Dillinger France S.A. v. United States

981 F.3d 1318

• Court: Court of Appeals for the Federal Circuit
• Decided: December 3, 2020
• Docket: 19-2395
• Status: Published

Opinion

Case: 19-2395 Document: 93 Page: 1 Filed: 12/03/2020

United States Court of Appeals for the Federal Circuit ______________________

DILLINGER FRANCE S.A., Plaintiff-Appellant

v.

UNITED STATES, SSAB ENTERPRISES LLC, NUCOR CORPORATION, Defendants-Appellees ______________________

2019-2395 ______________________

Appeal from the United States Court of International Trade in No. 1:17-cv-00159-GSK, Judge Gary S. Katzmann. ______________________

Decided: December 3, 2020 ______________________

MARC EDWARD MONTALBINE, Dekieffer & Horgan, PLLC, Washington, DC, argued for plaintiff-appellant. Also represented by JAMES KEVIN HORGAN, GREGORY S. MENEGAZ, ALEXANDRA H. SALZMAN.

KELLY A. KRYSTYNIAK, Commercial Litigation Branch, Civil Division, United States Department of Justice, Wash- ington, DC, argued for defendant-appellee United States. Also represented by JEFFREY B. CLARK, JEANNE DAVIDSON, TARA K. HOGAN; AYAT MUJAIS, Office of the Chief Counsel for Trade Enforcement and Compliance, United States Case: 19-2395 Document: 93 Page: 2 Filed: 12/03/2020

Department of Commerce, Washington, DC.

CYNTHIA CRISTINA GALVEZ, Wiley Rein, LLP, Washing- ton, DC, argued for defendant-appellee Nucor Corporation. Also represented by ALAN H. PRICE, STEPHANIE MANAKER BELL, TESSA V. CAPELOTO, MAUREEN E. THORSON, CHRISTOPHER B. WELD.

ROGER BRIAN SCHAGRIN, Schagrin Associates, Wash- ington, DC, for defendant-appellee SSAB Enterprises LLC. Also represented by NICHOLAS J. BIRCH, CHRISTOPHER CLOUTIER, GEERT M. DE PREST, ELIZABETH DRAKE, WILLIAM ALFRED FENNELL, PAUL WRIGHT JAMESON, LUKE A. MEISNER, KELSEY RULE. ______________________

Before NEWMAN, DYK, and HUGHES, Circuit Judges. DYK, Circuit Judge. Defendant Dillinger France S.A. (“Dillinger”) appeals a decision of the United States Court of International Trade (“Trade Court”). That decision affirmed the final anti- dumping determination of the U.S. Department of Com- merce (“Commerce”) for certain carbon and alloy steel cut- to-length plate from France. We affirm in part, vacate in part, and remand. BACKGROUND “Dumping occurs when a foreign firm sells a product in the United States at a price lower than the product’s nor- mal value.” Home Prods. Int’l, Inc. v. United States, 633 F.3d 1369, 1372 (Fed. Cir. 2011). Commerce is required to impose antidumping duties on imported merchandise that is being sold, or is likely to be sold, in the United States at less than fair value to the detriment of a domestic industry. 19 U.S.C. § 1673. Case: 19-2395 Document: 93 Page: 3 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 3

On April 28, 2016, Commerce initiated an antidumping duty investigation into certain carbon and alloy steel cut- to-length plate from France. Commerce chose Dillinger, a European producer of cut-to-length plate, as one of the mandatory importer respondents. Commerce assigned Dillinger a 6.15% antidumping margin. See Certain Carbon and Alloy Steel Cut-To-Length Plate from France, 82 Fed. Reg. 24,096, 24,098 (Dep’t of Commerce May 25, 2017). Dillinger appealed to the Trade Court, which initially sustained most of Commerce’s deter- mination but remanded to Commerce issues that are not involved in this appeal. The Trade Court then sustained Commerce’s remand results and the 6.15 percent duty. Dil- linger appeals the Trade Court’s judgment, contending that Commerce erred in the antidumping determination. We have jurisdiction under 28 U.S.C. § 1295(a)(5). DISCUSSION We review the Trade Court’s decision to sustain Com- merce’s final results and remand redeterminations de novo. See U.S. Steel Corp. v. United States, 621 F.3d 1351, 1357 (Fed. Cir. 2010). We will affirm Commerce unless its decision is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B)(i). I Dillinger raises three issues on appeal. We first ad- dress Dillinger’s argument that, in calculating normal value, Commerce improperly allocated costs between Dil- linger’s non-prime and prime products based on Dillinger’s books and records, which allocate cost based on likely sell- ing price rather than actual cost. 1 Because Dillinger’s

