on the factual issue of whether the physical differences between laminated CORE products and painted CORE products are minor and not commercially significant.

Commerce found, first, that "[t]he range of prices for laminated CORE products is within the range of prices for other painted CORE products for both reviews." Id. at 4. Even though grounded in record evidence, this finding has little or no probativity on the factual issue of whether laminated CORE products are distinguished from "other painted" CORE products by only minor and commercially insignificant physical differences. Commerce did not find that two otherwise identical CORE products, one of which is painted and the other of which is laminated, sell at the same or approximately the same price. The record shows that CORE products, regardless of whether painted or laminated, are produced and sold in various product specifications, in a very broad range of prices. Id. attachment 1 ("Model Match Variables"), attachment 2 ("Final Analysis Memo for Laminated CORE Products"). Setting forth price data in broad ranges is not informative on the question of what price difference may exist, typically or on average, between laminated and non-laminated, painted CORE products that do not differ except with respect to the presence of paint as opposed to laminate. Other evidence of record, however, bears directly on that question. See Letter from Union to the Sec'y of Commerce 6 (Nov. 20, 2006) (Admin.R.Doc. No. 3944) ("Union's Questionnaire Resp.") ("Laminating the steel increases ... the sales price by" a significant percentage over painted products.).

The second finding stated in the Remand Redetermination is that "[t]he prices for laminated CORE products are identical to prices for other painted CORE products at a greater frequency in the thirteenth administrative review than in the twelfth administrative review." Remand Redetermination 4. This finding, as stated in the Remand Redetermination, has little if any probative weight for the issue Commerce was required to resolve. It is simply a comparison with prior-review data, the significance of which is not explained.

Third, the Remand Redetermination states that "[a] comparable number of customers purchased both laminated CORE products and other painted CORE products from Union Steel in the twelfth administrative review and thirteenth administrative review." Id. This finding has little relevance to the issue Commerce was required to consider. Regardless of whether a purchaser is a distributor or a manufacturer, the fact that the same customer purchased Union's laminated CORE products as well as Union's non-laminated, painted CORE products does not signify that the customer considered the physical differences in the two products to be minor and commercially insignificant. A distributor may distribute various CORE products, as well as other steel products, and a manufacturer may purchase different CORE products for different applications. Finally, the comparison the Department makes to the corresponding data in the twelfth review, which is not at issue in this case, does not lend significance to the findings Commerce made in the thirteenth review.

The fourth stated finding is that "[t]he total cost of manufacture for laminated CORE products is within the range of the total cost of manufacture for other painted CORE products for both reviews." Id. at 5. Commerce approached the question of a differential in manufacturing costs by examining and comparing broad ranges of manufacturing costs for all types of painted and laminated CORE products. Here also, evidence based on broad ranges has little or no probativity on the issue of whether the two groups of products are distinguished only by minor and commercially insignificant physical differences. Because the record demonstrates that both painting and laminating are conducted on the same types of substrates, more probative evidence would address the question of whether the cost of laminating is equivalent to the cost of painting. Union stated in its response to the Department's questionnaire that its laminating process, on average, costs significantly more to perform than does painting. Union's Questionnaire Resp. 6. In its supplemental questionnaire response, Union stated that "[b]ecause PET film and PVC film are more expensive than the various paints used for other color coated products, including PVDF, and require more complicated processing know-how, the production cost and sales price are higher than other painted products." Letter from Union to the Sec'y of Commerce 20 (Feb. 2, 2007) (Admin.R.Doc. No. 4036) ("Union's Supplemental Questionnaire Resp."). It appears that Commerce declined to consider this evidence that laminated CORE products cost more to produce than painted CORE products. Observing in the Remand Redetermination that "Union proposes to isolate only the costs associated with the differences between laminate versus paint," Commerce dismisses this inquiry by concluding that "[a]s stated above, in the context of our model-match analysis, we look at these products from an overall perspective." Remand Redetermination 9. Commerce reasoned that "[a]ny analysis performed in calculating margins will not be limited to cost differences associated with this one small part of the overall manufacturing process and cost buildup" and gave as an example that it uses total cost of production in determining whether sales are made below cost. Id. The court disagrees with this reasoning. Evidence that laminating a particular CORE product costs more than painting that same CORE product is relevant to the question of whether the physical differences between laminated and non-laminated, painted CORE products are minor and commercially insignificant. Comparing the total cost of producing a laminated CORE product to the total cost of producing a non-laminated, painted CORE product is relevant to that question only if the two products are identical except with respect to the coating, i.e. paint or laminate. Were the physical differences between laminated and non-laminated, painted CORE products to have no significance in the marketplace, and were laminating more costly to perform than painting, it would be irrational for a manufacturer, absent an unusual circumstance, to produce laminated CORE products.2

