Peer Bearing Company-Changshan v. United States

OPINION AND ORDER STANCEU, Judge. Plaintiff Peer Bearing Company-Changshan (“CPZ”) contests a final determination (“Final Results”) of the International Trade Administration, United States Department of Commerce (“Commerce” or the “Department”), in the twentieth periodic administrative review of an antidumping duty order on tapered roller bearings and parts thereof (“subject merchandise”) from the People’s Republic of China (“PRC” or “China”). Compl. ¶1; Tapered Roller Bearings & Parts Thereof, Finished & Unfinished, From the People’s Republic of China: Final Results of Antidumping Duty Admin. Review, 74 Fed.Reg. 3,987 (Jan. 22, 2009) (“Final Results ”). 1 Corn *1357 merce assigned CPZ a 92.84% dumping margin in the Final Results, based on U.S. prices for subject merchandise made by CPZ that Commerce, in the absence of record data on sales transactions between CPZ and an unaffiliated importer in the United States, determined according to “facts otherwise available.” Final Results, 74 Fed.Reg. at 3,989; Issues & Decisions Mem., A-570-601, ARP 1-09, at 2 (Jan. 13, 2009) (Admin.R.Doc. No. 5486), available at http://ia.ita.doc.gOv/frn/summary/PRC/E 9-1219-l.pdf (“Decision Mem."). CPZ challenges this margin, arguing, inter alia, that Commerce unlawfully determined the U.S. prices according to facts otherwise available instead of opening the record to gather , data to determine actual export prices or using data already on the record to determine CPZ’s margin based on constructed export prices. Compl. ¶¶ 29-31; Pl.’s Mem. of Points & Authorities in Supp. of its Mot. for J. on the Agency R. 10-27 (“Pl.’s Mem.”). The issue of whether the U.S. prices should have been determined on an export price (“EP”) basis or a constructed export price (“CEP”) basis arose because CPZ sold subject merchandise to an unaffiliated U.S. importer that sold the merchandise to CPZ’s U.S. affiliate, Peer Bearing Company (“Peer”), which sold to unaffiliated U.S. customers. Compl. ¶¶ 14, 30. The record contained almost no data on CPZ’s sales to the U.S. importer because Commerce, contrary to the position taken by defendantintervenor The Timken Company (“Timken”) during the review, declined to make a request for these data during the review. CPZ seeks a remand instructing Commerce to “recalculate Plaintiffs antidumping duty margin using the constructed export price methodology....” Pl.’s Rule 56.2 Mot. for J. upon the Agency R. 2 (“PL’s Mot.”). CPZ also challenges several of the surrogate values that Commerce chose when determining the normal value of CPZ’s merchandise according to the methodology Commerce applies when the subject merchandise is produced in a non-market economy country such as the PRC. Compl. ¶¶ 33-37. CPZ seeks a remand instructing Commerce to recalculate the surrogate value for “steel wire input using a non-aberrational import value,” for “the steel bar input in order to remove aberrational import values,” and for the “steel scrap inputs consistent with Commerce’s practice in previous administrative reviewfs] and in the preliminary results of the review under appeal.” PL’s Mot. 2. Before the court is plaintiffs motion, made under USCIT Rule 56.2, for judgment upon the agency record, which is opposed by defendant and Timken, the petitioner in the investigation resulting in the antidumping duty order. The court concludes that Commerce, in assigning CPZ a dumping margin in the Final Results, based its U.S. price determinations on CEP starting prices, adjusted those starting prices according to a method the antidumping statute does not authorize, and incorrectly concluded that it had determined U.S. prices on an EP basis. The court further concludes that the surrogate values for steel wire rod, steel bar, and steel scrap from the production of cages were not determined according to law. The court orders a remand for correction of these errors. *1358 I. Background Commerce issued the antidumping duty-order on tapered roller bearings and parts thereof from China in 1987. Antidumping Duty Order; Tapered Roller Bearings & Parts Thereof, Finished or Unfinished, From the People’s Republic of China, 52 Fed.Reg. 22,667 (June 15, 1987). On June 1, 2007, Commerce invited parties to request an administrative review of the entries of subject merchandise made during the period June 1, 2006 to May 31, 2007 (the “period of review” or “POR”). Anti-dumping or Countervailing Duty Order, Finding, or Suspended Investigation; Opportunity To Request Admin. Review, 72 Fed.Reg. 30,542, 30,543 (June 1, 2007). On June 29, 2007, plaintiff requested that Commerce review entries of its subject merchandise. Letter from CPZ to Asst. Sec’y for Import Admin. (June 29, 2007) (Admin.R.Doc. No. 5323). On July 26, 2007, Commerce initiated the review at issue in this case. Initiation of Anti-dumping & Countervailing Duty Admin. Reviews & Request for Revocation in Part, 72 Fed.Reg. 41,057, 41,058 (July 26, 2007). In the preliminary results of the review (“Preliminary Results”), Commerce assigned CPZ a preliminary margin of 59.41% using a constructed export price methodology. Tapered Roller Bearings & Parts Thereof, Finished & Unfinished, from the People’s Republic of China: Prelim. Results of Antidumping Duty Admin. Review, 73 Fed.Reg. 41,033, 41,037, 41,039 (July 17, 2008) (“Prelim.Results”). In the Final Results, Commerce changed its position on the determination of U.S. prices of CPZ’s subject merchandise, choosing instead to calculate dumping margins on an “export price basis” that relied on “facts otherwise available.” Decision Mem. 2. The Final Results assigned CPZ a margin of 92.84%. Final Results, 74 Fed.Reg. at 3,989. A. Commerce’s Calculation of CPZ’s Dumping Margin Using Facts Otherwise Available CPZ’s response to Commerce’s anti-dumping questionnaire stated that Peer “sold the subject merchandise to the first unaffiliated U.S. customers and thus the constructed export price (‘CEP’) methodology should be used for reporting CPZ’s U.S. sales.” Pl.’s Mem. 5. CPZ reported to Commerce the prices paid by unaffiliated customers to Peer but not the prices paid by the unaffiliated U.S. importer to CPZ. Compl. ¶¶ 14, 30. Commerce did not inform CPZ that CPZ’s submission was deficient. On April 2, 2008, Commerce issued supplemental questionnaires, to which CPZ filed responses on April 29, 2008. Prelim. Results, 73 Fed.Reg. at 41,033. CPZ responded to Commerce’s request for evidence of price negotiations between CPZ and its importer by stating that “Peer issues a purchase order to [the unaffiliated importer] and copies CPZ.” Def.’s Opp’n to PL’s Mot. for J. upon the Agency R. 4 (“Def.’s Opp’n”). This purchase order, according to CPZ, contained the “agreed upon price between CPZ and the importer of record.... ” Id. In commenting on CPZ’s questionnaire responses, Timken advocated that Commerce require CPZ to report data on its sales to the importer so that Commerce could calculate CPZ’s margin on an export price basis. Letter from Timken to the Sec’y of Commerce 2-3 (Nov. 15, 2007) (Admin.R.Doc. No. 5354). Despite the supplemental questionnaire responses, and rejecting the request of Timken, Commerce declined to request that CPZ submit data on the transactions between CPZ and its unaffiliated importer. In the Preliminary Results, Commerce stated that it “calculated CEP for CPZ based on delivered prices to unaffiliated purchasers in the United States.” Prel *1359 ira. Results, 73 Fed.Reg. at 41,037 (“In accordance with Section 772(b) of the Act, we used CEP for CPZ’s sales where CPZ sold subject merchandise to its affiliated company in the United States, which in turn sold subject merchandise to unaffiliated U.S. customers.”). On January 22, 2009, Commerce published the Final Results, which stated that “[w]e have treated CPZ’s sale to the importer as the relevant sale for calculating dumping margins, and have calculated the margins on an export price basis.” Final Results, 74 Fed.Reg. at 3,988. In an issues and decisions memorandum (“Decision Memorandum”) incorporated by reference in the Final Results, Commerce stated a finding that “sales prices from CPZ to its importer for all transactions during the POR are not on the record of this review” and announced its decision, pursuant to section 776(a) of the Tariff Act of 1930 (“Tariff Act” or the “Act”), 19 U.S.C. § 1677e(a) (2006), to use facts otherwise available to calculate the U.S. prices. Decision Mem. 2; Final Results, 74 Fed.Reg. at 3,988. Noting the confidential nature of the matter, the Decision Memorandum referred to more detailed discussion in an internal memorandum to the file, to which Commerce referred as the “Analysis Memo.” Decision Mem. 2 n. 1 (citing Mem. from Int’l Trade Compliance Analyst, AD/CVD Operations Office 8, to the file (Jan. 13, 2009) (Admin.R.Doc. No. 5482) (“Analysis Mem.”)). The record contained a small number of sample purchase orders pertaining to sales transactions between CPZ and the unaffiliated importer. See Analysis Mem. 2. The sales data in the purchase orders accounted for fewer than 1% of the sales transactions between CPZ and the unaffiliated importer and only a small fraction of CPZ’s models of subject merchandise. See id. attachment 2; Pl.’s Mem. 25-26. The Analysis Memo stated that Commerce, to estimate CPZ’s export prices for all other transactions, would modify the price that CPZ reported for transactions between Peer and the ultimate customer using a factor derived from the relationship between the purchase order data and the sales data for the corresponding models contained in CPZ’s responses to Commerce’s questionnaires. Id. at 7. The Analysis Memo, the Decision Memorandum, and the signed, pre-publication version of the Final Results are dated the same day, January 13, 2009. As shown by the record documentation, that date is the date on which the administrative review was concluded. B. Commerce’s Choices of Surrogate Values for Steel Wire Rod, Steel Bar, and Steel Scrap On July 17, 2008, Commerce announced in the Preliminary Results that it would apply the non-market economy factors of production method to calculate the normal value of CPZ’s subject merchandise. 