Jtekt Corporation v. United States

OPINION AND ORDER

JTEKT Corporation, formerly Koyo Seiko Company, Ltd., 1 and Koyo Corporation of U.S.A. (collectively, “JTEKT”) brought an action pursuant to 28 U.S.C. § 1581(c) (2006) to contest the final determination of the United States Department of Commerce (“Commerce” or the “Department”) in the sixteenth administrative reviews (“AFBs 16 reviews” or “AFBs 16”) of antidumping duty orders on ball bearings and parts thereof (“subject merchandise”) from France, Germany, Italy, Japan, and the United Kingdom. Summons 1; Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, & the United Kingdom: Final Results of Anti-dumping Duty Admin. Reviews, 71 Fed. Reg. 40,064, 40,065 (July 14, 2006) (“Final Results”)-, Issues & Decision Mem. for the Antidumping Duty Admin. Reviews of Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, & the United Kingdom for the Period of Review May 1, 200U, through April 30, 2005, at 2 (July 14, 2006) (“Decision Mem.”). The reviews applied to imports of subject merchandise made during the period of May 1, 2004 through April 30, 2005 (“period of review” or “POR”). Final Results, 71 Fed.Reg. at 40,065.

Upon defendant’s consent motion, the court consolidated JTEKT’s action with five other cases. Consent Mot. to Consolidate 1. The five other groups of plaintiffs in the consolidated cases (referred to in this Opinion and Order collectively with their affiliates) are FYH Bearing Units USA, Inc. and Nippon Pillow Block Company Ltd. (collectively, “NPB”); NSK Corporation, NSK Ltd., and NSK Precision America, Inc. (collectively, “NSK”); American NTN Bearing Manufacturing Corp., NTN Bearing Corporation of America, NTN Bower Corporation, NTN Corporation, NTN Driveshaft, Inc., and NTN-BCA Corporation (collectively, “NTN”); Ñachi Technology, Inc., Nachi-Fujikoshi Corporation, and Ñachi America, Inc. (collectively “Ñachi”); and The Timken Company (“Timken”).

JTEKT, NPB, NSK, NTN, and Ñachi (collectively, “plaintiffs”), as well as Timken, which is both a plaintiff and the defen *1213 dant-intervenor (“defendant-intervenor”) in the consolidated cases, bring claims contesting various decisions and determinations that Commerce made in the Final Results. These claims are discussed in the respective sections of Part II of this Opinion and Order, as follows: (A) claims of JTEKT, NPB, NTN, and Ñachi challenging the application of Commerce’s “zeroing” methodology to non-dumped sales; (B) claims challenging the Department’s revised model-match methodology, the adoption of which JTEKT, NPB, NSK, NTN, and Ñachi oppose generally and the specific application of which JTEKT, NPB, NSK, and NTN challenge in certain respects; (C) JTEKT’s claim objecting to Commerce’s treating JTEKT and an affiliate as a single entity, (D) NSK’s claim that Commerce unlawfully deducted certain benefits expenses when determining the constructed export price of NSK’s subject merchandise, (E) NTN’s claim opposing Commerce’s reallocation of NTN’s freight expense on the basis of weight, (F) NTN’s claim opposing Commerce’s recalculation of NTN’s home market packing expenses, (G) NTN’s claim challenging the Department’s disallowance of NTN’s downward price adjustments to reflect certain discounts to home market customers, (H) Na-chi’s claim challenging Commerce’s use of facts otherwise available and adverse inferences in response to errors Ñachi made in reporting physical characteristics of subject bearings, and (I) Timken’s claim challenging Commerce’s use of Japanese interest rates, rather than U.S. interest rates, for a portion of the adjustment for imputed interest carrying costs in the calculation of constructed export prices of subject merchandise of NTN and Ñachi. As discussed in this Opinion and Order, the court grants relief on certain of these claims through an order of remand and, with respect to other claims, affirms Commerce’s decisions and determinations in the Final Results.

I. Background

The court sets forth below the procedural history of the administrative and judicial proceedings in general terms common to all plaintiffs. Additional background information specific to the individual claims is presented in Part II of this Opinion and Order.

A. Administrative Proceedings

On May 15, 1989, Commerce issued anti-dumping duty orders on imports of ball bearings from France, Germany, Italy, Japan, and the United Kingdom. 2 On June 30, 2005, Commerce initiated the sixteenth set of administrative reviews of these orders. Initiation of Antidumping & Countervailing Duty Admin. Reviews, 70 Fed. Reg. 37,749, 37,756-57 (June 30, 2005); Decision Mem. 2. Commerce issued the preliminary results of the administrative reviews (“Preliminary Results”) in March 2006, setting forth its analysis for certain of its initial determinations. Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, & the United Kingdom: Prelim. Results of Antidumping Duty Admin. Reviews, 71 Fed.Reg. 12,170 (Mar. 9, 2006) (“Prelim. Results ”). Later that year, Commerce issued the Final Results *1214 and incorporated by reference therein an internal Issues and Decision Memorandum (“Decision Memorandum”) containing the Department’s analysis of issues raised by interested parties subsequent to the Preliminary Results. Final Results, 71 Fed. Reg. at 40,065; see Decision Mem.

B. Judicial Review in the Consolidated Actions

On September 13, 2006, the court granted the consent motion of Timken to intervene on behalf of defendant. Upon defendant’s consent motion, the court ordered consolidation under Consolidated Court No. 06-00250 of JTEKT Corporation v. United States, No. 06-00250, Nippon Pillow Block Company Ltd. v. United States, No. 06-00258, Timken U.S. Corporation v. United States, No. 06-00271, NSK Ltd. v. United States, No. 06-00272, NTN Corporation v. United States, No. 06-00274, and Nachi-Fujikoshi Corporation v. United States, No. 06-00275. Order 1, Nov. 15, 2006; Consent Mot. to Consolidate 1. Each plaintiff and Timken filed a motion for judgment upon the agency record on February 8, 2007, which motions defendant opposes in the entirety and Timken, as defendant-intervenor, opposes with respect to certain claims. 3

Oral argument was held in camera on October 30, 2007. On June 18, 2008, the court requested additional briefing regarding certain matters, to which JTEKT, NPB, NSK, NTN, defendant, and Timken responded. See Letter from Timothy C. Stanceu, Judge, Ct. of Int’l Trade, to Counsel for Pis., Def., & Def.-Intervenor in Consol. Ct. No. 06-250 (June 18, 2008). In addition, defendant and Timken made five additional submissions, and defendant made one additional submission, to notify the court of supplemental authority.

II. Discussion

The court exercises jurisdiction pursuant to 28 U.S.C. § 1581(c), under which the Court of International Trade is granted exclusive jurisdiction of any civil action commenced under 19 U.S.C. § 1516a. 28 U.S.C. § 1581(c). The court reviews the Final Results on the basis of the agency record. See 28 U.S.C. § 2640(b) (2006); 19 U.S.C. § 1516a(b)(l)(B)(i) (2006). Upon such review, the court must “hold unlawful any determination, finding, or conclusion found,” 19 U.S.C. § 1516a(b)(l), “to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” Id. § 1516a(b)(l)(B)(i). “Substantial evidence is more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Consol. Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938).

A The Claims of JTEKT, NPB, NTN, and Ñachi Challenging the Department’s Zeroing Procedure Are Inconsistent with Controlling Judicial Precedent