Even if two hypothetical products cost the same to produce and to buy, it would not necessarily follow that the two products are "identical" for purposes of 19 U.S.C. § 1677(16)(A). Customers still might view the two products as different in a commercial sense. Regardless of relative cost of production, two products that have physical differences that are not minor and commercially insignificant but are approximately equal in commercial value may be compared, in some circumstances, under subparagraph (B) of 19 U.S.C. § 1677(16). In contrast, subparagraph (A) requires that the two products, even if costing the same to buy or produce, be "identical in physical characteristics." 19 U.S.C. § 1677(16)(A) (emphasis added); see Pesquera, 266 F.3d at 1383-84. Therefore, price and production cost are relevant to, but not determinative of, the issue of whether a physical difference has commercial significance. The Department's findings as to price and cost, being based on comparisons of broad price and cost ranges, fail to support even a conclusion that otherwise-identical laminated and non-laminated, painted CORE products sell for the same price and have the same cost of manufacture. As discussed above, the remaining finding, that the same customers purchased both laminated and other painted CORE products from Union, also lacks probativity on the critical issue in this case. These four findings, individually and in aggregate, therefore offer little, if any, support for an ultimate finding that the physical differences in question are minor and not commercially significant.

Although basing its overall conclusion on the four findings discussed above, the Remand Redetermination includes some additional findings. Commerce states that "there are physical differences that exist within each of the `other painted products.'" Remand Redetermination 5. However, the question of whether physical differences between some painted CORE products and other painted CORE products would preclude a proper identical merchandise finding under § 1677(16)(A) is not before the court in this case. The Department's grouping as "identical" all painted CORE products except for PVDF-painted CORE products may or may not be lawful, but in this litigation, Union does not challenge as overly broad the scope of the "other painted" type category, except to claim that the type category does not properly include laminated CORE products. A finding that various physical differences exist among the grouped painted CORE products does not indicate or suggest that the physical differences separating laminated CORE products from any non-laminated, painted CORE product— differences that Commerce itself acknowledges to exist—are minor and not commercially significant.

Similarly, with respect to "the differences between laminated and other painted CORE products articulated by Union," Commerce further states that "[w]e found that there were differences in physical characteristics, end uses, costs and selling prices across all of these products." Id. Summarizing record data, the Remand Redetermination states that "[t]hus, the record shows many differences and similarities between laminated and other CORE products, just as there are among and between all products." Id. at 6. Because, as the record evidence demonstrates, both laminated and non-laminated, painted CORE products are made to various specifications, the model-match methodology groups into the various type categories products that have some physical variations. That the model-match methodology, properly or not, made § 1677(16)(A) comparisons not at issue here, involving CORE products within the same type category that have different physical characteristics and end uses, does not justify the specific decision before the court, which was to compare as "identical" laminated CORE products and non-laminated, painted CORE products.

For the reasons the court has discussed, the various factual findings discussed above have little, if any, probative weight on the issue of whether laminated CORE products and the other CORE products that Commerce found identical to laminated CORE products are distinguished by only minor physical differences that are not commercially significant.