2 Prelim. Results, 73 Fed.Reg. at 41,033-34. Commerce identified several potential surrogate countries it determined to be at a similar level of economic development to that of China and chose India as the surrogate country. Id. at 41,034. In the Preliminary Results and Final Results, Commerce used average entered values in import data for India to value bearing-quality alloy steel wire rod, bearing-quality alloy steel bar, scrap from production of rollers and rings, and scrap from production of cages. 3 See Mem. from Case Ana *1360 lyst to the file 3 (June 30, 2008) (Admin.R.Doc. No. 5408) (“Prelim. Factor Valuation Mem. ”); Mem. from Case Analysts to file attachment 1 (Jan. 13, 2009) (Admin.R.Doc. No. 5484) (¡‘Final Factor Valuation Mem.”). The source of Commerce’s Indian import data was the online World Trade Atlas (“WTA”), a source that uses data provided by national governments. Prelim. Results, 73 Fed.Reg. at 41,037-38. Plaintiff initiated this action on February 4, 2009. Summons; Compl. On July 13, 2009, CPZ moved for judgment upon the agency record pursuant to USCIT Rule 56.2. Pl.’s Mot. Defendant and defendant-intervenor oppose this motion. Def.’s Opp’n; Def.-Intervenor The Timken Co.’s Opp’n to Mot. for J. on the Agency R. of PI. Peer Bearing Co.-Changshan (“Def.Intervenor’s Opp’n”). The court held oral argument on March 4, 2010. II. Discussion The court exercises subject matter jurisdiction under section 201 of the Customs Courts Act of 1980, 28 U.S.C. § 1581(c) (2006). Under the applicable standard of review, the court must hold unlawful any determination, finding, or conclusion found to be unsupported by substantial evidence on the record or otherwise not in accordance with law. Tariff Act, § 516A(b)(l)(B)(i), 19 U.S.C. § 1516a(b)(l)(B)(i). A Commerce’s Determination of U.S. Prices Was Contrary to Law CPZ claims that Commerce’s determination of CPZ’s dumping margin without using constructed export prices was unlawful. Pl.’s Mem. 10-27. According to CPZ, Commerce, having chosen to calculate the U.S. prices of CPZ’s subject merchandise on an export price basis, “was obligated to solicit additional information from CPZ to supplement/correct the record” instead of using facts otherwise available. Id. at 11. CPZ argues that “the adjusted prices that Commerce used as a proxy for CPZ’s U.S. sales prices were based on an arbitrary methodology, for which Commerce offered no explanation.” Id. at 27. Commerce stated in the Final Results that “[w]e have treated CPZ’s sale to the importer as the relevant sale for calculating dumping margins, and have calculated the margins on an export price basis.” Final Results, 74 Fed.Reg. at 3,988. The Final Results do not contain further discussion on the Department’s calculation of U.S. prices. In a single paragraph directed to the issue, the Decision Memorandum describes two findings, the first of which appears to be both a finding and a conclusion: “[w]e agree with Timken, and find that the relevant U.S. sales prices for purposes of calculating CPZ’s dumping margin are CPZ’s sales to the importer, rather than the sales prices from CPZ’s U.S. affiliate to unaffiliated customers.” Decision Mem. 2. Commerce then states as a finding that “[t]hough Timken in its case brief requested that the Department collect the necessary sales data, the sales prices from CPZ to its importer for all transactions during the POR are not on the record of this review.” Id. The latter finding is correct and is, in fact, an understatement: due to the Department’s having rejected Timken’s comment that Commerce should request export price data from CPZ, the record contained price data on fewer than 1% of the sales transactions between CPZ *1361 and the unaffiliated importer, covering only a small fraction of CPZ’s models of subject merchandise. See Analysis Mem. attachment 2; PL’s Mem. 25-26. Invoking the “facts otherwise available” authority provided in section 776(a)(1) of the Tariff Act, 19 U.S.C. § 1677e(a)(l), the paragraph in the Decision Memorandum then sets forth the determination at issue here: “[a]s facts available, in order to calculate CPZ’s U.S. prices on an EP basis, rather than a constructed export price basis, we will derive a ratio, based on the sales documentation on the record demonstrating the prices between parties, to apply to CPZ’s reported U.S. sales.” Decision Mem. 2-3 (emphasis added) (footnote omitted). The court concludes that this determination is contrary to law. The calculation of CPZ’s U.S. prices was not, as Commerce characterized it in the Final Results and Decision Memorandum, performed on an export price basis. See id. at 2. Commerce calculated U.S. prices using the CEP starting prices (the prices by which the merchandise was sold by Peer to the unaffiliated customers), rather than EP starting prices (which, with very few exceptions, it did not have), and proceeded to apply its own method of adjusting the CEP starting prices, a method at variance with the method of adjusting CEP starting prices that the Act requires. Commerce thereby determined the U.S. prices by a method conforming neither to the statutory requirements for determining export price nor to those for determining constructed export price. That method also failed to fulfill the statutory purpose of the review, which was to determine whether, and to what extent, the U.S. prices at which the subject merchandise was sold were below normal value. Because Commerce chose its “facts otherwise available” on the erroneous premise that in applying those facts it was determining U.S. prices on an export price basis, and because the U.S. “prices” Commerce determined had no meaningful relationship to the prices at which the subject merchandise actually was sold, the determination of U.S. prices in the review must be set aside as contrary to law. The court addresses these points in further detail below. 1. Contrary to its Characterization in the Final Results and in the Decision Memorandum, Commerce Did Not Determine U.S. Prices on an “Export Price Basis ” Commerce, until the conclusion of the review, analyzed the U.S. sales of CPZ’s merchandise according to a constructed export price methodology and collected questionnaire data accordingly. See Prelim. Results, 73 Fed.Reg. at 41,037. The record shows that the Department decided to change its method of determining the U.S. prices at or near the end of the review (apparently on the last day of the review). Final Results, 74 Fed.Reg. at 3,989; Decision Mem. 2. Commerce then sought to determine U.S. prices on an export price basis, using “facts otherwise available,” and concluded, contrary to law, that it had done so. Decision Mem. 2. Commerce’s method of adjusting the CEP starting prices in pursuit of an EP methodology is shown in an attachment to the Analysis Memo. Analysis Mem. attachment 2. Commerce adjusted all of the CEP starting prices in the review by a percentage that substantially reduced those prices. It derived that percentage by taking a simple average of two other percentages. Commerce derived each of those other percentages from a small number of sales of a single product code, i.e., an individual bearing part number. For each of the two part numbers, Commerce divided the price for that part number, as shown on an invoice in the sale from CPZ to the unaffiliated importer, by a simple average of the prices for the same part *1362 number that are shown on an invoice in the sale from Peer to the unaffiliated customer. Id. at 7. Commerce explained that “[bjecause we are no longer calculating U.S. net price on a CEP basis, we will not deduct the selling expenses related to activity incurred in the United States.” Id. at 6. Even though Commerce’s method of adjusting the CEP starting prices bore some mathematical relationship to a handful of prices in transactions between CPZ and the unaffiliated importer, the numerical results of Commerce’s resort to what it termed “facts otherwise available” in no way qualify as “export prices.” Had it actually determined U.S. price on an export price basis, Commerce would have been required by section 772(a) of the Act to use as Court No. 09-00052 Page 12 the “starting price,” 4 i.e., the price of the subject merchandise prior to the Department’s making the required adjustments, “the price at which the subject merchandise is first sold (or agreed to be sold) before the date of importation by the producer or exporter of the subject merchandise outside of the United States to an unaffiliated purchaser in the United States....” Tariff Act, § 772(a), 19 U.S.C. § 1677a(a) (emphasis added). Commerce could not do so because, except for an extremely small number of transactions, the prices at which CPZ sold the subject merchandise to the unaffiliated importer were not on the record. The absence of data necessary for determining U.S. prices on an export price basis for the vast majority of U.S. sales was the consequence of the Department’s earlier decision, made contrary to the comments of Timken, not to request from CPZ the information underlying CPZ’s sales to the unaffiliated importer. In the absence of those price data, Commerce, for almost all sales transactions in the review, determined the U.S. prices by making adjustments, of its own design, to the prices at which Peer sold the subject merchandise to unaffiliated customers in the United States. See Analysis Mem. 7 & attachment 2. These are the same prices that Commerce would have been required to use in determining constructed export prices in the Final Results had it adhered to a constructed export price methodology in the review. See 19 U.S.C. § 1677a(b). Because Commerce based its determinations of U.S. price on the CEP starting price and not the EP starting price, it was erroneous for Commerce to characterize its basis for determining the U.S. prices, and CPZ’s margin, as an “export price” basis. 