Plaintiffs claim that the Department’s policy of zeroing in administrative reviews *1215 violates U.S. antidumping laws and is inconsistent with international obligations of the United States. Citing 19 U.S.C. §§ 1673, 1677(35), 1677b(a), and 1677Í-1, NTN and Ñachi argue that zeroing precludes a fair comparison of normal value and export price, distorts margins by failing to account for all transactions, and does not merit deference under Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43, n. 9, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), because Commerce is inconsistently applying zeroing under the statute. Rule 56.2 Mot. & Mem. for J. on the Agency R. Submitted on behalf of Pls. NTN Corp., NTN Bearing Corp. of Am., Am. NTN Bearing Mfg. Corp., NTN-BCA Corp., NTN-Bower Corp., & NTN Driveshaft, Inc. 8-10 (“NTN Mem.”); Br. of Pls. Nachi-Fujikoshi Corp., Nachi Am., Inc. & Nachi Technology, Inc. in Support of Rule 56.2 Mot. for J. on the Agency R. 9-10, 13-15, 18 (“Nachi Mem.”). JTEKT, NPB, and Nachi also argue that it is unreasonable for Commerce to continue to apply the zeroing practice under the Charming Betsy doctrine, under which the laws of the United States should be interpreted so as not to conflict with U.S. international obligations. Mem. of P. & A. in Support of Mot. of Pis. JTEKT Corp. & Koyo Corp. of U.S.A. for J. on the Agency R. 44-47 (“JTEKT Mem.”); Mem. in Support of the Mot. for J. upon the Agency R. Submitted by Pls. Nippon Pillow Block Co. Ltd. & FYH Bearing Units USA, Inc. 29-30 (“NPB Mem.”); Nachi Mem. 15; see Murray v. Schooner Charming Betsy, 6 U.S. (2 Cranch) 64, 118, 2 L.Ed. 208 (1804). In support of this argument, these plaintiffs point to recent reports from panels and the Appellate Body of the World Trade Organization (“WTO”) concluding that the Department’s zeroing practice is inconsistent with the WTO obligations of the United States. JTEKT Mem. 45-47; NPB Mem. 29-30; NTN Mem. 5-8; Nachi Mem. 15-16. Finally, JTEKT, NPB, and NTN claim that the Court of International Trade should remand the determination to permit Commerce to implement, or consider implementing, adverse reports of the WTO on the application of zeroing in administrative reviews. See JTEKT Mem. 47; NPB Mem. 30-31; NTN Mem. 10-11. Defendant rejects plaintiffs’ various arguments, arguing that the Court of Appeals for the Federal Circuit (“Court of Appeals”) repeatedly has sustained the Department’s treatment of nondumped sales under the statute despite findings set forth in adverse reports from the WTO’s Dispute Settlement Body (“DSB”). Def.’s Resp. in Opp’n to Pis.’ Mots, for J. upon the Agency R. 72 (“Def. Resp.”) (citing Timken Co. v. United States, 354 F.3d 1334 (Fed.Cir.2004), cert. denied 543 U.S. 976, 125 S.Ct. 412, 160 L.Ed.2d 352 (2004), and Corns Staal BV v. Dep’t of Commerce, 395 F.3d 1343 (Fed.Cir.2005), cert. denied 546 U.S. 1089, 126 S.Ct. 1023, 163 L.Ed.2d 853 (2006) (“Corus I”)); see also Resp. of Timken U.S. Corp. to the Rule 56.2 Mots, of JTEKT Corp., et al. 28-30 (“Def.-Intervenor Resp.”).

The court concludes that plaintiffs’ arguments challenging zeroing fail to raise new issues not already settled by binding precedent of the Court of Appeals. In Timken, the Court of Appeals held that Commerce “reasonably interpreted § 1677(35)(A) to allow for zeroing” in the context of administrative reviews and also explained that the “fair comparison” requirement, which applies to the calculation of normal value under § 1677b(a), which was incorporated by the Uruguay Round Agreements Act (“URAA”), did not affect its holding. Timken, 354 F.3d at 1343. The Court of Appeals also rejected the argument that Commerce must adhere to a single practice on zeroing for both investigations and administrative reviews. Id. at 1344-45. In NSK, the Court of Appeals *1216 emphasized that it repeatedly has upheld the practice of zeroing and again affirmed the Department’s determination to apply zeroing to administrative reviews despite the Department’s determination to cease doing so in investigations. NSK Ltd. v. United States, 510 F.3d 1375, 1379-80 (Fed.Cir.2007) (“NSK III”) (citing Timken, 354 F.3d 1334; Corus I, 395 F.3d 1343; and Corns Staal BV v. United States, 502 F.3d 1370 (Fed.Cir.2007) (“Corus II”)).

Regarding the application of the Charming Betsy doctrine, the Court of Appeals in Timken rejected the “fair comparison” argument that had been raised thereunder. Timken, 354 F.3d at 1343-44. The Court of Appeals rejected a plaintiffs argument that it was required “to interpret the ‘fair comparison’ language in the U.S. anti-dumping statute in a manner consistent with U.S. international obligations, thereby adopting the holding in EC — Bed Linen and finding Commerce’s zeroing practice an unreasonable statutory interpretation.” Id. at 1343. Observing that “[t]he crux of [respondent’s] argument hinges on the Charming Betsy canon of claim construction,” id., the Court of Appeals rejected the respondent’s position, explaining that “[w]hile [respondent] relies on EC — Bed Linen for its persuasive value in an effort to convince us of the unreasonableness of Commerce’s zeroing practice, we do not find it sufficiently persuasive to find Commerce’s practice unreasonable.” Id. at 1344.

Finally, JTEKT and NPB argue that Commerce must comply with the findings set forth in adverse reports from the WTO DSB. JTEKT Mem. 45-47; NPB Mem. 29-30. JTEKT requests that the court remand the determination to Commerce so that Commerce may determine whether and how to comply with the decision of the Appellate Body of the WTO holding that zeroing in administrative reviews is impermissible. JTEKT Mem. 47 & n. 16. JTEKT asserts that its position is consistent with the holding of the Court of Appeals in Corns I, 395 F.3d at 1349, because JTEKT does not ask the court to act but rather to remand the matter to Commerce for reconsideration in light of recent legal developments in the WTO. Id. at 47, n. 16. NTN argues, similarly, that the principles articulated in recent DSB decisions on the .use of zeroing in administrative reviews should persuade the court to hold that the Department’s zeroing practice is unlawful and points to the U.S. decision, reached under Section 123 of the URAA, 4 to comply with the DSB ruling that zeroing is impermissible in investigations (“Section 123 Determination”). NTN Mem. 5-8, 10-11. Defendant responds that the Section 123 Determination applies only to new and continuing investigations and therefore has no effect upon this administrative review. Def. Resp. 77; see Def.-Intervenor Resp. 33-34.

Section 123 delineates specific procedures for determining whether and how the United States will comply with deci *1217 sions of the WTO. See 19 U.S.C. § 3533(g) (2006). The Section 123 Determination states that it will apply to current and future investigations as of the effective date. Antidumping Proceedings: Calculation of the Weightedi-Average Dumping Margin During an Antidumping Investigation; Final Modification, 71 Fed.Reg. 77, 722, 77,725 (Dec. 27, 2006) (“Section 123 Determination ”). With respect to timing in this matter, the Final Results were issued on July 14, 2006 and the Section 123 Determination was issued on December 27, 2006, with an original effective date of January 16, 2007, and a final amended effective date of February 22, 2007. Final Results, 71 Fed.Reg. 40,064; Antidumping Proceedings: Calculation of the Weighted-Average Dumping Margins in Antidumping Investigations; Change in Effective Date of Final Modification, 72 Fed.Reg. 3783 (Jan. 26, 2007) (“Section 123 Determination Am. Effective Date”); Section 123 Determination, 71 Fed.Reg. 77,-722. Because the effective date of the Section 123 determination occurred after the issuance of the Final Results, there is no basis for the court to consider whether the legal principles adopted therein should apply to this administrative review. The court, therefore, need not reach plaintiffs’ other arguments on this issue.

Even if the court concluded that the timing of the Final Results did not preclude application of the Section 123 determination, the court could not hold in favor of plaintiffs on this issue. The Court of Appeals repeatedly and consistently has upheld as reasonable the Department’s statutory interpretation that zeroing is permissible in administrative reviews. The Court of Appeals in Corns II, 502 F.3d at 1375, upheld the use of zeroing in an administrative review of an antidumping duty order and in NSK III, 510 F.3d at 1380, expressly rejected the argument that use of zeroing should be held unlawful based on a decision of the DSB and on statements by the United States indicating that the United States would comply with that decision. 5 With respect to a remand to allow Commerce time to consider implementation of a new practice, the Court of Appeals has stated explicitly that it is not for the courts to implement WTO decisions without explicit instructions from the Executive Branch:

until Commerce abandons zeroing in administrative reviews such as this one, a remand in this case would be unavailing. Therefore, because Commerce’s zeroing practice is in accordance with our well-established precedent, until Commerce officially abandons the practice pursuant to the specified statutory scheme, we affirm its continued use in this case.

NSK III, 510 F.3d at 1380 (emphasis added). Subsequently, the Court of Appeals reiterated its position:

The determination whether, when, and how to comply with the WTO’s decision on “zeroing,” involves delicate and subtle political judgments that are within the authority of the Executive and not the Judicial Branch. Neither Commerce nor the Department of Justice has requested, or even suggested, such a remand. It would be most inappropriate for this court on its own to direct Commerce to reopen the Final Results of the 15th review to consider the impact on its decision of the subsequent WTO ruling, and we decline to do so.

*1218 Koyo Seiko Co. v. United States, 551 F.3d 1286, 1291 (Fed.Cir.2008) (“Koyo II ”); see Corus I, 395 F.3d at 1349 (directing the application of current U.S. law until the law changes to prohibit zeroing in administrative reviews). In summary, Court of Appeals precedent entirely precludes the court from requiring Commerce to change its zeroing practice in administrative reviews based on the Section 123 determination, and it forecloses a remand under which Commerce would be directed or authorized to reconsider its practice in light of WTO decisions declaring unlawful the practice of zeroing in the context of administrative reviews.