1 It is unclear from Commerce’s final determination and brief whether Commerce’s calculation of normal value Case: 19-2395 Document: 93 Page: 4 Filed: 12/03/2020

books and records did not reasonably reflect the costs asso- ciated with the production and sale of the merchandise as required by 19 U.S.C. § 1677b(f), we vacate and remand for further proceedings on this issue. Dillinger sells plates designated as prime and non- prime. Non-prime plates are plates that are rejected after the production process for not meeting the standards for prime plate. Prime plate is sold with a warranty, whereas non-prime plate is not and thus cannot be used in applica- tions that require a warranty. In reporting costs to Com- merce, Dillinger reported the cost of non-prime plate as equal to the average actual cost of all plate because, accord- ing to Dillinger, “non-prime plate undergoes the same pro- duction process as prime plate and . . . is not less costly to produce simply because it cannot be sold at full price.” J.A. 1346. Commerce did not dispute that prime and non-prime plate undergo the same production process, but Commerce noted that Dillinger’s accounting system uses a different approach, valuing “non-prime plate at the likely selling price based on current market conditions and uses this amount to offset the cost of prime plates.” J.A. 1347. Com- merce accordingly adjusted Dillinger’s reported costs for non-prime plate “to reflect the sales values recorded in [Dil- linger’s] normal books and records” and allocated the dif- ference to the costs for Dillinger’s prime plate. Id. at 968,

involved determining constructed value (determining the sum of “the cost of materials and fabrication or other pro- cessing of any kind employed in producing the merchan- dise” and other factors under 19 U.S.C. § 1677b(e)), or involved determining cost of production so as to exclude home market sales made below cost of production under § 1677b(b)(3). In either event, § 1677b(f) applies, and the alleged errors would affect either calculation. See id. § 1677b(f). Case: 19-2395 Document: 93 Page: 5 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 5

1347. In doing so, Commerce reduced the cost of non-prime plate and allocated a greater portion of cost to prime plate based on the selling price of non-prime plate. Dillinger ar- gues that Commerce’s reliance on Dillinger’s books and rec- ords was improper because the books and records were not based on the costs associated with the production of its products. The applicable statutory provision, 19 U.S.C. § 1677b(f)(1)(A), provides that “[f]or purposes of subsec- tions (b) [sales at less than cost of production] and (e) [con- structed value] . . . , [c]osts shall normally be calculated based on the records of the exporter or producer of the mer- chandise, if such records are kept in accordance with the generally accepted accounting principles [(“GAAP”)] of the exporting country (or the producing country, where appro- priate) and reasonably reflect the costs associated with the production and sale of the merchandise.” Id. (emphasis added). Section 1677b(f)(1)(A) thus requires “that reported costs must ‘normally’ be used” only if (1) “they are ‘based on the records . . . kept in accordance with the [GAAP]’” and (2) “‘reasonably reflect’ the costs of producing and sell- ing the merchandise.” Thai Plastic Bags Indus. Co. v. United States, 746 F.3d 1358, 1365 (Fed. Cir. 2014) (quot- ing 19 U.S.C. § 1677b(f)(1)(A)). The dual nature of the test seems apparent from the face of the statute and is clear as well from our prior deci- sions and the legislative history. Before § 1677b(f), our case law had established that, “[a]s a general rule, an agency may either accept financial records kept according to [GAAP] in the country of exportation, or reject the rec- ords if accepting them would distort the company’s true costs.” Thai Pineapple Pub. Co. v. United States, 187 F.3d 1362, 1366 (Fed. Cir. 1999)). In IPSCO, Inc. v. United States, 965 F.2d 1056 (Fed. Cir. 1992), we held a method that “calculat[ed] costs for both limited-service and prime products on the basis of Case: 19-2395 Document: 93 Page: 6 Filed: 12/03/2020