2. Commerce Determined the U.S. Prices of CPZ’s Subject Merchandise According to a Method that the Act Does Not Authorize Section 751(a)(2)(A) of the Tariff Act directs generally that the dumping margins for each entry of subject merchandise in the review be determined according to the “export price” or the “constructed export price,” 19 U.S.C. § 1675(a)(2)(A), and provides, in section 772 of the Act, statutory definitions for these terms, 19 U.S.C. § 1677a. In those definitions, the statute specifies a different starting price for the export price determination than it does for the constructed export price determination and thereby prohibits Commerce from determining ex *1363 port price from the CEP starting price. See id. § 1677a(a) (identifying as the EP starting price the price at which the subject merchandise is first sold “outside of the United States” to an unaffiliated purchaser), § 1677a(b) (identifying as the CEP starting price the price at which the merchandise is first sold “in the United States” to an unaffiliated purchaser). The statute also prohibits Commerce from determining constructed export price according to a method that uses adjustments different from the ones the statute prescribes for the determination of CEP. See id. § 1677a(d) (providing for the deduction of selling expenses incurred and profit realized on the sale of the subject merchandise in the United States). Commerce, therefore, determined the U.S. prices by a method that did not conform to the statute. 3. Commerce Chose “Facts Otherwise Available” that Were Impermissible under the Act The Department’s U.S. price calculations were not only inconsistent with any method of determining U.S. price authorized by the Act, they were also decidedly inferior to U.S. prices determined according to CEP. If the court were to assume, arguendo, that it was permissible for the Department to invoke its authority to use facts otherwise available in this case (an issue addressed later in this Opinion and Order), it still would conclude that the Department’s choice of facts otherwise available in this case was impermissible. Commerce’s conclusion, made at or near the completion of the review, that it must determine U.S. prices on an EP basis required it, in effect, to do the impossible: base its U.S. price determinations on starting prices that, with a few insignificant exceptions, it did not have. In this case, the only statutorily permissible “facts otherwise available” on the record that may be used to determine U.S. prices are the CEP starting prices, adjusted by the method the statute requires. As the name implies, “constructed export price” is the method the statute uses to “construct,” i.e., approximate as closely as possible, an export price when an affiliation between the buyer and foreign seller precludes the use, as an EP starting price, of the price in the transaction by which the subject merchandise was sold outside of the United States, prior to importation. The Statement of Administrative Action (“SAA”) accompanying passage of the Uruguay Round Agreements Act confirms this essential point. The SAA explains that profit is deducted from the CEP starting price “consistent with the language of the Agreement,” i.e., the Uruguay Round’s Agreement on the Implementation of Article VI of the General Agreement on Tariffs and Trade 1994, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1A, 1868 U.N.T.S. 201 (1994), “which reflects that constructed export price is now calculated to be, as closely as possible, a price corresponding to an export price between non-affiliated exporters and importers.” Uruguay Round Agreements Act, Statement of Administrative Action, H.R. Doc. No. 103-316, Vol. 1, at 823 (1994), reprinted in 1994 U.S.C.C.A.N. 4040, 4163 (“SAA ”) (emphasis added). In essence, Commerce attempted to construct an export price using, as facts otherwise available, the same starting price as, but a different method of adjustment than, the method the statute requires for determining CEP. In contrast to the statutory CEP adjustments, Commerce’s adjustments to the CEP starting prices, even if they might be characterized as based on “facts otherwise available,” were not made based on information pertaining to the actual sales of subject merchandise that Peer made to the unaffiliated customers. Nor did these adjustments succeed in convert *1364 ing the CEP starting prices to anything resembling actual export prices. Commerce’s adjustment method was based on a simple average of two resale markup percentages (each of which occurred through a series of two resale transactions) that were so different as to refute any inference that a standard markup percentage applied across all of CPZ’s products in the review. As a result, the “U.S. prices” that emerged from the Department’s calculation were the CEP starting prices in a form that was significantly distorted by the adjustment method Commerce chose. These calculations yielded a price that no longer bore any meaningful relationship to the price at which any specific roller bearings or parts were actually sold. The resulting price was not, in the words of the SAA, “as closely as possible, a price corresponding to an export price between non-affiliated exporters and importers.” Id. Commerce’s choice of facts otherwise available does not fulfill the statutory purpose of remedying dumping, which is found to exist when the price at which subject merchandise is sold in the United States is less than normal value. See 19 U.S.C. §§ 1675(a)(2)(A), 1677(35)(A). Where, as here, Commerce attempts to conduct a periodic administrative review of calculated “export prices” that are not actual prices, and no adverse inference is appropriate because there was no failure of the respondent to cooperate, Commerce has failed to conduct a review that satisfies the basic requirement of the Act. See id. (requiring that dumping margins in a review be based on a comparison of export prices or constructed export prices with normal value). Commerce has considerable discretion in selecting “facts otherwise available,” but the discretion 19 U.S.C. § 1677e(a) provides is not so broad as to allow Commerce to choose its arbitrary, and unauthorized, method of adjusting the CEP starting prices rather than the adjustment method the Act prescribes in 19 U.S.C. § 1677a(c)-(d) and thereby reach a result that fails to satisfy § 1675(a)(2)(A). Although the court accords to Commerce’s constructions of the Act the deference required by 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 concludes in this case that Congress, in the plain language of the statute as well as the SAA, “has directly spoken to the precise question at issue,” id. at 842, 104 S.Ct. 2778, in establishing the requirements by which Commerce must conduct an administrative review of an antidumping duty order. In summary, because Commerce based CPZ’s margin on unlawfully-determined U.S. “prices” that were not prices at which subject merchandise was actually sold, the court must remand the Final Results and order that Commerce correct this fundamental error. In support of its argument that Commerce acted lawfully in determining the U.S. prices, defendant relies in part on AK Steel Corp. v. United States, 226 F.3d 1361 (Fed.Cir.2000). Def.’s Opp’n 13-14. This reliance is misplaced. As discussed infra, AK Steel holds that “the plain language of the EP definition precludes classification of a sale between two U.S. entities (ie., a U.S. affiliate of the producer and a U.S. purchaser) as an EP sale.” AK Steel Corp., 226 F.3d at 1370. The case does not hold that Commerce may base U.S. prices on CEP starting prices that are adjusted by a method that does not conform to the provisions in the statute governing CEP. A On Remand, Commerce Will Not Be Confined to Determining the U.S. Prices on a CEP Basis if it Reopens the Record to Obtain Export Price Information Plaintiff argues that the court should order Commerce on remand to de *1365 termine the U.S. prices on a CEP basis, contending that this is required by 19 U.S.C. § 1677a(b) and Commerce’s administrative practice. 