B. The Court Rejects Plaintiffs’ Various Claims Related to the Model-Match Methodology Except for NPB’s Claim Relating to Sampling Months and NTN’s Claim Relating to Commerce’s Rejection of NTN’s Proposed Bearing Design-Type Categories

In determining a dumping margin, Commerce compares the U.S. price of the subject merchandise with the price of comparable merchandise (the “foreign like product”) in the home market. 19 U.S.C. § 1677b (2006). To do so, Commerce first attempts to match sales of the subject merchandise with sales of identical merchandise in the home market. 19 U.S.C. § 1677(16)(A) (2006). In the absence of identical merchandise, Commerce attempts to match a sale of subject merchandise in the United States with a sale of similar merchandise in the home market. See id. § 1677(16)(B)-(C). If Commerce finds there are no sales of similar merchandise in the home market, Commerce will calculate a constructed value. Id. § 1677b(a)(4).

The model-match methodology is the means by which Commerce identifies similar merchandise. For the initial fourteen administrative reviews of the subject merchandise, Commerce applied a model-match methodology in which it compared bearings on the basis of eight characteristics (the “family model-match methodology”). Issues & Decision Mem. for the Antidumping Duty Admin. Reviews of Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, Singapore, & the United Kingdom for the Period of Review May 1, 2003, through April 30, 2004, at 19-26 (Sept. 16, 2005) (“AFBs 15 Decision Mem.”). The bearings that matched according to those eight characteristics were grouped in the same “family” for purposes of determining a foreign like product. Id. In the fifteenth administrative reviews of the bearings orders (“AFBs 15”), Commerce adopted a different methodology (“new model-match methodology”) in which Commerce applies a multi-step process. Id.; see Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, Singapore, & the United Kingdom: Final Results of Antidumping Duty Admin. Reviews, 70 Fed.Reg. 54,711, 54,712 (Sept. 16, 2005) (“AFBs 15 Final Results ”). Commerce applied this new model-match methodology in the AFBs 16 reviews. Decision Mem. 12-27.

In the new model-match methodology, Commerce first matches a ball bearing model sold in the United States to one sold in the home market according to the following four physical characteristics: load direction, bearing design, number of rows of rolling elements, and precision rating. AFBs 15 Decision Mem. 19. For bearing design, in AFBs 16 reviews, Commerce recognized the following eight design types: angular contact, self-aligning, deep groove, integral shaft, thrust ball, housed, insert, and hub units. Decision Mem. 77. A match requires consistency with respect to all four of these physical characteristics. AFBs 15 Decision Mem. 19. If there is such consistency, Commerce then identifies the most appropriate home market ball bearing model according to four addi *1219 tional, quantitative characteristics: load rating, outer diameter, inner diameter, and width. Id. Commerce excludes any potential matches in which the sum of the deviations in the four quantitative characteristics exceeds 40%. Id. Finally, Commerce excludes matches for which the Department’s difference-in-merehandise adjustment (“DIFMER”) exceeds 20%. See Decision Mem. 19 (“Because we applied our normal methodology of disregarding potential matches with a difference-in-merchandise adjustment of greater than 20 percent, we regard all the matches we actually made to be approximately equal in commercial value.”); Imp. Admin. Policy Bulletin 92.2 (July 29, 1992), http://ia.ita. doc.gov/policy/index.html (last visited Dec. 18, 2009).

Plaintiffs challenge the new model-match methodology on various grounds. Certain of the plaintiffs challenge specifics of the application of the methodology in the AFBs 16 reviews, including the Department’s decision to consider only eight physical characteristics, the refusal by Commerce to increase the number of months in which it searched for matches, the Department’s rejection of all but one of the proposed additional bearing design types, and the alleged unlawfulness of certain matches. For the reasons discussed below, the court rejects these various claims except for two claims, as discussed below, and affirms, in other respects, the model-match methodology as applied in the AFBs 16 reviews.

1. Contrary to the Claims of JTEKT, NPB, NSK, NTN, and Ñachi, the Court Concludes that Commerce Acted Lawfully in Deciding to Change its Model-Match Methodology

Plaintiffs raise various arguments relating to the standard Commerce must meet to change its model-match methodology. JTEKT and NTN claim that Commerce failed to give compelling reasons to change the methodology and that Commerce must use the prior methodology as a matter of fairness due to respondents’ reliance, NPB contends that Commerce failed to provide a reasoned explanation for its departure from past practice, and NSK claims that the Department’s determination to change methodologies was not reasonable because Commerce failed to show that the new methodology would yield a more accurate dumping margin. JTEKT Mem. 26-28; NTN Mem. 20-26; NPB Mem. 9; Mem. of P. & A. in Support of NSK’s Mot. for J. on the Agency R. 11-12 (“NSK Mem.”). Defendant argues that Commerce need only have acted reasonably in changing the methodology and that Commerce has done so. Def. Resp. 15.

JTEKT, NPB, and NTN argue, further, that substantial record evidence does not support the claim of greater accuracy, that the new methodology actually yields less accurate dumping margins than the predecessor, that it is not possible to select a single most similar model for a ball bearing, and that Commerce, in deciding that price-to-price comparisons are more accurate than constructed values, contradicted the position it took in earlier reviews. JTEKT Mem. 24-25, 29-38; NPB Mem. 13-18; NSK Mem. 11-15. JTEKT, NPB, and NTN also reject as unsupported by substantial record evidence the Department’s rationale that technology improvements enable Commerce to implement the new model-match methodology. JTEKT Mem. 39-40; NPB Mem. 11-13; NTN Mem. 22-23. Defendant and defendantintervenor urge the court to reject these arguments and uphold the new methodology as reasonable and, as applied in this case, supported by the record evidence. See Def. Resp. 15-21; see Def.-Intervenor Resp. 12-22.

The court will review a change in methodology for reasonableness. SKF USA, Inc. v. United States, 537 F.3d 1373, *1220 1377-78 (Fed.Cir.2008) (“SKF I”). The more demanding standards that plaintiffs advocate, i.e., that Commerce must set forth compelling reasons for the change or that the change must be demonstrated to produce a more accurate dumping margin, are not correct statements of the law. Id. at 1378 (stating that a review for reasonableness does not conflict with the substantial evidence standard). Further, the Court of Appeals previously has rejected arguments identical or similar to those advanced by plaintiffs, noting that “this statute ‘is silent with respect to the methodology that Commerce must use to match a U.S. product with a suitable home-market product.’ ’” Id. at 1379 (quoting Koyo Seiko Co. v. United States, 66 F.3d 1204, 1209 (Fed.Cir.1995) (“Koyo I ”)). Concluding that “Congress has granted Commerce considerable discretion to fashion the methodology used to determine what constitutes ‘foreign like product’ under the statute,” the Court of Appeals deferred to the Department’s choice of methodology as a reasonable construction of the antidumping statute. Id. (citing Pesquera Mares Australes Ltda. v. United States, 266 F.3d 1372, 1384 (Fed.Cir.2001), which cites, in turn, Koyo I, 66 F.3d at 1209). The Court of Appeals explained that Commerce was reasonable in seeking to improve accuracy, to select a model that would yield more priee-to-price comparisons, and to capitalize on technological advances that enable implementation of a more accurate methodology. Id. at 1380.

JTEKT, NPB, and NSK argue that the change in methodology impermissibly was applied retroactively despite plaintiffs’ reliance on the previous methodology. JTEKT Mem. 25, 41-43; NPB Mem. 18-22; NSK Mem. 17-22. NPB argues specifically that it relied upon the prior methodology for fourteen years and that principles of fairness preclude Commerce from changing the methodology. NPB Mem. 18-22. NSK contends that the longstanding agency practice carries the weight of law, explaining that the old methodology was well-established, confirmed as lawful by the courts, and unaffected by any intervening statute that would require modification of the methodology. NSK Mem. 19 (citing Nat’l Corn Growers Ass’n v. Baker, 840 F.2d 1547, 1555 (Fed.Cir.1988)). NSK further contends that it relied on this longstanding practice of applying the family model-match methodology and that it was not put on notice that the new methodology would apply in the review at issue. NSK Mem. 20-22. Defendant argues that the reliance arguments lack merit because Commerce had announced in earlier reviews that it was considering a revision of the methodology, provided respondents notice early in the proceedings of the determination to apply the revised methodology, and also provided an opportunity to comment on the determination. Def. Resp. 21-22; see Def.-Intervenor Resp. 23-27.

The Court of Appeals previously has rejected an argument that retroactivity and respondents’ reliance on the old methodology preclude Commerce from modifying its methodology in a current administrative review. SKF I, 537 F.3d at 1381. The Court of Appeals reasoned that there is an inherent retroactivity to antidumping review determinations and that a change in methodology, like any application of a methodology in an antidumping review, permissibly involves a retroactive effect. Id. The Court of Appeals also rejected the assertion of plaintiff SKF of detrimental reliance, explaining that “SKF does not dispute that Commerce has consistently found that SKF continues to sell at dumped prices” and that “SKF cannot properly analogize its situation to that in Shikoku, where ‘[t]he record contained] evidence that plaintiffs adjusted their prices in accordance with methodology consistently applied by Commerce in an attempt to comply with United States anti-dumping law.’ ” Id. (quoting Shikoku *1221 Chemicals Corp. v. United States, 16 CIT 382, 386, 795 F.Supp. 417, 420 (1992)). As in SKF, plaintiffs do not dispute that Commerce consistently has found that JTEKT, NPB, NSK, and NTN continue to sell at dumped prices.