their relative prices” to be “an unreasonable circular meth- odology” because it “contravened the express requirements of the statute which set forth the cost of production as an independent standard for fair value.” Id. at 1061; see also id. at 1060 (“The legislative history confirms the statute’s unambiguous intent to provide cost of production as an in- dependent yardstick for deciding whether home and export sales prices are suitable for fair value comparisons.”). We relied on section 1677b(e), the provision that “expressly co- vers actual production costs,” for computing constructed value, and section 1677b(b), which “disregards, under spec- ified circumstances, home or export market sales at less than the cost of production.” Id. at 1059 (citing 19 U.S.C. § 1677b(b), (e) (1988)). Here, there is no dispute that Com- merce relied on the likely selling price of non-prime plate in its determination of cost. Thus, if IPSCO governs, Com- merce’s reliance on Dillinger’s books and records was im- permissible. Commerce argues that IPSCO should not govern be- cause the Tariff Act was amended to add § 1677b(f). When Congress added § 1677b(f), Congress did not repeal §§ 1677b(b) or (e), the sections we relied on in IPSCO, which still require determination of “the cost of materials and fabrication or other processing of any kind,” id. § 1677b(e), 2 and there is no indication that Congress in- tended for the addition of section 1677b(f) to overrule IPSCO. “Section 224 of [the Uruguay Round Agreements Act] add[ed] new section 773(f) to the [Tariff] Act to

2 Subsection (b) at the time of our decision in IPSCO required determination of “cost of producing the merchan- dise,” IPSCO, 965 F.2d at 1060 (quoting 19 U.S.C. § 1677b(b) (1988)), and has since been amended to require determination of “the cost of materials and of fab- rication or other processing of any kind.” 19 U.S.C. § 1677b(b)(3)(A). Case: 19-2395 Document: 93 Page: 7 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 7

incorporate the provisions of the [Antidumping Agree- ment 3] regarding the calculation of costs. In addition, sec- tion 773(f) harmonize[d] the methods of calculating cost for purposes of examining sales below cost and determining constructed value.” H.R. Rep. No. 103-826, pt. 1, at 91 (1994). The legislative history indicates Congress’s clear intent for Commerce to “continue its current practice of cal- culating costs,” id., and that such costs should “accurately reflect the resources actually used in the production of the merchandise in question,” S. Rep. No. 103-412, at 75 (1994). In codifying this rule, Congress noted that “[u]nder [then-]existing U.S. law and practice, Commerce normally calculate[d] costs on the basis of records kept by the ex- porter or producer of the merchandise, provided such rec- ords [were] kept in accordance with [GAAP] of the exporting (or producing) country and reasonably reflect[ed] the costs associated with the production and sale of the merchandise” and that “[u]nder new section [1677b(f)], Commerce [would] continue its current practice.” H.R. Rep. No. 103-826, pt. 1, at 90–91. Congress also concluded that “[c]osts shall be allocated using a method that reasonably reflects and accurately captures all of the actual costs incurred in producing and selling the product under investigation or review.” State- ment of Administrative Action (“SAA”), H.R. Rep. 103-316 (1994), as reprinted in 1994 U.S.C.C.A.N. 4040, 4172. Con- gress “expect[ed] [Commerce], in determining whether a producer’s or exporter’s records reasonably reflect the costs associated with the production and sale of the product in