5 Pl.’s Mot. 2, Proposed Order 1; Pl.’s Mem. 24-25. The court disagrees. Due to the importance of ensuring that U.S. price is determined by an appropriate method, the court concludes that Commerce, in the circumstances of this case, should not be precluded from reopening the record to obtain information that might allow it to determine the U.S. prices on an EP basis. In support of its argument that the statute, in 19 U.S.C. § 1677a(b), requires that the CEP basis be used, plaintiff contests various findings of fact, as stated in the Analysis Memo, on which the Department concluded that an EP basis was appropriate. 6 Pl.’s Mem. 13-18. CPZ contends that the “sales process in this case is between two affiliated parties, CPZ and Peer, who set the terms of the sale” and that the sale made by CPZ to the unaffiliated importer “is not the relevant first sale to an unaffiliated party” because the merchandise in that sale was subsequently resold to Peer and CPZ had knowledge of that fact. Id. at 19. According to CPZ, “[a]ll the material terms of sale: price, quantity, specifications were established by Peer,” which issued simultaneous purchase orders to CPZ and the importer. Id. at 15. CPZ argues that “[b]ecause a U.S. seller affiliated with the producer is involved in the U.S. sale process, the sales cannot be reported as EP sales.” Id. at 19. It also argues that Commerce failed “to consider the fact that CPZ had knowledge of the final destination of the sales, the extent of Peer’s involvement in the sales process and the fact that both CPZ and Peer treated these sales as affiliated party sales in their normal books and records.” Id. at 18. Collectively, these arguments are essentially a contention that the “sales” between CPZ and the unaffiliated importer were not “sales” for purposes of determining U.S. prices, and that the real sales should be deemed to have occurred between CPZ and Peer. Although CPZ points to some record evidence in support of its characterization of the “sales process,” the current record *1366 evidence is not such that Commerce would be compelled to reach findings of fact in agreement with that characterization. The record also contains evidence that CPZ actually did sell the subject merchandise to the unaffiliated importer. Based in part on CPZ’s own questionnaire response, Commerce found in the Analysis Memo that “[i]n this case, based upon the facts on the record, the subject merchandise is first sold before the date of importation by CPZ to an unaffiliated purchaser ... in the United States.” Analysis Mem. 3. Commerce observed in the Analysis Memo that CPZ stated in its questionnaire response that it sold the subject merchandise to an unaffiliated company in the United States and that Peer then resold the subject merchandise to the ultimate customer. Id. at 3-4. In its brief to the court, CPZ admits that “[djuring the period of review CPZ made sales of TRBs [i.e., tapered roller bearings]” to the United States through the unaffiliated importer, who then sold to Peer. Pl.’s Mem. 17. Under the applicable standard of review, it is, of course, possible that the same record could be found to contain substantial evidence for two opposite findings. In ordering a remand in this case, the court does not hold or imply that Commerce is bound by the factual findings it previously made in the Analysis Memo pertaining to the U.S. price issue. In addition to reopening the record to admit additional evidence, Commerce, on remand, also has the discretion to reconsider the record evidence, and its previous findings, bearing on the issue of an appropriate determination of U.S. prices. The current state of the record, upon which the Department reached factual conclusions on the nature of the sales between the various entities, is missing a large amount of information pertaining to the sales between CPZ and the unaffiliated importer. If on remand Commerce reopens the record to collect export price information and decides that U.S. price cannot or should not be determined on a CEP basis, then it must include an explanation of its reasoning. In the context of this case, the reasoning would need to include specific findings of fact and relevant conclusions of law. The Final Results and Decision Memorandum do not include an adequate explanation on the question of the determination of U.S. prices, and the Analysis Memo, even if presumed to be incorporated by reference into the Decision Memorandum, also is wanting. In it, Commerce appears to have decided that it must determine the U.S. prices of CPZ’s subject merchandise on an EP basis, rather than a CEP basis, because the first sale of the subject merchandise to an unaffiliated purchaser occurred outside of the United States, before the date of importation, when CPZ sold the merchandise to the unaffiliated importer. See Analysis Mem. 4 (“The record does not sufficiently support a conclusion that the first sale to an unaffiliated purchaser took place after the date of importation or inside the United States.”). The court cannot conclude for certain that Commerce reasoned in this way, however, because other discussion in the Analysis Memo explains Commerce’s reasoning for its decision in the context of “classifying” sales as EP sales. As Commerce stated, “we find that the sales must be classified as EP sales because the record does not sufficiently support CPZ’s assertion that [the unaffiliated importer] is not the first unaffiliated purchaser under the standard set forth in AK Steel. ” Analysis Mem. 5 (citing AK Steel Corp., 226 F.3d at 1372). In the Analysis Memo, Commerce concluded that the United States Court of Appeals for the Federal Circuit (“Court of Appeals”) in AK Steel Corp. “held that section 772 of the Act is unambiguous: the critical factors for *1367 whether a sale should be classified as EP or CEP are the location of the sale and whether the parties to the sale are affiliated or unaffiliated.” Id. at 4-5 (citing AK Steel Corp., 226 F.3d at 1369-70). The Analysis Memo does not base its decision to pursue an EP basis on a construction of the Act, nor do the Final Results or Decision Memorandum. From the Analysis Memo, it appears that the Department’s conclusion may have been based, at least in part, on the holding in AK Steel Corp. If that was the Department’s reasoning, it was not correct. The holding in AK Steel Corp. does not require Commerce to determine the U.S. prices on an export price basis in this case. The question in AK Steel Corp. was “whether a sale to a U.S. purchaser can be properly classified as a sale by the producer/exporter, and thus an EP sale, even if the sales contract is between the U.S. purchaser and a U.S. affiliate of the producer/exporter and is executed in the United States.” AK Steel Corp., 226 F.3d at 1368. Even though the sales in question were made between a U.S. affiliate of the producer/exporter and an unaffiliated U.S. customer, Commerce decided the sales were export price sales, considering the role of the U.S. affiliate to be too minor to satisfy the “PQ Test” that it applied to determine whether a particular sale is an EP sale or a CEP sale. 7 Id. Based on the plain meaning of the export price definition in section 772(a) of the Act, 19 U.S.C. § 1677a(a), the Court of Appeals concluded that the sales transactions at issue in the appeal could not qualify as export price sales, reasoning that “[a] transaction, such as those here, in which both parties are located in the United States and the contract is executed in the United States cannot be said to be ‘outside the United States.’ ” AK Steel Corp., 226 F.3d at 1370. The court held that “the plain language of the EP definition precludes classification of a sale between two U.S. entities (i.e., a U.S. affiliate of the producer and a U.S. purchaser) as an EP sale.” Id. In summary, AK Steel Corp. presented the question of whether a particular sale was properly classified as the starting price for a determination of U.S. price on an EP basis or the starting price for determining U.S. price according to a CEP basis, a question to be resolved, in large part, by determining where the sale at issue took place. On the findings of fact Commerce made, this case presents a different factual circumstance than did AK Steel Corp., which involved a Korean steel producer who sold subject steel to an affiliated Korean exporter, who then sold the merchandise to a U.S. affiliate, who in turn sold the merchandise to an unaffiliated U.S. purchaser. Id. at 1365. This case, viewed according to Commerce’s findings, presents a circumstance in which the foreign producer and its U.S. affiliate each make a sale of the subject merchandise to an unaffiliated U.S. purchaser (but not the same purchaser) and in which the sale by the foreign producer, made outside of the United States, occurs first. The issue of “classification” that was involved in AK Steel Corp. does *1368 not arise on the facts of this case, as found by Commerce. On those facts, the sale by CPZ to the importer cannot be the starting price for a CEP sale, and the sale by Peer to the un affiliated customer cannot be a EP sale. The issue presented instead is whether the statute contains a requirement for the determination of U.S. price on an EP basis on facts such as those found by Commerce, under which the sale by CPZ to the importer occurred before the sale by Peer to the unaffiliated customers. Commerce apparently concluded that there is such a requirement, but in doing so it relied, at least in part, on its interpretation of the holding in AK Steel Corp. 8 The plain language of subsections 772(a) and (b) of the Act does not specify that Commerce must determine U.S. price on an export price basis on the facts it found. See 