B. The Department's Determination that Laminated CORE Products Are Not "Marketed Separately" from Other CORE Products Is Unsupported by Substantial Record Evidence

In addressing the issue of whether the physical differences distinguishing laminated CORE products from the other CORE products which the model-match methodology considered, under 19 U.S.C. § 1677(16)(A), to be "identical" to laminated CORE products have commercial significance, the Remand Redetermination relies on another finding that lacks probativity. Specifically, the Department finds that "laminated products are included with painted and other products in the same product brochure, which is used by the company to market their products to customers." Id. From that finding, the Department reaches the unwarranted conclusion that "[i]f laminated CORE were such a different or unique product as claimed by Union Steel, the Department might reasonably expect to see record evidence that laminated CORE is marketed separately" and determines that "[t]he record evidence, however, supports the opposite conclusion." Id. The only evidence on which Commerce relies expressly for that factual determination consists of a Union brochure and a brochure of its affiliate, Union Coating Co., Ltd. ("Unico"). See Letter from Union to the Sec'y of Commerce exhibit A-28, at 552-607 (Nov. 13, 2006) (Admin.R.Doc. No. 3933) ("Union's Section A Resp."). Those brochures, and others on the record, do not amount to substantial evidence that laminated CORE products are not marketed separately from other CORE products. Instead, the brochures, as well as other evidence of record consisting of Union's questionnaire responses, refute such a finding.

The inclusion of different products in a company's brochure does not, by itself, mean that the physical differences between those products are minor and not commercially significant. Commerce had no factual basis on which to assume that a manufacturer would never choose to describe commercially different products in the same brochure. What is meaningful instead is the information the product brochures contain about the physical differences between laminated CORE products and non-laminated, painted CORE products.

Although the record contains a Union brochure and a Unico brochure that describe both laminated and non-laminated, painted CORE products, the texts of these brochures differentiate between the two groups of products. The record describes three lines of PET-film laminated CORE marketed by Union, "Unipet," "Unilux," and "White Board," each of which is listed on the page of Union's brochure labeled "High-tech Steel"3 and not on the page labeled "Pre-painted Steel." Id. exhibit A-28, at 575-76. The record also contains a separate brochure for Unipet, which is described as "a laminated steel sheet with the printed PET (Polyethylene Telephthalaid [sic]) film on primer-coated galvanized steel sheet." Union's Supplemental Questionnaire Resp. exhibit B-22. Even though, as Union acknowledges, Unico "do[es] not have a separate product brochure for laminated products," the record shows that Unico has two brands consisting of only laminated CORE products, "Bristar" and "Natulami."4 Id. at 18 & exhibit B-22. Specifically, excerpts from material that appeared on Unico's website describe Bristar and Natulami as "LAMINATED Steel Plate." Id. exhibit B-22. A Unico brochure indicates under a "Product Characteristics" heading that Bristar and Natulami are coated with either PVC or a combination of PVC and PET, while Unico's non-laminated, painted CORE products are not. Union's Section A Resp. exhibit A-28, at 595-603.

The brochures on the record do not constitute substantial evidence for the Department's finding that laminated CORE products are not marketed separately from other CORE products. Together with other record evidence, the brochures establish, to the contrary, that Union produces and markets separate product lines of PET-film-laminated CORE products, i.e., the Unipet, Unilux, and White Board lines of products. Id. exhibit A-28, at 576. The record also shows that Union's affiliate, Unico, markets two lines of CORE products, Bristar and Natulami, laminated with either PVC film or a combination of PVC film and PET film. See, e.g., Union's Section A Resp. exhibit A-28, at 594. The court, therefore, concludes that the record evidence does not sustain Commerce's finding that these products are not "marketed separately."

C. The Record on the Whole Does Not Contain Substantial Evidence Establishing that the Physical Differences in Question Are Minor and Commercially Insignificant

Despite its apparent reliance on the statutory construction affirmed in Pesquera, the Remand Redetermination does not contain a finding that the two groups of products in question are, in a commercial sense, essentially equal or interchangeable. This omission is noteworthy, for two reasons. First, language in the Pesquera opinion suggests that physical differences are minor and commercially insignificant if the two products under consideration can be described as "essentially equal or interchangeable." Pesquera, 266 F.3d at 1382 (quoting The American Heritage Dictionary 639 (2d ed.1991)). Second, as a matter of logic, it is difficult to envision how the physical differences separating the two groups of products at issue in this case could be considered to be minor and commercially insignificant unless the two groups of products are viewed by customers as generally equal or interchangeable in the marketplace. The court is unable to find on the record substantial evidence to support a finding that the two product groups are viewed in this way, and what evidence exists is inconsistent with such a finding. Cf. Pesquera, 266 F.3d at 1378 (describing the Department's findings that in various countries the highest grade of salmon was the "superior" grade, which Commerce found equivalent to the Japanese "premium" grade, and that superior grade salmon would contain some salmon that, like "super-premium" grade salmon, would be without defects).