In SKF, the Court of Appeals noted that Commerce had found that plaintiffs were selling at dumped prices for the periods of review prior to the one at issue, May 1, 2003 to April 30, 2004. Id. at 1377, 1381 (addressing the period of review from May 1, 2003 to April 30, 2004). Commerce had determined the following margins for SKF entities for the three periods of review preceding the one at issue in SKF: for the POR beginning May 1, 2002, SKF entities had margins of 1.38%, 2.49%, and 5.25%; for the POR beginning May 1, 2001, SKF entities had margins of 3.38%, 5.08%, and 6.70%; and for the POR beginning May 1, 2000, SKF entities had margins of 3.70% and 8.51%. Antifriction Bearings & Parts Thereof From France, Germany, Italy, Japan, Singapore, & the United Kingdom: Final Results of Antidumping Duty Admin. Reviews, Rescission of Admin. Reviews in Part, & Determination To Revoke Order in Part, 69 Fed.Reg. 55,574, 55,578-80 (Sept. 15, 2004) (“Final Results POR 2002-03”); Ball Bearings & Parts Thereof From France & Japan; Am. Final Results of Antidumping Duty Admin. Reviews, 68 Fed.Reg. 43,712, 43,712 (July 24, 2003) (“Am. Final Results POR 2001-02”); Ball Bearings & Parts Thereof From France, Germany, Italy, Japan, & Singapore: Final Results of Antidumping Duty Admin. Reviews, Rescission of Admin. Review in Part, & Determination Not To Revoke Order in Part, 68 Fed.Reg. 35,623, 35,625 (June 16, 2003) (“Final Results POR 2001-02 ”); Ball Bearings & Parts Thereof From France, Germany, Italy, Japan, & the United Kingdom; Final Results of Antidumping Duty Admin. Reviews, 67 Fed.Reg. 55,780, 55,781 (Aug. 30, 2002) (“Final Results POR 2000-01 ”).

The period of review for the AFBs 16 reviews is May 1, 2004 through April 30, 2005. The court notes that Commerce determined in prior reviews that plaintiffs JTEKT, NPB, NSK, and NTN sold at dumped prices. For example, Commerce determined the following margins for plaintiffs for the three periods of review preceding the one at issue in this action: for the POR beginning May 1, 2003, 12.78% for Koyo (JTEKT), 15.51% for NPB, 8.25% for NSK, and 5.93% for NTN; for the POR beginning May 1, 2002, 5.56% for Koyo (JTEKT), 3.37% for NPB, 2.46% for NSK, and 2.74% for NTN; and for the POR beginning May 1, 2001, 4.98% for Koyo (JTEKT), 4.82% for NPB, 2.68% for NSK, and 4.51% for NTN. Notice of Correction to Am. Final Results of Anti-dumping Duty Admin. Review: Ball Bearings & Parts Thereof from Japan, 70 Fed.Reg. 69,316, 69,316 (Nov. 15, 2005) (“Corrected Am. Japan Final Results POR 2003-0P’); Notice of Am. Final Results of Antidumping Duty Admin. Reviews: Ball Bearings & Parts Thereof from Japan, 70 Fed.Reg. 61,252, 61,252 (Oct. 21, 2005) (“Am. Japan Final Results POR 2003-Op’); Ball Bearings & Parts Thereof from France, Germany, Italy, Jar pan, Singapore, & the United Kingdom: Final Results of Antidumping Duty Admin. Reviews, 70 Fed.Reg. 54,711, 54,713 (Sept. 16, 2005) (“Final Results POR 2003-Op’); Final Results POR 2002-03, 69 Fed.Reg. at 55,580; Am. Final Results POR 2001-02, 68 Fed.Reg. at 43,712; Final Results POR 2001-02, 68 Fed.Reg. at 35,625. In the AFBs 16 reviews, none of the plaintiffs actually has demonstrated detrimental reliance on the old methodology-

For the reasons stated above, the court will affirm the Final Results with respect to the Department’s decision to depart from the previous model-match methodology-

*1222 2. The Court Rejects NPB’s Claim that Commerce Must Include Additional Physical Characteristics in its Model-Match Methodology

While contesting the adoption of the new model-match methodology, NPB argues in the alternative that Commerce should have included additional physical characteristics in the methodology to account for certain specialty bearings. NPB Mem. 22. NPB explains that its specialty bearings should not be considered to be foreign like products of standard bearings sold in the United States and identifies seven additional characteristics that it submits Commerce, at a minimum, should have applied when determining matches. Id. at 23-24. Specifically, NPB argues that Commerce should have incorporated the following additional physical characteristics into its methodology: “types of seals,” “greased v. ungreased,” “ceramic v. nonceramic,” “diameter of second inner dimension,” “diameter of second outer dimension,” “diameter of second width dimension,” and “diameter of third width dimension.” Id. at 24. NPB argues that Commerce impermissibly rejected NPB’s request to include these additional characteristics, challenging the Department’s finding that NPB submitted the request too late in the proceedings and the finding by Commerce that NPB did not explain how Commerce could implement the proposed additional physical characteristics in the methodology. Id. at 25-26.

In their oppositions to NPB’s motion for judgment upon the agency record, neither defendant nor defendant-intervenor addressed the issue of NPB’s proposal for additional physical characteristics. See Def. Resp.; Def.-Intervenor Resp. NPB urges the court to enter judgment against defendant, relying on USCIT Rule 56(e)(2). Reply Br. in Supp. of the Rule 56.2 Mot. for J. upon the Agency R. Submitted by Pis. Nippon Pillow Block Co., Ltd. and FYH Bearing Units USA, Inc. 2-3 (“NPB Reply”) (citing, inter alia, Taylor v. City of N. Y., 269 F.Supp.2d 68, 75 (E.D.N.Y.2003), and Precision Specialty Metals, Inc. v. United States, 24 CIT 1016, 1016, 116 F.Supp.2d 1350, 1353 (2000)). However, USCIT Rule 56(e)(2) is inapposite in a case where the motion for judgment upon the agency record is made pursuant to USCIT Rule 56.2. Under USCIT Rule 56.2, defendant is deemed to have opposed the issue whether or not the opposition was specifically pled. Therefore, the court rejects NPB’s argument under US-CIT Rule 56(e)(2) that the court should enter judgment against defendant on this issue.

Regarding the timing of NPB’s proposal to incorporate additional physical characteristics in the model-match methodology, Commerce stated in its Decision Memorandum that NPB was the only respondent to suggest additional physical characteristics and did not do so until its case brief, which NPB submitted on April 25, 2006, less than three months before the issuance of the Final Results. Decision Mem. 18; Final Results, 71 Fed.Reg. 40,064. Commerce explained that respondents must submit their arguments regarding physical characteristics earlier in the review, rather than after the Preliminary Results, in order for Commerce to have a reasonable opportunity to consider them and to allow interested parties to comment. Decision Mem. 23. NPB concedes that in the subject reviews, it first proposed the use of additional physical characteristics in its case brief. 6 NPB Mem. 25. Instead, NPB *1223 relies on 19 C.F.R. § 351.309(c)(2), which requires a party to submit all of its arguments in its case brief, including arguments presented before the publication of the preliminary results. NPB Mem. 26. NPB adds that it raised the same arguments in the prior review. Id. Finally, NPB rejects the Department’s argument that interested parties would not have an opportunity to comment and argues instead that parties wishing to comment could have done so in their rebuttal briefs. Id.

NPB’s reliance on 19 C.F.R. § 351.309(c)(2) is misplaced. In this regulatory provision, Commerce requires parties to submit all arguments in the case brief, including those already before the agency. See 19 C.F.R. § 351.309(c)(2) (2009). The regulation ensures that Commerce has before it all of a party’s arguments for consideration prior to the Final Results. It does not guarantee that Commerce will adopt the positions suggested therein, nor does it preclude Commerce from deciding, according to specific circumstances, that a change in methodology advocated at that time would not be practicable.