19 U.S.C. § 1677a(a) (referring to the first sale to an unaffiliated purchaser that occurs outside of the United States), § 1677a(b) (referring to the first sale to an unaffiliated purchaser that occurs in the United States, by the producer or exporter or a seller affiliated with the producer or exporter). The statute directs Commerce to determine the “export price (or constructed export price) of each entry of the subject merchandise.” Id. § 1675(a)(2)(A)(i) (emphasis added). Commerce determines “the dumping margin for each such entry,” id. § 1675(a)(2)(A)(ii), but “dumping margin” is “the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise,” id. § 1677(35)(A) (emphasis added). Although the statute is silent on the narrow question presented, the SAA appears to speak to that question. Under the subheading “Identification of the Starting Price,” the SAA provides, in pertinent part, that “[i]f the first sale to an unaffiliated purchaser in the United States ... is made by the producer or exporter in the home market prior to the date of importation, then Commerce will base its calculation on export price. ” SAA at 822, reprinted in 1994 U.S.C.C.A.N. at 4163 (emphasis added). The SAA provides, further, that “[i]f ... the first sale to an unaffiliated person is made by (or for the account of) the producer or exporter or by a seller in the United States who is affiliated with the producer or exporter, then Commerce will base its calculation on constructed export price.” Id. (emphasis added). Thus, the SAA might be read as an indication that where the first sale to an unaffiliated purchaser occurs prior to importation and outside of the United States, the statute requires that the price in that sale be used as the starting price in an EP determination, even if the price in a subsequent resale of the subject merchandise conforms to the definition of the CEP starting price set forth in 19 U.S.C. § 1677a(b). The court does not hold, however, that Commerce could not reach a different interpretation of the SAA language. Because Commerce did not discuss a construction of the statute in choosing an EP basis, and because of the principles of delegation by Congress to an agency and deference to the agency’s statutory constructions by the courts that the Supreme Court discussed in Chevron, 467 U.S. at 842-43, 104 S.Ct. 2778, the court refrains from opining on a hypothetical statutory construction that Commerce has not made in this case. *1369 The court also refrains from opining on whether Commerce, if it chooses not to reopen the record, still might conclude, based on any such statutory construction that Commerce might make on remand, that the missing export price information is “necessary information” for which Commerce must substitute facts otherwise available. 9 On the record in its current state, however, there is no alternative to the Department’s determining U.S. prices on a CEP basis, for the reasons discussed previously: only CEP starting prices are on the record (with insignificant exceptions), and those prices, adjusted by the method required by the Act, are the only record information the statute will permit to be used in place of the missing EP information as facts otherwise available. Therefore, the court will order Commerce, on remand, to determine the U.S. prices on a CEP basis unless it decides to reopen the record to obtain additional information, including, specifically, prices that qualify as the starting prices for an EP determination under 19 U.S.C. § 1677a(a). B. Commerce’s Choice of Data to Value Alloy Steel Wire Rod Was Not Supported by Substantial Evidence on the Record Plaintiff claims that substantial evidence did not support Commerce’s finding that Indian import data, which showed an average price of $3,877 per metric ton, were the “best available information” for valuing CPZ’s input of bearing-quality alloy steel wire rod. 10 Pl.’s Mem. 33-35. Plaintiff advocates that Commerce instead should have used record import data from Indonesia ($1,184 per metric ton) or the Philippines ($1,327 per metric ton), arguing, inter alia, that the Indian import data do not reliably indicate the price CPZ would have had to pay in a market economy because they indicate prices “almost 400% higher than benchmark prices available on the record.” Id. at 33. Plaintiff points to U.S. import data under a ten-digit tariff subheading specific to bearing-quality steel, which lists an average value of $1,391 per metric ton, as corroboration of the price reflected by the Indonesian and Filipino data. Id. at 34. 11 *1370 When Commerce determines the normal value of subject merchandise of a producer in a non-market economy country, “the valuation of the factors of production shall be based on the best available information regarding the values of such factors in a market economy country or countries considered to be appropriate by the administering authority.” 19 U.S.C. § 1677b(c). Commerce “normally will value all factors in a single surrogate country” and “normally will use publicly available information to value factors.” 19 C.F.R. § 351.408(c)(l)-(2) (2007). Commerce identified its criteria for “best available information” by stating that its “preference is to use, where possible, a range of publicly available, non-export, tax-exclusive, and product-specific prices for the POR ... with preference to data from a single surrogate country.” Decision Mem. 12. Commerce’s practice is to reject “aberrational” data that does not reliably indicate the price a producer would have paid for the input. See Antidumping Duties; Countervailing Duties, 62 Fed. Reg. 27,296, 27,366 (May 19, 1997) (“Preamble ”). Commerce’s findings on what constitutes the best available information for valuing a factor of production must be supported by substantial evidence. 19 U.S.C. § 1516a(b)(l)(B)(i). Substantial evidence 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). Commerce has considerable discretion in choosing the surrogate values that most accurately reflect the price that the non-market economy producer would have paid had it purchased the factor of production from a market economy country. Nation Ford Chemical Co. v. United States, 166 F.3d 1373, 1377 (Fed.Cir.1999). However, the court “cannot evaluate the substantiality of evidence supporting” Commerce’s conclusion “ ‘merely on the basis of evidence which in and of itself justified it, without taking into account contradictory evidence ....’” Suramerica de Aleaciones Laminadas, C.A. v. United States, 44 F.3d 978, 985 (Fed.Cir.1994) (quoting Universal Camera Corp. v. NLRB, 340 U.S. 474, 487, 71 S.Ct. 456, 95 L.Ed. 456 (1951)). Commerce obtained its value of $3,877 per metric ton for CPZ’s alloy steel wire rod from the average entered value of imports shown in subheading 7228.50.90 of the Harmonized Tariff Schedule of the Republic of India (“Indian HTS”). Decision Mem. 14; see Final Factor Valuation Mem. attachment l. 12 Filipino import data showed a value of $1,327 per metric ton, and the value in Indonesian import data was $1,184 per metric ton. Letter from CPZ to the Sec’y of Commerce exhibit 2 (Aug. 7, 2008) (Admin.R.Doc. No. 5420) (“CPZ’s Surrogate Value Letter”). The record evidence does not indicate a reason for this striking difference in price, which exists despite certain similarities in the import data from the three countries. The Indonesian and Filipino data, like the Indian data, are from the World Trade Atlas and cover the same time period as the Indian data. Compare id. with Final Factor Valuation Mem. attachment 1. Like India, Indonesia and the Philippines are countries Commerce considered to be at levels of economic development comparable to that of China. Prelim. Results, 73 Fed.Reg. at 41,034. The Indonesian and Filipino import data, like the Indian import data, pertained to information collect *1371 ed under the internationally-harmonized six-digit subheading 7228.50, which pertains to the article description “Other bars and rods of other alloy steel; angles, shapes and sections, of other alloy steel; hollow drill bars and rods, of alloy or non-alloy steel: Other bars and rods, not further worked than cold-formed or cold-finished.” Pl.’s Mem. 33-34; see World Customs Org., Harmonized Commodity Description and Coding System Explanatory Notes 72.28, at 1097-1098 (2d ed. 1996) {“Explanatory Notes”). No party disputes that this subheading describes CPZ’s input. The record evidence shows one difference existing among the import data sets, but it does not explain the great disparity in price, nor does it constitute substantial evidence that the Indian data were superi- or to the other two data sets. The Indian data that Commerce used pertained to an eight-digit India-specific subheading that applies to goods classified in the six-digit subheading 7228.50, except that it excludes goods classified as “Of engine valves and cold heading steel.” Indian HTS 7228.50.10, 90, available at http:// compendium.iift.ac.in/index.asp (last visited Jan. 28, 2011). In the Decision Memorandum, Commerce concluded that the additional specificity to the input being valued that was provided by the eight-digit breakout subheading supported the choice of the Indian tariff data over the tariff data from Indonesia and the Philippines. Decision Mem. 14. However, the eight-digit subheading, even after the two items are excluded, still includes a very broad variety of different products. Data on the record support this point. CPZ’s Surrogate Value Letter exhibit 7. The additional specificity provided by the exclusion of the two items, when viewed in the context of the record as a whole, does not support a finding that the Indian data are the best available information. The same record demonstrates that the values obtained from the Indonesian and Filipino import data corroborate each other and are further corroborated by the U.S. import data. The record also reveals that all three of these values are greatly at variance with the value drawn from the Indian data. Finally, although the Department, as stated in its regulations, 19 C.F.R. § 351.408(c)(2), normally values all factors of production in a single country, this preference is not a reason by which the court may sustain a finding that a particular data set constitutes the “best available information” for purposes of 19 U.S.C. § 1677b(c), where, as here, substantial evidence did not support that choice. Defendant and defendant-intervenor argue that the Indian import data were not aberrational because “[h]igh values alone do not support a comparison of import data and a subsequent rejection of import data-not without specific evidence that the import data are aberrational.” See Def.’s Opp’n 29; Def.