The Remand Redetermination reaches an overall finding that "although we agree that there are certain physical differences between laminated and other painted products, we disagree that these differences result in a significant commercial difference that would render the products non-comparable." Remand Redetermination 5 (emphasis added). This formulation of a factual determination is conclusory. It fails to address the central issue that Commerce was required to resolve on remand. Two groups of products that are not "identical" might not, for purposes of identification of the foreign like product, necessarily be "non-comparable" under 19 U.S.C. § 1677(16)(B) or (C). However, when proceeding under § 1677(16)(A), as opposed to § 1677(16)(B) or (C), Commerce is held to the "identical" standard, under which differences in physical characteristics in the two groups of products being compared may exist, but if they exist, these differences must be minor and not commercially significant. In this case, Commerce insisted on matching as "identical," according to § 1677(16)(A), Union's sales of non-laminated, painted CORE products in the United States with sales in Korea that included both laminated and non-laminated, painted CORE products. All but one of the findings of fact that the Department made to support that determination had little, if any, probative weight on the issue of the commercial significance of the physical differences. The finding of fact that laminated CORE products are not "marketed separately" from non-laminated, painted CORE products, although it is vaguely expressed, arguably would have had some probativity with respect to that issue. However, that finding is unsupported by substantial record evidence and, instead, is contradicted by the evidence consisting of the brochures and questionnaire responses.

After reviewing the record as whole, the court concludes that record evidence consisting of the brochures, the questionnaire responses, and Union's reported price and cost data does not constitute substantial evidence to support a finding that the physical differences between laminated and non-laminated, painted CORE products are minor and not commercially significant. The brochures and questionnaire responses establish that laminated and non-laminated, painted CORE products result from different coating processes and are comprised of different materials. See, e.g., Union's Section A Resp. exhibit A-28, at 576, 591-93; Union's Questionnaire Resp. 5-6. According to a questionnaire response, these differences gave laminated CORE products different physical characteristics from those of the non-laminated, painted CORE products. Union stated in the questionnaire response that "[c]ompared to the normal painted products, laminated products have physical properties such as unrestricted expression of various patterns, superior durability, environmentally-friendly material, etc." Union's Questionnaire Resp. 6; see also Union's Section A Resp. exhibit A-28, at 576. According to another questionnaire response, the different physical characteristics are significant to Union's customers. Union stated that its "salesmen work with their customers to make sure that they obtain the right product for their needs" and that "[s]hould that need best be served with a laminated product, Union would sell the laminated product to the customer." Union's Supplemental Questionnaire Resp. 22.

A finding that the physical characteristics that distinguish laminated CORE products have no commercial significance must confront the record evidence showing that Union and Unico have invested in production equipment and other resources necessary to produce and market laminated CORE products while concurrently producing and marketing non-laminated, painted CORE products. The product brochures and questionnaire responses establish that Union produces three separate lines of PET-film-laminated CORE products and that its affiliate, Unico, produces two separate lines of CORE products laminated with either PVC film or a combination of PVC film and PET film. In sum, the record evidence refutes a finding that the physical differences between laminated and non-laminated, painted CORE products are minor and not commercially significant.