The Court of Appeals has held that “Commerce has considerable discretion in defining ‘identical in physical characteristics.’ ” Pesquera Mares, 266 F.3d at 1384 (quoting 19 U.S.C. § 1677(16)(A)). Commerce exercised that discretion in requiring NPB and other respondents to submit questionnaire responses according to the physical characteristics Commerce had identified. Letter from Baker & McKenzie, LLP to Dep’t of Commerce 570-72 (Sept. 27, 2005) (Admin.R.Doc. No. 92) (“NPB Questionnaire Resp.”) (stating, at B-2 through B-4 of the submission, that NPB reported its bearings according to the eight product characteristics indicated by Commerce and also added several fields). NPB was aware of those physical characteristics at the time of submitting its questionnaire responses and could have advocated in a questionnaire response that Commerce adopt its detailed proposal. See id. It would have been preferable had Commerce notified parties that proposals for use of additional physical characteristics in the model-match methodology would not be considered if submitted after a specified date. Nevertheless, the Department’s apparent failure to do so in this case does not compel the court to conclude that Commerce acted contrary to law in rejecting NPB’s proposal for reasons of impracticability due to time constraints and the desirability of allowing meaningful comment by other parties. At or about the time NPB submitted its questionnaire response, NPB reasonably should have expected that Commerce, absent an objection lodged at this earlier stage of the reviews, would proceed with its analysis according to the physical characteristics that Commerce had identified. Commerce explained that there were practical issues with the additional characteristics proposed by NPB that Commerce could not resolve at a point so late in the review, i.e., after Commerce had issued the Preliminary Results. Decision Mem. 23. Commerce also explained that it had an insufficient time period in which to determine how to amend its methodology to incorporate the suggested additional physical characteristics and to allow interested par *1224 ties the opportunity to comment. Id. at 23-24.

NPB having failed to raise its proposal sooner, the court cannot conclude that Commerce erred in determining that it did not have enough time to review, and possibly implement, the proposal of NPB to use additional physical characteristics in the model matching process. Due to the statutory time constraints under which Commerce must conduct its administrative reviews, Commerce must be considered to have a measure of discretion in deciding whether a modification of its model-match methodology is practicable. That NPB had raised the same issue in a prior review does not alter the court’s conclusion. Commerce could not know whether NPB intended to submit its proposal in the subject reviews until NPB actually did so, which NPB first did in its case brief.

For these reasons, the court concludes that the Department’s decision in the Final Results to decline to consider NPB’s proposal for additional physical characteristics was supported by substantial record evidence and adequate reasoning and was within the Department’s discretion over procedural questions relevant to the conduct of its administrative reviews. The court, therefore, affirms that decision.

3. The Department’s Decision Rejecting NPB’s Proposal to Expand the Choice of Sample Months Misconstrues the Department’s Regulation

NPB claims that Commerce, when searching for a ball bearing model to match with a particular subject ball bearing that a respondent sold in the United States, should have expanded the time period surrounding the date of the U.S. sale during which Commerce searched that respondent’s home market sales data. NPB Mem. 27. NPB argues that Commerce, by confining its search for possible matches to the “sampled months” immediately preceding and following the month in which the target U.S. sale occurred, unnecessarily reduced the number of identical matches and increased the number of matches for which only similar, but not identical, merchandise was involved. Id. NPB argues that the Department’s regulations do not require such a limited time period for the choice of sampled months and that the regulations allow Commerce to use a longer time period, such as a five-month time period, as NPB advocated. Id. at 28.

In the AFBs 16 reviews, the “sampled months” Commerce selected were February, June, August, September, and November of 2004, and February, March, and May of 2005. Prelim. Results, 71 Fed.Reg. at 12,174. According to the Department’s explanation of its method of sampling transactions in the Preliminary Results, see 19 U.S.C. § 1677f-l(a), Commerce resorted to sampling of transactions for respondents who made more than 10,-000 constructed export price (“CEP”) sales transactions in the United States, or more than 10,000 home market sales transactions, during the period of review. Prelim. Results, 71 Fed.Reg. at 12,172-74. Commerce explained that for a respondent who engaged in more than 10,000 CEP sales of merchandise subject to a particular order, Commerce selected “sample weeks” — one week from each two-month period during the period of review, for a total of six weeks — and then reviewed the individual U.S. CEP sales transactions that occurred during those sample weeks. Id. at 12,172. Commerce further explained that for home market sales, Commerce selected as sampled months the months corresponding to the sample weeks selected for CEP sales in addition to a sampled month prior to the period of review and a sampled month following the period of review. Id. at 12,173-74. For NPB, Commerce “analyzed CEP sales of ball bearings in sample weeks because NPB had more than 10,000 CEP transactions of BBs [ie., ball bearings] during the *1225 POR” and “analyzed HM [i.e., home market] sales using sample months because NPB made over 10,000 POR home-market transactions.” Mem. from Int’l Trade Compliance Analyst, AD/CVD Operations 5, to The File 2 (Mar. 2, 2006) (Admin.R.Doc. No. 220). Commerce stated that for model match purposes it first searches the home market sales during the sampled month corresponding to the sampled week in which the U.S. sale occurred. Decision Mem. 86. Next, Commerce searches the preceding home-market sampled month, which, because not all months are sampled months, could result in a total span of several months. Id. Finally, Commerce searches the home market sampled month subsequent to the target sampled month, which, in these administrative reviews, also could result in a span of several months. Id. (stating that Commerce “matched each bearing model sold in the United States in one of the sampled weeks to the corresponding home-market month, then to the immediately preceding sampled home-market month, and then to the immediately following sampled home-market month”); see Prelim. Results, 71 Fed.Reg. at 12,-172-74.

In the Decision Memorandum, Commerce reported its decision not to include additional sampled months as proposed by NPB. Decision Mem. 86. Commerce based its decision on 19 C.F.R. § 351.414(e)(2), which addresses the selection of the “contemporaneous months,” ie., the months to which Commerce limits the weighted averaging of prices in home market sales for use in making a comparison with an individual U.S. sale. 7 See Decision Mem. 86. Commerce stated in the Decision Memorandum its position that “[t]he contemporaneous months should not pass beyond the most recent of the three months prior or the two months subsequent to the month of the U.S. sale in which there was a sale of a foreign like product.” Id. (citing 19 C.F.R. § 351.414(e)(2)). “Given the fact that sample home-market months are separated by a month or more between each other in either direction, extending the window period by a month in each direction often results in extending the window period beyond the time period our regulation allows.” Id. If, for example, Commerce were to match a bearing sold in the United States with sales of a matching foreign like product that occurred in August 2004, 8 *1226 Commerce first would search home market sales in the target sampled month, August 2004; second, in the preceding sampled month, June 2004; and third, in the following sampled month, September 2004. As a result, subject bearings sold during a sample week in August 2004 potentially could be matched to home market sales in a sampled month that could only have occurred during the four-month period of June through September 2004. Were Commerce also to search for matching sales in the additional preceding and following sampled months, in this case February and November of 2004, that time period would be ten months long.

Neither defendant nor defendant-intervenor addressed NPB’s “sampled month” issue in their oppositions to plaintiffs’ motion for judgment upon the agency record. See Def. Resp.; Def.-Intervenor Resp. The court will not rule in favor of NPB on that basis. As discussed previously, the United States is deemed to have opposed summary judgment on an issue raised in a Rule 56.2 motion for judgment upon the agency record. See USCIT Rule 56.2.

In considering the sampling month issue on the merits, the court concludes that the premise underlying the Department’s analysis, as stated in the Decision Memorandum, was in error. Commerce erroneously concluded that its regulation required it to reject NPB’s proposal. Decision Mem. 86 (finding that “[gjiven the fact that sample home-market months are separated by a month or more between each other in either direction, extending thé window period by a month in each direction often results in extending the window period beyond the time period our regulation allows”). But that was not the case. The regulation in question, § 351.414(e)(2), provides that “[n formally, the Secretary will select as the contemporaneous month the first of the following which applies.” 19 C.F.R. § 351.414(e)(2) (emphasis added). Under the plain meaning of the regulation, Commerce is not precluded entirely from choosing as the “contemporaneous month” a sampled month outside of the total time span contemplated by § 351.414(e)(2). Instead, Commerce is free to exercise its discretion not to follow the normal procedure set forth in that provision. The regulation allows for an atypical circumstance under which it may be reasonable or appropriate to depart from the normal procedure. 9 The court notes that § 351.414(e)(2) is a general provision that does not address specifically the special circumstances in which Commerce resorts to sampling under 19 U.S.C. § 1677f-l(a).

In identifying the error in the Department’s analysis, the court does not hold or imply that Commerce, on the administrative record before it, was required to accept NPB’s proposal to expand the choice of sampling months. NPB may be correct that the proposal it advanced would have increased the accuracy of the model-match methodology; the review of a larger data base presumably would lead to more identical matches. Nevertheless, Commerce has considerable discretion in choosing the timing for conducting the comparison called for in the average-to-transaction method that normally applies during administrative reviews. See 19 C.F.R. § 351.414(c), (e). NPB’s proposal may or may not have been practicable in *1227 the context of the subject reviews and in this case may have caused Commerce to exceed the time period recommended by its regulation. But in this case, Commerce did not reject NPB’s proposal for considerations of practicability. Instead, it erroneously treated the method set forth in that regulatory provision as an ironclad requirement. The court, therefore, must remand for reconsideration the Department’s decision to reject NPB’s proposal to expand the choice of sampled months. The court must review the Department’s decision on the basis of the reasoning that Commerce put forth. See Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 285-86, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) (stating that the court “may not supply a reasoned basis for the agency’s action that the agency itself has not given”). In this circumstance, the reasoning Commerce set forth in the Decision Memorandum cannot be reconciled with the plain meaning of 19 C.F.R. § 351.414(e) and the discretion Commerce is granted thereunder.