-Intervenor’s Opp’n 32-33. The general principle they advance is of dubious merit when applied in this circumstance, which is the enormous disparity between the value shown in the Indian data and the data from the other two countries. Absent record evidence to explain it, a disparity of this size calls into serious question a finding that the Indian data were the best available information on the record with which to value the factor of production. Because nothing on the record resolves that serious question, the court is unable to conclude that Commerce’s choice was based on substantial evidence. Defendant argues, further, that the Indian data were the best available information because it “is possible that a comparison of a larger set of appropriate data sources ... would lead to the opposite conclusion, that the Indian data are not so different from other sources.” Def.’s *1372 Opp’n 32. This argument lacks merit. The absence of evidence corroborating the price reflected by the Indian import data is evidence that detracts from, rather than supports, a finding that this price is based on the best available information. Finally, both defendant and defendantintervenor argue that import data pertaining to the United States do not bear on whether the Indian import data were the best available information because China and the United States are not at comparable levels of economic development. Def.’s Opp’n 33; Def.-Intervenor’s Opp’n 33-35. This argument misreads the statute, which directs Commerce to value factors of production using the best available information from countries at a level of economic development comparable to that of China. See 19 U.S.C. § 1677b(c)(l), (4). The statute does not prohibit Commerce from considering, for corroboration purposes, record evidence consisting of prices for a commodity in a market economy country when determining which information from countries at a level of economic development comparable to China is the best available information. In summary, the enormous disparity between the value for alloy steel wire rod shown in the Indian import data and the values shown in all the other information on the record calls into serious question a finding that the Indian data were the best available information with which to value the factor of production. Nothing on the record resolves that serious question. The two competing data sets yield values that corroborate each other and are further corroborated by the U.S. data. The court, therefore, is unable to conclude that Commerce’s choice was based on substantial evidence. C. In Valuing Alloy Steel Bar, Commerce Failed to Compare the Indian Import Data with the Import Data from Indonesia and the Philippines and Did Not Provide a Rational Explanation Plaintiff claims that substantial evidence did not support Commerce’s finding that WTA Indian import data pertaining to Indian tariff subheading 7228.30.29, 13 reflecting a value of $1,607 per metric ton, were the best available information to value CPZ’s input of bearing-quality alloy steel bar. Pl.’s Mem. 35-37; Final Factor Valuation Mem. attachment 1. According to CPZ, the Indian data were aberrational as used to value this input because they indicate prices “more than 60% higher than the benchmark prices.... ” PL’s Mem. 35. The benchmarks to which CPZ refers include import data from the United States, Indonesia, and the Philippines. 14 Id. at 36. From its examination of the record and its consideration of the reasoning stated in the Decision Memorandum, the court concludes that Commerce’s choice of the Indian HTS import data is not accompanied by a rational explanation of why those data are the best available information on the record, which also contained import data from Indonesia and the Philippines. Here, as elsewhere, Commerce must provide a rational explanation for its choice. See Motor Vehicle Mfrs. Ass’n v. State Farm Mut., 463 U.S. 29, 43, 103 S.Ct. *1373 2856, 77 L.Ed.2d 443 (1983) (an agency must “articulate a satisfactory explanation for its action including a ‘rational connection between the facts found and the choice made.’ ”) (citing Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962)). As one of the stated reasons for its choice of the Indian import data, Commerce cites the aforementioned regulatory preference for valuing factors of production using data from a single surrogate country. Decision Mem. 15-16. That preference, according to 19 C.F.R. § 351.408(c)(2), plainly is a factor in the Department’s decision to choose one data set over another. However, because the statute requires Commerce to compare the chosen data set with other data sets on the record and thereby determine what is the best available information, the regulatory preference cannot suffice as adequate reasoning if it is the only factor that Commerce considers. Cf. 19 C.F.R. § 351.408(c)(2) (providing that “the Secretary normally will value all factors in a single surrogate country”) (emphasis added); 19 U.S.C. § 1677b(c)(l) (requiring Commerce to identify the “best available information” regarding the value of a factor of production in a market economy country or countries considered to be appropriate by the administering authority). The preference for using data from a single country might support a choice between data sets that, upon a fair comparison, are otherwise seen to be fairly equal, but the Decision Memorandum contains no indication that Commerce actually made a comparison of the Indian data with the Indonesian and the Filipino data. Despite the absence of any mention of an actual comparison of the three sets of import data, Commerce states in the Decision Memorandum that “because the WTA India data are contemporaneous with the POR, are publicly available, and represent a broad market average, we find that they represent the best available information for purposes of valuing the steel bar input.” Decision Mem. 16. It would appear from the record that the Indonesian and Filipino import data are also contemporaneous with the POR, are also publicly available, and also represent a broad market average. CPZ’s Surrogate Value Letter exhibit 1. Additionally, a fair comparison of the three import data sets necessarily would have given due consideration to the record evidence showing that the Indonesian and Filipino data sets present values that corroborate each other rather closely and are at considerable variance with the value obtained from the Indian import data. The record information consisting of Indonesian and Filipino import data pertaining to the internationally-harmonized subheading 7228.30 revealed average values of $964 per metric ton and $1,088 per metric ton, respectively. Id. The Indonesian value and Filipino value are further corroborated by data on U.S. imports of bearing-quality alloy steel bar, which, unlike the other import data on the record, pertained specifically to bearing-quality alloy steel bar and reflected an average value of $1,040 per metric ton. 15 As another reason offered in connection with its choice of the Indian data, Commerce states that “the existence of higher *1374 prices alone does not necessarily indicate that price data is distorted or misrepresented, and, thus, the existence of a higher price is not sufficient to exclude a particular surrogate value, absent specific evidence the value is otherwise aberrational” and that “[ojther than pointing out that the Indian import data reflect a high value, we find that CPZ has not demonstrated that the WTA Indian data are unreliable.” Decision Mem. 15. This explanation fails to offer a reason why Commerce considered the Indian data to be better information than the Indonesian or Filipino data. It is, instead, merely the reason Commerce offered for rejecting CPZ’s objection that the value obtained from the Indian data is aberrational. Moreover, Commerce characterized that value as the “existence of higher prices,” but a fair characterization would recognize it as a substantially higher price that is not corroborated by the other evidence on the record. In summary, the record evidence demonstrating that the Indonesian and Filipino values corroborate each other rather closely, particularly when viewed together with the additional corroboration