Defendant-intervenor United States Steel Corporation ("U.S. Steel") argues that the record supports a finding that the physical differences between laminated and non-laminated, painted CORE products are minor because Unico's brochure shows that both groups of products "are used for the exact same purposes." U.S. Steel's Remand Comments 11. U.S. Steel's contention is unwarranted by the record evidence. Unico's brochure shows that laminated CORE products share many applications with certain non-laminated, painted CORE products, e.g., the outer casings on home appliances such as refrigerators and microwaves and applications as interior building materials. Union's Section A Resp. exhibit A-28, 595, 597, 599, 601, 603. Neither this brochure nor other record evidence shows, however, that customers perceive no significant differences between these two groups of products. To the contrary, the record contains evidence that Union's sales personnel recommend laminated CORE products to satisfy the requirements of certain customers, Union's Supplemental Questionnaire Resp. 22, and that laminated CORE products are more durable than non-laminated, painted CORE products, Union's Questionnaire Resp. 6.

Defendant and defendant-intervenors also argue that the physical differences between laminated and non-laminated, painted CORE products were minor and not commercially significant because laminated products were not typically more expensive to produce than non-laminated, painted CORE products. See, e.g., Def.'s Post-Oral Argument Resp. 6. The evidence does not support how these parties characterize the costs of production, and their characterization would have little probative weight in any case. The evidence shows that typically, but not invariably, laminated CORE products were more expensive to produce than non-laminated, painted CORE products: in a comparison of thirty-nine sets of otherwise identical CORE products, laminated products were more expensive to produce than non-laminated, painted CORE products in thirty-four instances. Pl.'s Post-Oral Argument Comments 9 n. 5 & attachment 1. As discussed previously, even if the evidence showed that production costs were not typically different, such evidence would not be determinative on the question of whether the two groups of products are distinguished only by physical differences that are minor and not commercially significant.

Defendant and defendant-intervenors attempt to discredit the data showing that the weighted-average cost to produce laminated CORE products substantially exceeded the weighted-average cost to produce non-laminated, painted CORE products, Union's Supplemental Questionnaire Resp. exhibit B-24, contending that, because the products within both groups have substantially different costs of production, a weighted-average reflects only the relative cost to produce the mix of products sold during the period of review. See, e.g., Def.'s Post-Oral Argument Resp. 9. Regardless of the limits inherent in weighted averages, this argument is unconvincing. Again, the record as a whole fails to support a finding that otherwise identical laminated CORE products and non-laminated, painted CORE products cost the same to produce. See Pl.'s Post-Oral Argument Comments 9 n. 5 & attachment 1 (showing that, for thirty-four of thirty-nine sets of otherwise identical products, the laminated product were more expensive to produce); Union's Questionnaire Resp. 6 (indicating that lamination increases the total cost of manufacture).

Defendant-intervenor Nucor Corporation argues that the evident difference in the costs of producing these groups of products is not relevant because it is mostly due to savings from economies of scale associated with the far greater production of non-laminated, painted CORE. Nucor's Post-Oral Argument Comments 6-11. This contention appears to be based on an unsupported assumption concerning economies of scale, but even taken at face value the contention acknowledges that laminated CORE products cost more to produce than non-laminated, painted CORE products.

Defendant-intervenor U.S. Steel argues that the record evidence demonstrates no "significant" differences between the costs to produce laminated and non-laminated, painted CORE products because (1) raw materials costs were in the same range, with different varieties of both groups of products having low, median, and high production costs, (2) the differences in the costs of paint and laminate are not substantial, and (3) with respect to Unico, the overhead expenses associated with non-laminated, painted CORE products are greater than the overhead expenses associated with laminated CORE products. U.S. Steel's Post-Oral Argument Resp. 4-11. The evidence on which these arguments rely fails to show that laminated CORE products were not typically more expensive to produce than non-laminated, painted CORE products. First, evidence that costs were in similar ranges and were concentrated at similar levels says little about whether production costs for laminated and non-laminated, painted CORE products were typically similar because both laminated and non-laminated CORE products were produced in various sizes and specifications. Remand Redetermination attachment 1 ("Model Match Variables"), attachment 2 ("Final Analysis Memo for Laminated CORE Products"). Second, U.S. Steel's statement that the costs of paint and laminate are similar is not supported by the record, which shows that most of Union's paints cost less than PET film or PVC film. See Letter from Union to the Sec'y of Commerce exhibit D-43, at 27 (Mar. 2, 2007) (Admin.R. Doc. No. 4066). Third, U.S. Steel's argument about overhead costs is unpersuasive because overhead is only one component of Unico's manufacturing cost, and Unico's total per-unit cost of manufacture for laminated CORE products is greater than its total per-unit cost of manufacture for non-laminated, painted CORE products. See Union's Supplemental Questionnaire Resp. exhibit B-24.