3 The Antidumping Agreement means the Agree- ment on Implementation of Article VI of the General Agree- ment on Tariffs and Trade 1994. Uruguay Round Agreements Act §§ 121(9), 101(d)(7), PL 103–465, Decem- ber 8, 1994, 108 Stat 4809. Case: 19-2395 Document: 93 Page: 8 Filed: 12/03/2020

question, to examine the recorded production costs with a view to determining as closely as possible the costs that most accurately reflect the resources actually used in the production of the merchandise in question.” S. Rep. No. 103-412, at 75. Thus, the legislative history of section 1677b(f), consistent with its plain meaning, indicates Con- gress intended that Commerce rely on a producer’s or ex- porter’s books and records if they are in accordance with GAAP and reasonably reflect the costs of production. Nonetheless, Commerce argues that our decision in Thai Pineapple, decided after the Tariff Act was amended to include section 1677b(f) (but deciding issues raised un- der the pre-amended Tariff Act), supports Commerce’s po- sition here. In Thai Pineapple, in determining costs of production and constructed value, Commerce relied on a producer’s allocation methodology for “material cost” for pineapple fruit, which the producer used to make canned pineapple products and juice products. 187 F.3d at 1366. The producer’s books and records “allocate[ed] a range of 82 to 91% of the pineapple fruit costs to canned pineapple fruit production, and 9 to 18% to production of juice prod- ucts.” Id. “Commerce’s allocation of the cost of the raw pineapple fruit between canned pineapple fruit and other products was not based on the selling price or output value of these products.” Id. at 1369. “Thus, unlike [IPSCO], the selling price of the [subject] products was not a factor in determining the cost of raw material component in Commerce’s calculation of [costs of production and constructed value].” Id. Instead, “Com- merce’s methodology reflected the raw material allocations of [the producer] as shown by their books and records.” Id. We held that, “[t]o the extent that the records of [the pro- ducer] reasonably reflect the costs of production, Com- merce may rely upon them.” Id. at 1367. The government relies on footnote 5 of Thai Pineapple, but that part of the decision simply “note[d] that this rule is now codified in 19 U.S.C. § 1677b(f)(1)(A) (1996).” Id. at 1366 n.5. Thai Case: 19-2395 Document: 93 Page: 9 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 9

Pineapple is not inconsistent with IPSCO as to the deter- mination of production costs. 4 The government also relies on PSC VSMPO-Avisma Corp. v. United States, 688 F.3d 751 (Fed. Cir. 2012), but that case does not support Commerce’s position. In PSC, we affirmed Commerce’s method of basing the costs of chlo- rine “upon what [the exporter] would have to spend to pur- chase the chlorine necessary for its titanium production process.” Id. at 757. Thus, PSC concerned the use of pur- chase price to determine cost rather than using likely sell- ing price of the end product to allocate costs as here. There is no dispute that Dillinger’s records were kept in accordance with GAAP. However, Dillinger’s records that Commerce relied on for the cost of non-prime and prime plate were based on “likely selling price” rather than costs of producing and selling the merchandise. J.A. 1347. Because Dillinger’s books and records were based on “likely selling price” rather than cost of production, id., Commerce erred in relying on them. A remand is required for Com- merce to determine the actual costs of prime and non-prime products.

4 To the extent that Thai Pineapple disagreed with IPSCO, it was to distinguish Commerce’s use of a weight- based methodology in IPSCO. In IPSCO, we sustained Commerce’s weight-based allocation because “[t]he steel pipe was manufactured from the same raw material and underwent one production process,” but in Thai Pineapple, we found that “pineapple fruit [was] not a homogeneous raw material like the raw material used to make the pipe in [IPSCO], and the production process [was] entirely dif- ferent for the various pineapple products produced.” Thai Pineapple, 187 F.3d at1369. Accordingly, we found that Commerce’s determination not to use a weight-based meth- odology was reasonable. Id. Case: 19-2395 Document: 93 Page: 10 Filed: 12/03/2020