provided by the U.S. import data, and the record evidence showing that the Indian value is substantially higher than the corroborated value, casts doubt on whether substantial evidence supported the choice of the Indian import data as the best available information on the record before the Department. Based on that record as a whole, and even considering the preference established in 19 C.F.R. § 351.408(c)(2), the court is unable to sustain the choice of the Indian data on the reasoning Commerce has put forth, from which a fair comparison of the three import data sets is lacking. On remand, Commerce must review and reconsider its choice of the Indian data to value the bearing-quality alloy steel bar and must reach a determination that is supported by substantial record evidence and an adequate explanation. D. Plaintiffs Claim Contesting Commerce’s Valuation of Steel Scrap Resulting from Production of Rollers and Rings Must Be Rejected Plaintiff contests as unlawful, or inadequately explained, Commerce’s choice of import data with which to value the scrap resulting from CPZ’s production of rollers and rings. PL’s Mem. 37-39. The court finds no merit in this claim. In the Final Results, Commerce used Indian import data under subheading 7204.29.90 of the Indian HTS, which pertains to scrap of non-stainless alloy steel, to value scrap generated from CPZ’s roller and ring production. Decision Mem. 8. This was a change in position from the Preliminary Results. See Prelim. Factor Valuation Mem. 3. In the Preliminary Results, Commerce valued the roller and ring scrap according to Indian HTS import data under subheading 7204.41, id., which subheading, according to the internationally-harmonized article description, pertains to ferrous waste and scrap other than waste and scrap of cast iron, alloy steel, or tinned iron or steel, Explanatory Notes 72.04, at 1075. CPZ argues that this change in position was in error and that Commerce’s explanation for changing the surrogate value “is either insufficient or incorrect as a matter of law.” PL’s Mem. 39. Plaintiffs argument fails to confront the essential point that in the Final Results, Commerce based its choice of subheading 7204.29.90 on a finding of fact that the steel CPZ used in its roller and ring production was alloy steel. See Decision Mem. 8. Nowhere in its submissions does CPZ contest that finding of fact, based on which the scrap resulting from roller and ring product is properly classified under the Indian HTS steel scrap subheading that encompasses scrap of alloy steel, subheading 7204.29.90, and not under sub *1375 heading 7204.41, from which scrap of alloy steel is specifically excluded. E. Commerce’s Choice of Data for the Valuation of Steel Scrap Resulting from Production of Cages Lacks Essential Findings of Fact and Appears to Be Based on a Misclassification In the Final Results, Commerce valued the scrap resulting from CPZ’s cage production according to data in Indian HTS subheading 7204.49, changing its position from the Preliminary Results, in which Commerce valued the scrap according to import data in Indian HTS subheading 7204.41. Decision Mem. 8; Prelim. Factor Valuation Mem. 3. Commerce changed its position after considering a comment by Timken, in which Timken argued that the value derived from import data under subheading 7204.41 was aberrational in that the value determined for the scrap, illogically, would be higher than the value shown by the Indian HTS import data for the steel product used in making the cages. Decision Mem. 8. The Department explained that its decision to change its position based on its finding of an aberrational value was consistent with a similar decision it made in a previous final determination involving the same two Indian HTS subheadings, Nails from the PRC. Id. at 9 (“Given the nearly identical facts here, in which the surrogate value for cage steel scrap exceeds the surrogate value for the direct material input, and in keeping with Department precedent, we have revalued the steel scrap generated from CPZ’s cage production using Indian HTS 7204.49.00 for the Final Results.”) (citing Certain Steel Nails from the People’s Republic of China: Final Determination of Sales at Less Than Fair Value & Partial Affirmative Determination of Critical Circumstances, 73 Fed.Reg. 33,977 (June 16, 2008); Issues & Decision Mem., A-570-909, ARP 6-08, at 38 (June 6, 2008), available at h ttp://ia.ita.doc.gov/frn/summary/PRC/E8-13474-l.pdf). Indian HTS subheadings 7204.41 and 7204.49 are six-digit subheadings, and therefore the associated article descriptions are internationally harmonized. See Explanatory Notes 72.04, at 1075. According to the relevant internationally-harmonized article description, subheading 7204.41 pertains to ferrous waste and scrap (other than waste and scrap of cast iron, alloy steel, or tinned iron or steel) in the form of “[t]urnings, shavings, chips, milling waste, sawdust, filings, trimmings and stampings, whether or not in bundles.” Id. Subheading 7204.49 is a “basket” subheading (“Other”) that pertains to other forms of such scrap, including “[articles of iron or steel, definitely not usable as such because of breakage, cutting-up, wear or other reasons....” Id.; see also Note 8(a) to Section XV, Harmonized Commodity Description and Coding System. Only goods failing to satisfy the article description for subheading 7204.41 are classifiable in subheading 7204.49. See Explanatory Notes 72.04, at 1075. The Decision Memorandum contains no findings of fact under which it could be concluded that the scrap resulting from CPZ’s cage production would be excluded from Indian HTS subheading 7204.41 and therefore be classifiable under subheading 7204.49. Nor is the court able to identify record evidence from which any such findings could be made. To the contrary, one CPZ questionnaire response suggests that this scrap would satisfy the article description for subheading 7204.41 because this scrap was “the result of the machining of raw materials.” Letter from CPZ to the Sec’y of Commerce 3 (Aug. 11, 2008) (Admin.R.Doc. No. 5422). The court is unable to sustain the Department’s choice of data pertaining to Indian HTS subheading 7204.49 without appropriate findings of fact, supported by substantial evidence on the record, that identify the physical nature of this scrap for purposes of tariff *1376 classification. The Decision Memorandum not only fails to include such findings of fact but vaguely suggests, in its discussion of the Department’s previous decision in Nails from the PRC, that the Department considered it acceptable to determine a value based on a tariff classification other than the one that pertained to the good being valued. See Decision Mem. 9 (“The Department concluded that, although the relevance of the HTS description is an important factor in the selection of a surrogate value, it is not the sole consideration, and cannot be relied upon when it produces unreasonable results.”). If that was the Department’s reasoning (and the point is not entirely clear), then the court would disagree that Commerce, in the circumstances of this case, could base its valuation of the scrap on an incorrect tariff classification. In summary, the Department’s choice of data from Indian HTS subheading 7204.49 is unsupported by findings of fact essential to the correct tariff classification of the scrap being valued. As a result, the court is unable to sustain a finding that this choice constituted the best available information on the record. III. Conclusion and Order For the reasons discussed in the foregoing, and after consideration of all submissions made in this action, the court concludes that Commerce did not act in accordance with law when it calculated CPZ’s dumping margin. Commerce acted contrary to law in determining the U.S. price of CPZ’s subject merchandise and in determining surrogate values for alloy steel wire rod, alloy steel bar, and steel scrap from production of cages. The court, therefore, will remand the Final Results to Commerce for redetermination consistent with this Opinion and Order. ORDER Upon review of the Final Results and all papers and proceedings herein, it is hereby ORDERED that plaintiffs Rule 56.2 Motion for Judgment upon the Agency Record be, and hereby is, GRANTED in part and DENIED in part; it is further ORDERED that Commerce shall issue upon remand a redetermination (“Remand Redetermination”) that complies in all respects with this Opinion and Order, is based on determinations that are supported by substantial record evidence, and is in all respects in accordance with law; it is further ORDERED that Commerce, in preparing the Remand Redetermination in accordance with this Opinion and Order, shall redetermine the margin for Peer Bearing Company-Changshan (“CPZ”) based on redetermined U.S. prices of CPZ’s subject merchandise that are calculated according to a method that complies with law; it is further ORDERED that Commerce may reopen the record to obtain additional information in preparation for issuing the Remand Re-determination, but if it does not do so and does not obtain price information qualifying for use as starting prices for a determination of export prices according to 19 U.S.C. § 1677a(a), then, as required by law, it must determine the U.S. prices on a constructed export price basis, whether or not it relies on its authority to use facts otherwise available under 19 U.S.C. § 1677e(a); it is further ORDERED