For the various reasons discussed above, the court concludes that the record as a whole does not contain substantial evidence that the physical differences separating laminated CORE products and non-laminated, painted CORE products are minor and not commercially significant. Therefore, it was unlawful for Commerce to compare these groups of products as "identical" according to 19 U.S.C. § 1677(16)(A).

D. The Department's Claimed Reliance on a Practice Not to Alter a Model-Match Methodology Absent Compelling Reasons Does Not Suffice to Sustain the Remand Redetermination

The Remand Redetermination relies on a claimed "practice ... not to alter a model-match methodology developed at an earlier stage of a proceeding absent `compelling reasons' for the modification." Remand Redetermination 4. In support of this reliance, the Remand Redetermination cites various decisions of the Court of International Trade for the proposition that the practice is grounded in a reasonable construction of the antidumping statute. Id. (citing Fagersta Stainless AB v. United States, 32 CIT ___, 577 F.Supp.2d 1270 (2008); SKF USA Inc. v. United States, 31 CIT 951, 491 F.Supp.2d 1354 (2007), aff'd 537 F.3d 1373 (Fed.Cir.2008) and Mittal Steel USA, Inc. v. United States, 31 CIT 1395 (2007)). After making various factual findings, the Remand Redetermination states as follows: "[t]he Department finds on remand that the record evidence does not support the assertion that meaningful physical and commercial differences between laminated and other painted CORE products justify a departure from its previous model match methodology." Id. at 10.

The prior decisions of the Court of International Trade cited in the Remand Redetermination are not binding on the court in this case. More important, however, is that those prior decisions do not state a principle under which the court may affirm the Remand Redetermination. None of the cited cases hold that the Department is free to do what it did here, which was to refuse, under a "compelling reasons" standard, to modify a model-match methodology on a record that does not permit the Department to find that laminated and "other painted" CORE products are "identical in physical characteristics" as required by 19 U.S.C. § 1677(16)(A). Whatever the merits of its claimed practice, the Department cannot comply with the definition of "foreign like product" in § 1677(16)(A) unless it can find, based on substantial evidence, that the physical differences involved are minor and not commercially significant. The analysis undertaken by the Court of Appeals in Pesquera confirms this point. Pesquera, 266 F.3d at 1384 (stating, after affirming the Department's construction of the statute, that "[w]e must also determine whether Commerce's commercial practice determination here is supported by substantial evidence and whether Commerce has provided an adequate explanation for its determination."). In the circumstances of this case, the fact that Commerce is applying a model-match methodology that it used in prior reviews is insufficient to justify a determination that lacks such a valid finding. The cases the Remand Redetermination cites do not hold to the contrary.

In Fagersta, a factual issue critical to the challenge to the model-match methodology was whether stainless steel wire rod ("SSWR") made using a process of electro-slag refining, which plaintiff Fagersta Stainless AB defined as "a separate and significant processing stage ... [that] imparts unique material qualities, primarily superior fatigue resistance, to the finished wire rod product," Fagersta, 32 CIT at ___, 577 F.Supp.2d at 1274 (alterations in original), differed physically from other SSWR in a way that was commercially significant, id. at ___, 577 F.Supp.2d at 1278-79. Basing its decision on the record evidence, the Court of International Trade sustained the Department's finding that the physical difference was not commercially significant. Id. at ___, 577 F.Supp.2d at 1279-80. In contrast, the record in this case will not support such a finding.

SKF USA did not address the narrow issue presented by this case. In SKF USA, the Court of International Trade upheld the Department's decision to revise the long-standing "family" methodology for matching various models of ball bearings in favor of a new methodology that involved, inter alia, a difmer adjustment. See SKF USA, 31 CIT at 953-60, 491 F.Supp.2d at 1358-63. The case did not involve a decision to modify, or refuse to modify, a model-match methodology that was grounded solely in the "identical in physical characteristics" criterion of 19 U.S.C. § 1677(16)(A).