I. Commerce Failed to Provide an Adequate Explanation for its Rejection of NTN’s Proposal for Additional Ball Bearing Design Types

NTN claims that Commerce erred in refusing to recognize and apply the additional ball bearing design types that NTN proposed for use in the model matching process. See NTN Mem. 26-30. NTN proposed designation of additional design types, all but one of which Commerce rejected. Decision Mem. 76-80. NTN argues that the ball bearing design types that Commerce identified, i.e., angular contact, self-aligning, deep groove, integral shaft, thrust ball, housed, and insert, are overly broad and fail to account for significant physical characteristics. NTN Mem. 27. In the AFBs 16 reviews, Commerce adopted, in response to NTN’s objection, only one additional design type, “hub units incorporating angular contact bearings.” Decision Mem. 77 (“We do find, however, that NTN provided evidence ... that demonstrates that NTN’s hub units incorporating angular contact bearings are significantly different from standard angular contact bearings as well as housed bearings to warrant a bearing-design designation distinct and separate from the seven bearing-design types we identified in our questionnaire.”). Commerce took the position that to include an additional design type in its model matching process, it had “to be satisfied that the classification is substantially different from each of the design types” already included. Id. Commerce concluded that NTN had not demonstrated the need for separate classifications for the other design types NTN proposed. Id. (stating that “NTN did not provide compelling evidence that each of its reported bearing-design types was so significantly different that it merited its own classification, distinct and separate from the seven bearing-design types we identified in our questionnaire”). Commerce also “found that, on many occasions, NTN’s numerous bearing-design designations were distinguishable due to a single element of difference or an element of difference that is not pertinent, such as a different width of inner race or the type of bore, ... etc.,” id., and that many of the differences correlate directly with load ratings, physical dimensions, and occasionally, the precision rating of bearings, which are characteristics that Commerce already uses to distinguish dissimilar products in its methodology. Id. at 77-78. In response to NTN’s arguments in the reviews that the function or application of different bearings warrant a separate design designation, Commerce stated that

it is the rolling element that is dispositive as to whether a bearing can be considered similar with respect to the purposes for which bearings are used *1228 (e.g., a ball bearing cannot be considered similar to a cylindrical roller bearing under any circumstance), not whether a specific application for one bearing differs from the specific application of another.

Id. at 78 (internal quotation marks and citation omitted). Commerce explained that “[bjecause all bearing models designated by NTN as different design types have a ball as a rolling element and fall within one of our seven bearing-design designations,” Commerce “consider[s] them equally similar in component material or materials for model-match purposes, as instructed by section 771(16)(B)(ii) of the Act [i.e., 19 U.S.C. § 1677(16)(B)(ii) ].” Id. at 77.

NTN maintains that its proposed additional design types, as presented and discussed in NTN’s questionnaire response and case brief, were required to avoid unreasonable comparisons of dissimilar products. NTN Mem. 28. NTN further argues that it set forth detailed information in its questionnaire response demonstrating the need for these design types. Id. at 28-29. Finally, NTN argues that it relied on the Department’s longstanding practice of using NTN’s design types in the previous methodology and that it was improper for Commerce to change its methodology. Id. at 29-30.

Defendant responds that NTN, in response to the design-type portion of the questionnaire, provided general descriptions of its design types, technical drawings, and brochures but did not explain the submissions or explain why each of the design types merited designation separate from the general design types that Commerce included in the questionnaire. Def. Resp. 25. Defendant argues that Commerce has broad discretion to develop a methodology for determining “like” or “similar” merchandise under 19 U.S.C. § 1677(16)(B). Id. at 26. According to defendant, NTN also failed to show that the claimed differences in design type were so substantial that the DIFMER adjustment would not account for the differences. Id. at 27-29; see Def.-Intervenor Resp. 52-53. Pointing to its questionnaire responses, NTN disputes defendant’s statement that NTN failed to provide the information, explanation, and analysis necessary to substantiate its request. Reply Br. of Pis. NTN Corp., NTN Bearing Corp. of Am., Am. NTN Bearing Mfg. Corp., NTN-BCA Corp., NTN-Bower Corp., & NTN Driveshaft, Inc. 3-4 (“NTN Reply”).

The Court of Appeals in Koyo II, 551 F.3d 1286, confronted an issue similar to that posed by NTN’s proposal for additional design types. In Koyo II, the plaintiff — which was NTN in that case as well-argued that its design types were significantly different from each other and that it had relied reasonably on the Department’s acceptance of additional design types because Commerce had accepted its design types in prior reviews. Id. at 1292. The Court of Appeals held that “NTN has not demonstrated that Commerce’s choice of design types, including its adjustments, was unreasonable” and that “even if Commerce had accepted NTN’s proposals in the past, it was not required to do so in future reviews.” Id. The Court of Appeals explained that “NTN’s claim that its design types are superior does not show that Commerce’s use of its own types was unreasonable.” Id.

The court observes that some of the additional design-type categories proposed by NTN appear to describe ball bearings that fall within one, and only one, of Commerce’s accepted design-type categories. However, the court also observes that some of the bearings described in NTN’s proposal for additional design types appear to fall within more than one of the Depart- *1229 merit's design-type categories. NTN raises this specific objection in support of its claim that Commerce should not have rejected its proposal. See NTN Mem. 28 (“Commerce’s design codes do not take into account bearings, which fall into more than one category, such as bearings that are both ‘angular contact’ and ‘deep groove’.... ”). Defendant does not respond to this issue in its briefs, nor is this issue addressed in the Decision Memorandum. Without a suitable explanation, the court is not able to discern how Commerce applied its model matching methodology to those of NTN’s bearings that appear to fall within more than one accepted design type. The answer to this question is relevant to the court’s consideration, in the entirety, of Commerce’s decision to reject all of NTN’s proposed design types other than the design-type category for hub units incorporating angular contact bearings. The court, therefore, will remand for reconsideration Commerce’s decision in the Final Results to reject NTN’s additional design types.

5. JTEKT, NPB, and NSK Failed to Demonstrate that Certain Identified Ball Bearing Matches Were Unlawful

At the invitation of the court, several plaintiffs filed supplemental briefs that set out, inter alia, certain matches they considered to be unreasonable. Supplemental Br. of Pis. JTEKT Corp. & Koyo Corp. of U.S.A. 2-3, Exs. I-K (“JTEKT Supplemental”); Pis. Nippon Pillow Block Co. Ltd. & FYH Bearing Units USA, Inc. Supplemental Br. 5-8 (“NPB Supplemental”); NSK’s Supplemental Mem. in Supp. of Mot. for J. on the Agency R. 4-5 (“NSK Supplemental”); NTN Corp., NTN Bearing Corp. of Am., Am. NTN Bearing Mfg. Corp., NTN-BCA Corp., NTN-Bower Corp., & NTN Driveshaft, Inc.’s Supplemental Br. Pursuant to the Ct.’s Letter of June 18, 2008, at 3-7 (“NTN Supplemental”). JTEKT, NPB, NSK, and NTN assert that the methodology implemented by Commerce in the AFBs 16 reviews resulted in unlawful matches.

JTEKT identified three matches that it views are unreasonable. JTEKT Supplemental 2-3. The exhibits attached to its brief set forth a chart comparing physical or commercial characteristics of the U.S. model to one or two home market models. Id. at Exs. I-K. With respect to the first match, JTEKT states that “[b]ecause the purpose and applications of these models are quite different, there are remarkable differences between the U.S. model and the two home market models in terms of appearance and specifications.” Id. at Ex. I, at 2. JTEKT asserts that “the home market models require higher specifications to ensure their performance than the U.S. model.” Id. JTEKT further asserts that “[djespite these differences, the Department matched the U.S. model with these home market models based on their coincidentally similar costs of production.” Id. The second and third matches identified by JTEKT as unreasonable rely on substantively similar allegations. See id. at Exs. J-K. JTEKT’s arguments essentially identify physical characteristics that were not separately recognized within the Department’s new model-match methodology. JTEKT, however, has not demonstrated generally that the Department’s matches were unreasonable according to the physical characteristics applied by the Department’s methodology, or why that methodology, through the application of the DIFMER, failed to address the asserted “remarkable differences between the U.S. model and the two home market models in terms of appearance and specifications.” See id. at Ex. I, at 2.