II We next consider Dillinger’s argument that Com- merce’s use of the average-to-transaction method in deter- mining the dumping margin was improper. The dumping margin is the “amount by which the normal value exceeds the export price or constructed export price of the subject merchandise.” 19 U.S.C. § 1677(35)(A). To determine the dumping margin, Commerce uses one of three methods: the average-to-average method, the transaction-to-transaction method, and the average-to-transaction method. 5 Here, Commerce’s decision used the average-to-transaction

5 The average-to-average method compares the weighted average of the normal values to the weighted av- erage of the export prices. 19 U.S.C. § 1677f-1(d)(1)(A)(i). Commerce “will use the average-to-average method unless [Commerce] determines another method is appropriate in a particular case.” 19 C.F.R. § 351.414(c)(1) (2020). The transaction-to-transaction method compares the normal values of individual transactions to the export prices of individual transactions. 19 U.S.C. § 1677f- 1(d)(1)(A)(ii). Commerce “will use the transaction-to-trans- action method only in unusual situations, such as when there are very few sales of subject merchandise and the merchandise sold in each market is identical or very simi- lar or is custom-made.” 19 C.F.R. § 351.414(c)(2). The average-to-transaction method compares weighted average of the normal values to the export prices of indi- vidual transactions for comparable merchandise. 19 U.S.C. § 1677f-1(d)(1)(B). Commerce may use the average- to-transaction method if “there is a pattern of export prices . . . for comparable merchandise that differ significantly among purchasers, regions, or periods of time,” and Com- merce “explains why such differences cannot be taken into account using [the average-to-average or transaction-to- transaction methods].” Id. § 1677f-1(d)(1)(B)(i)–(ii). Case: 19-2395 Document: 93 Page: 11 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 11

method, which may be used if “there is a pattern of export prices . . . for comparable merchandise that differ signifi- cantly among purchasers, regions, or periods of time.” 19 U.S.C. § 1677f-1(d)(1)(B)(i). This provision addresses situations “where targeted dumping may be occurring.” SAA, 1994 U.S.C.C.A.N. at 4178. Targeted dumping occurs where “an exporter may sell at a dumped price to particular customers or regions, while selling at higher prices to other customers or regions.” Id. at 4177–78. To determine a pattern of export prices for comparable merchandise that differ significantly among purchasers, regions, or periods of time, Commerce used the Cohen’s d test. The Cohen’s d coefficient is a “generally recognized statistical measure” of the extent of the difference between the weighted-average price of a test group and the weighted-average price of a comparison group. J.A. 958. Here, the test groups were export prices for a purchaser, region, and time period, and the corresponding comparison groups were all other export prices (i.e., the export sales to all other purchasers, regions, or time periods). If the Co- hen’s d coefficient is equal to or greater than 0.8, then Com- merce considers the difference between the average prices of the test group and the average prices of the comparison group to be significant, and thus the test group passes the Cohen’s d test. Commerce next applied the “ratio test,” in which Com- merce calculated the sales value for all test groups that passed the Cohen’s d test. “If the value of sales to purchas- ers, regions, and time periods that pass the Cohen’s d test account for 66 percent or more of the value of total sales, then the identified pattern of prices that differ significantly supports the consideration of the application of the aver- age-to-transaction method to all sales as an alternative to the average-to-average method.” J.A. 959. In its final determination, Commerce determined that 95.78 percent of Dillinger’s U.S. sales passed the Cohen’s d Case: 19-2395 Document: 93 Page: 12 Filed: 12/03/2020