Mittal does not stand for a principle under which the court may uphold the Remand Redetermination. Although Mittal involved the model-match methodology applied to CORE products from Korea, the question presented was whether the Department acted lawfully in rejecting a request of the petitioner to require respondents to provide more specific product information in a supplemental questionnaire, on the premise that had these data been collected, petitioners possibly would have been able to uncover a compelling reason for changing the model-match methodology. Mittal, 31 CIT at 1397-98. The Court of International Trade upheld, as supported by substantial record evidence, the Department's decision not to request the additional information. Id. at 1400.

In conclusion, the judicial decisions on which the Remand Redetermination relies do not support a principle under which the decision challenged here can be sustained. The existence of a practice of declining to modify an existing model-match methodology absent "compelling reasons" does not suffice to support unlawful comparisons under 19 U.S.C. § 1677(16)(A). That the record fails to support, with substantial evidence, a finding that the merchandise being compared is in fact "identical in physical characteristics" within the meaning of that provision must be viewed, in the context of this case, as a "compelling reason" to change the model-match methodology to bring that methodology into compliance with law.

III. CONCLUSION

The Remand Redetermination must be set aside as contrary to law. The current record does not contain substantial evidence to support a determination that laminated CORE products may be compared with "other painted," i.e., non-laminated, painted CORE products under the "identical in physical characteristics" requirement of the foreign like product definition in 19 U.S.C. § 1677(16)(A). Therefore, Commerce, on the second remand, may not compare laminated CORE and non-laminated, painted CORE products under § 1677(16)(A) unless it reopens the record, conducts a sufficient investigation on the question of whether the physical differences distinguishing laminated and non-laminated, painted CORE products are minor and not commercially significant, and reaches appropriate, probative findings that are supported by substantial evidence on the reopened record. If Commerce decides not to proceed in this way, on the second remand it must modify the model-match methodology applied in the Final Results to avoid comparing laminated CORE and non-laminated, painted CORE as products "identical in physical characteristics" under § 1677(16)(A).

ORDER

Upon consideration of all proceedings and submissions herein, and upon due deliberation, it is hereby

ORDERED that the Final Results of Redetermination Pursuant to Remand (Dec. 28, 2009) ("Remand Redetermination") be, and hereby are, set aside as contrary to law; it is further

ORDERED that Commerce, upon remand, shall review and reconsider its "model-match" methodology, including its decision, made in Certain Corrosion-Resistant Carbon Steel Flat Products from the Republic of Korea: Notice of Final Results of the Thirteenth Admin. Review, 73 Fed.Reg. 14,220 (Mar. 17, 2008) ("Final Results"), and made again in the Remand Redetermination, to deny Union's request for a revision of that model-match methodology, by which Commerce compared the types of subject merchandise in plaintiff's U.S. sales with the types of foreign like products in plaintiff's sales in its home market; it is further

ORDERED that, on remand, the Department may reopen the record to re-investigate the question of whether the physical differences that have been established to exist between laminated and non-laminated, painted CORE products are minor and commercially insignificant; it is further

ORDERED that if Commerce does not reopen the record to re-investigate the question of whether the physical differences that have been established to exist between laminated and non-laminated, painted CORE products are minor and commercially insignificant, then it must alter the model-match methodology that was applied in the Final Results and Remand Redetermination so that laminated and non-laminated, painted CORE products are not compared according to 19 U.S.C. § 1677(16)(A) and recalculate Union's margin accordingly; it is further

ORDERED that the second redetermination must comply in all respects with this Opinion and Order, be supported by substantial evidence, and be otherwise in accordance with law; and it is further

ORDERED that the Department shall have ninety (90) days from the date of this Opinion and Order to file its second redetermination upon remand in this proceeding, that plaintiff and defendant-intervenors shall have thirty (30) days from the filing of the second redetermination upon remand to file comments thereon with the court, and that defendant shall have fifteen (15) days thereafter to file any reply to such comments.