NPB also failed to demonstrate that Commerce applied an unreasonable methodology or exceeded its discretion in matching certain bearings. NPB states that “any number of unlawful comparisons *1230 with inappropriate foreign like products” occurred, NPB Mem. 26, and explains that Commerce “compared high temperature bearings NPB sold in the Japanese market to standard bearings that it sold in the U.S. market.” Id. NPB explained that its “high temperature bearings contain heatproof grease, special heat-proof seals, and may contain ceramic bearings” while its standard bearings “have standard grease, standard seals and stainless steel bearings.” NPB Supplemental 5. NPB provides two examples of matches it considers unreasonable. First, NPB contends that Commerce “incorrectly compared the sales of NPB’s U.S. model UC 211-32 G5 made in November 2004, to sales of Japanese model NA 211-32, also made in November 2004,” arguing that Commerce “should have used the alternative model UC 211, a highly similar model to the UC 211-32 G5,” which “has the same set screw design” and “differs only in that its inner diameter is slightly larger (55 mm. for the UC 211 versus 50.8 mm. for the UC 211-32 G5).” Id. at 6. NPB states that the “appropriate comparison model” was used in other sample months but that “the NA 211-32 was preferred in November only because it had the same inner diameter as the U.S. model.” Id. NPB acknowledges, however, that “[t]he fundamentally different design of the NA 211-32 is reflected in the adjustment for the difference in merchandise vis-á-vis the UC 211-32 G5,” which “adjustment is 18% of the total cost of production of the UC 211-32 G5, just short of the 20% that the Department uses as the maximum level for permissible comparisons.” Id. at 6-7. Nonetheless, NPB asserts that Commerce should have used the other match, which yielded “a difference in merchandise [ie., DIFMER] of only 1%” and a percentage margin of 9%, as opposed to 84%. Id. at 7. NPB alleges, with respect to the first match, that the “single incorrect comparison by the Department contributes about one-tenth of the overall dumping margin.” Id. at 6. NPB advances another match it considers unreasonable, asserting that the two models had fundamentally different designs, differed in inner diameter, and were sold in different months. Id. at 7. According to NPB, this match “contributes several percent to the dumping margin.” Id.

NPB did not explain why the court must conclude that the matches to which it objects were unlawful. The existence of matches that NPB considers best do not compel the court to conclude that Commerce erred. As NPB acknowledged, “[tjhese specific model match comparisons highlight the larger issue ... that is, additional physical characteristics are necessary to select the single most similar model matches.” Id. at 8. As the court concluded above, however, Commerce was reasonable in limiting its methodology to the physical characteristics upon which it relied in conducting the administrative review at issue. The existence of isolated examples of matches that NPB asserts are preferable than the matches Commerce made pursuant to its methodology does not require the conclusion that the Department’s methodology is unreasonable or that Commerce exceeded its discretion in making those matches. NPB fails to allege that the model-match methodology was misapplied by Commerce, and as applied to the specific matches to which NPB objects, the court cannot conclude that the results of the methodology were unreasonable.

NSK argues that “gross dissimilar matches infect much of the database used to calculate NSK’s antidumping duty margin.” NSK Mem. 13. NSK identified “two matches of dissimilar merchandise that the family methodology would have deemed unacceptable.” NSK Supplemental 2. NSK highlights differences in load rates, inner diameters, and outer diame *1231 ters, which are physical characteristics for which the new model-match methodology permits a 40% sum-of-the-deviations limit. Id. at 2-3. NSK also highlights the different uses of the matched bearings. Id. at 2. NSK’s discussion of these examples does not present a reason for the court to conclude that Commerce exceeded its discretion or otherwise acted contrary to law in applying its new model-match methodology. Rather, the matches are offered only to illustrate, in nonspecific terms, the alleged superiority of the prior family model-match methodology. As the court stated above, however, the Court of Appeals has affirmed the Department’s methodology as reasonable. SKF II, 537 F.3d at 1379-80. As affirmed by the Court of Appeals, the model-match methodology incorporated the 40% sum-of-the-deviations feature that NSK finds specifically objectionable. Id. at 1379. NSK’s argument that it has presented “an extensive list of dissimilar bearing matches” in an affidavit it submitted with its motion for judgment upon the agency record is similarly unconvincing. NSK Supplemental 3 (citing NSK Mem., Ex. 12).

NTN also contests several matches: a specialized, farm implement bearing to a differently specialized set screw type bearing; a bearing with a pillow block housing to a bearing with square or round flange housing; a bearing with a dust cover to one without a dust cover; a bearing with a pillow-type housing to a bearing with a cartridge-type housing; and an eccentric locking collar bearing that has a spherical outer diameter to a set screw bearing that has a straight outer diameter. NTN Supplemental 3-9. In highlighting certain matches as unreasonable, NTN is challenging implicitly the reasonableness of the methodology itself. The variations that NTN highlights among the matched bearings result from differing physical characteristics or design types that are not recognized by the Department’s model-matching methodology as applied in the AFBs 16 reviews. Thus, NTN’s basic objection is to Commerce’s choice of physical characteristics and design types. NTN argues that “[t]he sum-of-the-deviations may only be taken into consideration, however, after the bearing design has been correctly considered by the Department” and “the [DIFMER] adjustment is only relevant for comparison between similar bearings, that is, once the bearings have been determined to be of the same design type, appropriately defined.” Id. at 4-5. Stating its argument in general terms, NTN submits that “the first decision in the Department’s model match methodology must be whether the paired bearings are alike in physical characteristics and function.” Id. at 5. As discussed above, the court is remanding the Final Results for reconsideration of the decision to reject NTN’s proposed additional design types. The Department’s redetermination may affect specific matches.

The court again recalls that “Congress has granted Commerce considerable discretion to fashion the methodology used to determine what constitutes ‘foreign like product’ under the statute,” id. at 1379 (citing Pesquera Mares, 266 F.3d at 1384), and that “Commerce’s interpretation of the statute [with respect to the new model match methodology] merits deference .... ” Id. In alleging that certain matches were unreasonable, plaintiffs make a number of general arguments: they elevate the importance of certain physical characteristics beyond that which Commerce concluded was appropriate in its methodology, see, e.g., JTEKT Supplemental 2-3, Exs. I-K; they set forth alternative matches that they contend would have been better matches, see, e.g., NPB Supplemental 5-8; and they argue that the matches yielded by the family model-match methodology were more accurate, see, e.g., NSK Mem. 13, Ex. 12 and NSK *1232 Supplemental 2-3. However, because the Department’s methodology has been found to be reasonable and because Commerce has a significant degree of discretion when deciding which additional design types it will apply in its model-match methodology, the court will not deem particular matches unreasonable on the basis of physical characteristics unless Commerce abused its discretion or otherwise acted unreasonably. Overall, the Department’s model matching decisions must be based on findings supported by substantial record evidence, and they must be reasonable determinations of foreign like products. Commerce is not required to refrain from making matches of bearings that are not identical, so long as it is reasonable to conclude that a home market bearing is sufficiently similar to the subject bearing to satisfy the reasonableness requirement of 19 U.S.C. § 1677(16)(B)-(C).

As discussed above, the Court of Appeals held that the Department’s new methodology is lawful. SKF II, 537 F.3d at 1379-80. A respondent’s showing that a particular match resulting from the new methodology would not have sufficed under the old methodology is not a sufficient basis upon which the court can conclude that Commerce acted unlawfully. With the possible exception of a problem posed by overlapping bearing design types, as discussed in the prior section with respect to NTN’s merchandise, plaintiffs have not demonstrated that the new methodology yielded unreasonable matches. Commerce applied a general methodology affirmed by the Court of Appeals as reasonable. Id. The methodology affirmed by the Court of Appeals was highly similar to that used in the AFBs 16 reviews in physical characteristics, the 40% sum-of-the-deviations limit, and the 20% DIFMER limit. See Decision Mem. 16-27, 76-80; AFBs 15 Decision Mem. 19, 24-25.

. In summary, the court will order that Commerce, upon remand, reconsider its rejection of NPB’s proposal to expand the sample months in which Commerce searches for home market matches and its rejection of NTN’s proposal for additional design-type categories for ball bearings but will affirm the use of the new model-match methodology in the AFBs 16 reviews in other respects.

C. Commerce Lawfully Decided to Treat JTEKT and Its Affiliate as a Single Entity

Commerce determined it appropriate to treat as a single entity (i.e., “collapse”) JTEKT and another producer of subject merchandise with which JTEKT is affiliated. Decision Mem. 61-63; Final Results, 71 Fed.Reg. at 40,065 (incorporating the Decision Memorandum). JTEKT claims a lack of substantial evidence to support this determination. JTEKT Mem. 11. To the contrary, the court concludes that the record contains substantial evidence to support the findings underlying Commerce’s determination to treat JTEKT and its affiliate as a single entity in the AFBs 16 reviews.