test and that this “confirm[ed] the existence of a pattern of prices that differ[ed] significantly among purchasers, re- gions, or time periods.” Id. at 1306. Commerce accordingly used the average-to-transaction method for all U.S. sales to calculate the dumping margin. Dillinger raises two challenges to Commerce’s determi- nation of a pattern. First, Dillinger contends that Com- merce’s use of the Cohen’s d test and the ratio test to determine a pattern “ignor[ed] the word ‘pattern’ in section 1677f-1(d)(1)(B)(i).” Appellant’s Br. 15. Dillinger appears to argue that Commerce’s ratio test improperly aggregated sales across categories (purchasers, regions, or time peri- ods) and that comparing aggregated sales across categories cannot be done to establish a pattern. Id. at 21 (stating Commerce’s methodology “does not analyze the categories of purchasers, regions and time periods individually”). Such aggregation is not inconsistent with the statute, which requires that Commerce determine that there is “a pattern of export prices . . . for comparable merchandise that differ significantly among purchasers, regions, or pe- riods of time.” 19 U.S.C. § 1677f-1(d)(1)(B)(i). The statute is silent as to how Commerce must determine a “pattern.” See id. §§ 1677, 1677f-1. “[I]f the statute is silent or ambig- uous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permis- sible construction of the statute.” Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843 (1984). We find that Commerce’s interpretation of pattern was reason- able. Dillinger relies on a determination from the World Trade Organization (“WTO”), which reached the opposite conclusion in interpreting Article 2.4.2 of the Anti-Dump- ing Agreement. Appellate Body Report, United States – Antidumping and Countervailing Measures on Large Resi- dential Washers from Korea, WTO Doc. WT/DS464/AB/R, at 25–31 (adopted Sep. 7, 2016). The WTO Appellate Body Case: 19-2395 Document: 93 Page: 13 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 13

determined that Commerce’s methodology of using the Co- hen’s d test and the ratio test “is inconsistent” with deter- mining “a pattern of export prices which differ significantly among different purchasers, regions, or time periods” be- cause the methodology “aggregates prices found to differ among different purchasers, among different regions, and among different time periods for the purposes of identify- ing a single pattern.” Id. at 25, 31. The WTO “oversee[s] the application of the various WTO agreements and serve[s] as the framework for mem- ber governments to conduct their trade relations under those agreements.” SAA, 1994 U.S.C.C.A.N. at 4043. “WTO decisions are ‘not binding on the United States, much less this court.’” Corus Staal BV v. Dep’t of Com- merce, 395 F.3d 1343, 1348 (Fed. Cir. 2005) (quoting Tim- ken Co. v. United States, 354 F.3d 1334, 1344 (Fed. Cir. 2004)). 6 Dillinger’s other arguments regarding the interpreta- tion of “pattern” are not adequately developed, and we de- cline to consider them. See Agile Def., Inc. v. United States, 959 F.3d 1379, 1384 n.* (Fed. Cir. 2020) (because party “fail[ed] to adequately develop [an] argument,” the court “decline[d] to consider it on appeal”); SmithKline Beecham Corp. v. Apotex Corp., 439 F.3d 1312, 1320 (Fed. Cir. 2006) (declining to consider argument that “[did] not amount to a developed argument”).

6 Dillinger also argues that Commerce’s application of the Cohen’s d test applied “an irrebuttable presumption” that a 0.8 Cohen’s d coefficient indicates that a price differ- ence is significant. Appellant’s Br. 24. The record does not indicate that Commerce’s use of the Cohen’s d test or its thresholds is irrebuttable. To the contrary, Commerce con- sidered Dillinger’s objections to its methodology and pro- vided its reasons for rejecting them. Case: 19-2395 Document: 93 Page: 14 Filed: 12/03/2020

Second, Dillinger argues that Commerce’s use of the Cohen’s d test to determine a pattern among export prices was not in accordance with the law because Dillinger’s products are custom-made. Thus, in Dillinger’s view, Com- merce was not permitted to use the average-to-transaction test and instead should have used the default average-to- average test. 7 But there is nothing in § 1677f-1 or the reg- ulations promulgated thereunder that requires Commerce to consider custom products differently when determining whether “there is a pattern of export prices . . . that differ significantly among purchasers, regions, or periods of time” so long as such comparison is made between “comparable merchandise.” 19 U.S.C. § 1677f-1(d)(1)(B). Here, Dillinger has not shown how Commerce failed to use “comparable merchandise.” “Comparable merchan- dise” was defined by product control numbers (“CON- NUMs”), which have certain “physical characteristics” that were subject to notification and comment during Com- merce’s investigation. J.A. 958, 1310. In making its com- parison, Commerce rejected Dillinger’s assertion that “its made-to-order products are inferably so unique and em- brace such a wide range of grades within a given [CONNUM] that any comparison of U.S. prices on a CONNUM basis must take into account these inter- CONNUM variations.” Id. at 1309–10. We see no error in Commerce’s determination.