that Commerce, if it reopens the record and collects the information described above and decides to determine U.S. prices on an export price basis, shall provide a reasoned explanation for its decision to do so, stating the relevant findings of fact and conclusions of law; it is further *1377 ORDERED that Commerce, in preparing the Remand Redetermination in accordance with this Opinion and Order, shall review, reconsider, and redetermine the surrogate values for alloy steel wire rod, alloy steel bar, and scrap from the production of cages; it is further ORDERED that Commerce shall file the Remand Redetermination no later than ninety (90) days from the date of this Opinion and Order; it is further ORDERED that plaintiff shall be allowed thirty (30) days from defendant’s filing of the Remand Redetermination to file any comments thereon; and it is further ORDERED that defendant shall be allowed fifteen (15) days from the filing of plaintiffs comments in which to file any rebuttal to such comments. Footnotes [1] The merchandise subject to the order is "tapered roller bearings and parts thereof, finished and unfinished, from the PRC; flange, take up cartridge, and hanger units incorporating tapered roller bearings; and tapered roller housings (except pillow blocks) *1357 incorporating tapered rollers, with or without spindles, whether or not for automotive use.” Tapered Roller Bearings & Parts Thereof, Finished & Unfinished, From the People’s Republie of China: Final Results of Antidumping Duty Admin. Review, 74 Fed.Reg. 3,987, 3,988 (Jan. 22, 2009) ("Final Results "). [2] This method is described in section 773(c) of the Tariff Act of 1930 ("Tariff Act” or the "Act”), 19 U.S.C. 1677b(c) (2006) ("[T]he administering authority shall determine the normal value of the subject merchandise on the basis of the value of the factors of production utilized in producing the merchandise and to which shall be added an amount for general expenses and profit plus the cost of containers, coverings, and other expenses.”). [3] The International Trade Administration, U.S. Department of Commerce ("Commerce” *1360 or the "Department”) reduces the constructed value by the amount a producer recovers from the sale of by-products or scrap. Issues & Decisions Mem., A-570-601, ARP 1-09, at 3-4 (Jan. 13, 2009) (Admin.R.Doc. No. 5486), available at http://ia.ita.doc.gov/frn/summary/ PRC/E9-1219-1 .pdf (“Decision Mem.’’). In its final determination, Commerce granted an offset for Peer Bearing Company-Changshan’s ("CPZ”) sales of steel scrap from production of cages, rollers, and rings. Final Results, 74 Fed.Reg. at 3,988. [4] With respect to both export price and constructed export price, Commerce uses the term “starting price” to refer to the unadjusted price at which subject merchandise is sold to the unaffiliated purchaser. 19 C.F.R. § 351.402(a) (2007) ("In order to establish export price, constructed export price, and normal value, the Secretary must make certain adjustments to the price to the unaffiliated purchaser (often called the 'starting price') in both the United States and foreign markets.”). [5] During the review, and again before the court, CPZ argued that Commerce should follow the practice it followed in Certain Cut-to-Length Carbon Steel Plate from Romania: Preliminary Results of Admin. Review & Notice of Intent To Rescind in Part, 69 Fed.Reg. 54,108, 54,115 (Sept. 7, 2004), and subsequent preliminary phases of the same proceeding. PL’s Mem. of Points & Authorities in Supp. of its Mot. for J. on the Agency R. 22-23 ("PL's Mem.”). Before the court, CPZ cited various other administrative proceedings which it claims show that Commerce departed from a practice in reaching the decision to pursue an export price ("EP”) basis in the review at issue. Pl.’s Mem. 20-23. The court does not reach the issue of whether Commerce departed from a practice in the final results of this administrative review ("Final Results”), or gave adequate reasons for doing so, having concluded that the determination of U.S. prices in the Final Results was otherwise contrary to law. [6] In addition to contesting the Department's specific factual findings, CPZ also argues that using constructed export price (“CEP”) is consistent with the purpose of the antidumping law because CPZ, after the sale to Peer, "continues to have the opportunity to control or participate in pricing decisions in the U.S. market through its U.S. affiliate.” PL's Mem. 24. This argument impliedly presumes that Commerce would not have been required by law to determine U.S. price on an EP basis had the record permitted it to do so. Because the court concludes that Commerce has discretion as to whether to reopen the record to obtain export price information and that, if Commerce does so, Commerce must address the issue of whether it was required by law to determine U.S. price on an EP basis, the court, in ordering a remand, does not reach the question raised by this argument. [7] As the Court of Appeals for the Federal Circuit ("Court of Appeals”) explained in AK Steel Corp. v. United States, 226 F.3d 1361, 1365 (Fed.Cir.2000): Using the PQ Test, Commerce classifies sales made by U.S. affiliates as EP sales if the following criteria are met: (1)the subject merchandise was shipped directly from the manufacturer to the unrelated buyer, without being introduced into the inventory of the related shipping agent; (2) direct shipment from the manufacturer to the unrelated buyer was the customary channel for sales of this merchandise between the parties involved; and (3) the related selling agent in the United States acted only as a processor of sales-related documentation and a communication link with the unrelated U.S. buyer. [8] In its opinion in AK Steel Coip., the Court of Appeals recognized that Commerce, in choosing between an EP basis and a CEP basis, identified the first sale to an unaffiliated purchaser, but the discussion is not part of the holding of the case. 226 F.3d at 1367 ("To isolate an arm’s length transaction under the current statute, Commerce looks to the first sale to a purchaser that is not affiliated with the producer or exporter.”). [9] Section 776(a) of the Tariff Act provides, in pertinent part, that if "necessary information is not available on the record ... the administering authority ... shall ... use the facts otherwise available in reaching the applicable determination under this subtitle.” 19 U.S.C. § 1677e(a) (emphasis added). Relevant to identifying the "applicable determination” are subparagraphs (i) and (ii) of section 751(a)(2)(A) of the Act, under which Commerce is to make, respectively, the following two determinations in a periodic administrative review: "(i) the normal value and export price (or constructed export price) of each entry of the subject merchandise, and (ii) the dumping margin for each such entry.” Id. The Act, in section 771(35)(A), defines “dumping margin” as “the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise.” Id. § 1677(35)(A). [10] Commerce failed to state explicitly that its chosen surrogate value for steel wire rod was based on the best available information. Instead, it found that "CPZ has failed to prove the inadequacy of the Indian data or to demonstrate another value to be more appropriate.” Decision Mem. 12. The court construes Commerce’s choice of the surrogate value for steel wire rod as an implicit finding that the Indian import data were the best available information on the record. [11] Subheading 7228.50.10.10, Harmonized Tariff Schedule of the United States (2007) (“HTSUS”), pertains to: "Other bars and rods of other alloy steel; angles, shapes and sections, of other alloy steel; hollow drill bars and rods, of alloy or non-alloy steel: Other bars and rods, not further worked than cold-formed or cold-finished: Of tool steel (other than high-speed steel): Of ball-bearing steel.” In reaching its conclusion, the court does not consider CPZ's modified value of $1,367 per metric ton. See Case Br. of Peer Bearing Co.Changshan & Peering Bearing Co. 8 (Aug. 26, 2008) (Admin.R.Doc. No. 5431) ("CPZ’s Case Br.’’) (explaining that $1,367 per metric ton excludes what CPZ defines as "aberrationally high monthly data”). [12] This source shows values in Indian rupees. For ease of comparison, the court discusses values in U.S. dollars, as did the parties. [13] Commerce’s findings incorrectly referred to "Indian HTS 7288.3029." Decision Mem. 16. Commerce used import data pertaining to subheading 7228.30.29. Mem. from Program Manager to file attachment 1 (Jan. 13, 2009) (Admin.R.Doc. No. 5484). [14] CPZ also put evidence on the record of several of its own purchases from an economically developed market economy country, as well as data pertaining to exports from Japan to the United States and Japan to India. Letter from CPZ to the Sec’y of Commerce exhibit 4 (Aug. 7, 2008) (Admin.R.Doc. No. 5420) ("CPZ’s Surrogate Value Letter ”). [15] For its brief, CPZ removed the “aberrationally high” prices before calculating an average of $966 per metric ton. PL’s Mem. 36; CPZ’s Case Br. 11-12. $1,040 USD/MT is the average value including all transactions. CPZ’s Surrogate Value Letter exhibit 1. HTSUS subheading 7228.30.20 includes "Other bars and rods of other alloy steel; angles, shapes and sections, of other alloy steel; hollow drill bars and rods, of alloy or non-alloy steel: Other bars and rods, not further worked than hot-rolled, hot-drawn or extruded: Of ball-bearing steel.”