7 Dillinger does not argue on appeal that Commerce should have used the transaction-to-transaction method, even though the regulations state that Commerce “will use the transaction-to-transaction method only in unusual sit- uations, such as when there are very few sales of subject merchandise and the merchandise sold in each market is identical or very similar or is custom-made.” 19 C.F.R. § 351.414(c)(2) (emphasis added). Case: 19-2395 Document: 93 Page: 15 Filed: 12/03/2020

DILLINGER FRANCE S.A. v. UNITED STATES 15

III Finally, we consider Dillinger’s argument that Com- merce erred in determining that Dillinger’s factory sales and sales from its affiliated service centers constituted a single level of trade in France and thus concluding that a level of trade adjustment was not warranted. Commerce is required to establish normal value “to the extent practicable, at the same level of trade as the export price.” 19 U.S.C. § 1677b(a)(1)(B)(i). If Commerce is una- ble to find sales in the foreign market at the same level of trade as the sales in the United States, normal value shall be “increased or decreased to make due allowance for any difference (or lack thereof) between the export price . . . and [normal value] that is shown to be wholly or partly due to a difference in level of trade.” Id. § 1677b(a)(7)(A). “[T]he level of trade adjustment is designed to ensure that the normal value and U.S. price are being compared . . . at the same level of trade, that is, at the same marketing stage in the chain of distribution that begins with the manufac- turer.” Micron Tech., Inc. v. United States, 243 F.3d 1301, 1314 (Fed. Cir. 2001). Commerce will grant a level of trade adjustment where “there is a difference between the actual functions performed by the sellers at the different levels of trade in the two markets.” SAA, 1994 U.S.C.C.A.N. at 4168. Dillinger makes sales directly from its factories to end users and distributors and from affiliated service centers to downstream customers. Dillinger argues that Com- merce erred in determining that inventory maintenance performed on service center sales did not require a finding of a separate level of trade. “Substantial differences in sell- ing activities are a necessary, but not sufficient, condition for determining that there is a difference in the stage of marketing.” 19 C.F.R. § 351.412(c)(2). Commerce deter- mined that inventory maintenance alone did not make a substantial difference between the selling activities Case: 19-2395 Document: 93 Page: 16 Filed: 12/03/2020

commonly performed by Dillinger’s factories and service centers, and we find this determination to be supported by substantial evidence and in accordance with law. In addition to the selling functions performed by its fac- tories, Dillinger’s affiliated service centers also perform service center functions such as cutting, sawing, drilling, and bending. Dillinger argues that Commerce “improperly ignored processing activities such as cutting and sawing of plate into smaller sizes for resale.” Appellant’s Br. 51. Commerce agreed that Dillinger’s service centers per- formed service center functions such as cutting and sawing “to make downstream sales.” J.A. 1330. It also determined that “these items (i.e., cutting, sawing, drilling[,] and[] bending) are not selling functions . . . contained in the list provided in [Commerce’s] standard section A question- naire. . . . Instead, . . . these items are performed in con- nection with the further processing of the merchandise, which are part of the cost to produce the downstream prod- uct.” Id. We see no error in Commerce’s refusal to consider these processing activities to be selling functions. CONCLUSION We have considered the parties’ remaining arguments and find them unpersuasive. We vacate the Trade Court’s judgment sustaining Commerce’s decision to rely on Dil- linger’s books and records to determine cost. We affirm the Trade Court’s judgment sustaining Commerce’s determi- nations of the pattern requirement of the average-to-trans- action method and level of trade. We remand the case to the Trade Court for Commerce to recalculate the dumping margin consistent with this opinion. AFFIRMED IN PART, VACATED IN PART, AND REMANDED COSTS No costs.