Under its regulations, Commerce will treat two or more affiliated producers as a single entity when two conditions are satisfied. The first condition is that the producers at issue “have production facilities for similar or identical products that would not require substantial retooling of either facility in order to restructure manufacturing priorities.... ” 19 C.F.R. § 351.401(f)(1) (2009). The second condition is met if Commerce concludes that “there is a significant potential for the manipulation of price or production.” Id.

JTEKT does not contest that it is affiliated with the other producer for purposes of 19 U.S.C. § 1677(33) (which defines the term “affiliated persons”) and 19 C.F.R. § 351.401(f)(1), nor does it take issue with *1233 the finding of fact by Commerce, see Decision Mem. 58, that JTEKT and its affiliated producer “produce similar or identical merchandise.” JTEKT Mem. 13. JTEKT argues, instead, that the second criterion in 19 C.F.R. § 351.401(f)(1) — ie., a significant potential for the manipulation of price or production between the two companies — is not satisfied by the record evidence when considered according to the totality of the circumstances. JTEKT Mem. 13-14.

The Department’s regulations set forth three factors that the Secretary of Commerce “may consider” in identifying a significant potential for manipulation of price or production under § 351.401(f)(1). See 19 C.F.R. § 351.401(f)(2). The three factors are “[t]he level of common ownership,” id. § 351.401(f)(2)(i), “[t]he extent to which managerial employees or board members of one firm sit on the board of directors of an affiliated firm,” id. § 351.401(f)(2)(ii), and “[wjhether operations are intertwined, such as through the sharing of sales information, involvement in production and pricing decisions, the sharing of facilities or employees, or significant transactions between the affiliated producers,” id. § 351.401(f)(2)(iii).

Commerce concluded in the Decision Memorandum that “facts here meet each of the factors listed under 19 CFR 351.401(f)(2) as indicative of ... a potential” for the manipulation of price or production, stating that “[w]e found common ownership between the companies due to JTEKT’s holdings in the affiliate, shared board members, and intertwined operations of the two companies.” Decision Mem. 61. Commerce added that it “found that Koyo’s standing as the largest shareholder of the affiliate to be indicative of the intertwined nature of the companies.” Id. The Decision Memorandum explained that the finding regarding intertwined operations of the two companies “was based on a sharing of sales information, which Commerce found to be implicit in the sharing of board members, and significant sales transactions between the two companies.” Id. Commerce stated further that “[a]n analysis of Koyo’s sales data, including its acquisition costs of products from the affiliate, led us to conclude that the shared transactions were significant and the potential existed for Koyo to be involved in the production and pricing decisions of the affiliate.” Id. The decision cited for support an internal Commerce memorandum, the “Koyo Collapsing Memorandum.” 10 Id.; Mem. from Int’l Trade Compliance Analyst to Office Dir., AD/ CVD Operations 5 (Mar. 2, 2006) (Admin.R.Doc. No. 218) (“Collapsing Mem.”).

JTEKT does not dispute that it is the largest shareholder in the affiliate but points to record evidence indicating that its level of ownership is a minority share, not a controlling share. JTEKT Mem. 14. JTEKT argues that under Japanese law, a minority shareholder of a company cannot compel the company to disclose its sales and cost data. Id. at 14-15. With respect to shared board members, JTEKT asserts that these members “serve on a very limited basis as part-time (‘non-stationed’) auditors” on the board of directors of the affiliate, id. at 15, “are not permitted to vote, and have no control over the board’s decisions as to corporate business matters — such as prices or production.” Id. at 16. JTEKT argues that these directors, whose role is limited to conducting internal business audits, would have been acting ultra vires, and in breach of their fiduciary duty, were they to have shared with *1234 JTEKT pricing and cost information of the affiliate to which they had access. Id. JTEKT argues that the third reason Commerce gave for its decision to collapse the two entities, a significant level of sales transactions in which JTEKT purchased subject merchandise from the affiliate, does not provide a basis for that decision because these purchases were conducted at arm’s length and, on average, prices were higher than the affiliate’s cost of production. Id. at 17-18.

The court concludes that Commerce’s finding that “there is a significant potential for the manipulation of price or production” is supported by substantial record evidence. 19 C.F.R. § 351.401(f)(1). It is uncontested that JTEKT is the largest shareholder of the affiliate (although not holding a majority share) and that JTEKT and its affiliate share board members. JTEKT Mem. 14. JTEKT disputes the level of control it may exercise over its affiliate via its minority share and also contests the Department’s conclusion that the shared board members would have access to, or use, price and cost information for the affiliate upon attending board meetings. Id. at 14-16; Collapsing Mem. 4. However, under 19 C.F.R. § 351.401(f)(2), Commerce, in considering the level of common ownership and the extent to which there are shared board members, may exercise a significant degree of discretion. See Decision Mem. 61 (stating that “facts here meet each of the factors listed under 19 CFR 351.401(f)(2) as indicative of ... a potential ” for the manipulation of price or production (emphasis added)). The regulation does not require Commerce to find that a respondent exercises absolute control over its affiliate or that a board member may obtain and use price or cost information. The second factor in § 351.401(f)(2) is, therefore, satisfied. Commerce also satisfied the third factor, ie., “[wjhether operations are intertwined, such as through the sharing of sales information, involvement in production and pricing decisions, the sharing of facilities or employees, or significant transactions between the affiliated producers,” 19 C.F.R. § 351.401(f)(2)(ni), in finding that JTEKT made a certain number of sales to its affiliate and of those sales a certain percentage were sold at or below the cost of production. Collapsing Mem. 4. Even if the court assumes that the shared board members are not permitted to share pricing information, and did not do so, and takes at face value JTEKT’s statement that under Japanese law, a minority shareholder of a company cannot compel the company to disclose its sales and cost data, the court still must conclude that the record contains substantial evidence to support the findings underlying the Department’s determination, in particular the level of ownership and the substantial volume of transactions between the companies, some of which were found to be below-cost sales. On the basis of findings supported by substantial evidence, Commerce permissibly reached the ultimate finding of a significant potential for manipulation of prices according to the factors in the Department’s regulation, which regulation JTEKT does not challenge. The court therefore sustains the Department’s determination to treat JTEKT and its affiliate as a single entity in the AFBs 16 reviews.

D. Commerce Lawfully Deducted Certain “Additional Benefit Expenses” when Determining the Constructed Export Price of NSK’s Subject Merchandise

When determining constructed export price (“CEP”) according to 19 U.S.C. § 1677a(b), Commerce is required to make a deduction from the price at which subject merchandise is first sold in the United States to an unaffiliated purchaser, to account for expenses of that sale, if those *1235 expenses are incurred by, or for the account of, the producer or exporter or the affiliated seller in the United States. 19 U.S.C. § 1677a(d)(l) (2006). Commerce has provided in its regulations that in establishing constructed export price, “the Secretary will make adjustments for expenses associated with commercial activities in the United States that relate to the sale to an unaffiliated purchaser, no matter where or when paid.” 19 C.F.R. § 351.402(b) (2009).

NSK claims that Commerce, when calculating the constructed export price of NSK’s subject ball bearings that were sold in the United States, erred in deducting certain expenses from the price at which the merchandise was sold to an unaffiliated purchaser. NSK Mem. 22-24. The expenses in question were additional benefits expenses that were paid to Japanese nationals located in the United States, who also were paid base salaries for certain activities that Commerce determined to relate to the sale of subject merchandise to unaffiliated customers in the United States. Decision Mem. 74-75. NSK does not dispute that “[t]he base wage for these workers is an expense that is associated with commercial actions in the United States that relate to sales activity, and thus should be deducted by Commerce from the CEP.” NSK’s Reply Mem. of P. & A. in Supp. of NSK’s Mot. for J. on the Agency R. 11-12 (“NSK Reply”). NSK maintains, however, that the additional benefits expenses “are general and administrative (“G & A”) in nature and otherwise are not associated with commercial actions in the United States, let alone sales activity.” Id. at 12. According to NSK, the additional benefits expenses “are completely unrelated to the activities of these employees outside Japan.” NSK Mem. 23.

NSK, in presenting its argument, refers to certain information about the additional benefit expenses for which it claims confidential treatment. Commerce, in the Decision Memorandum, characterized that information generally as follows: “According to NSK, the expenses associated with certain additional benefits for these Japanese workers are designed to mimic the parent company’s compensation structure while the workers are residing outside Japan, regardless of whether the worker is in the United States or some other country.” Decision Mem. 74. Based on its review of the confidential record information, the court finds accurate the general characterization that Commerce included in the Decision Memorandum to provide a description of the confidential information for purposes of public disclosure.

The court concludes that substantial record evidence, including specifically NSK’s December 5, 2005 supplemental questionnaire response, is sufficient to establish that the employees in question were engaged in economic activity in the United States related to the sale of the subject merchandise. Letter from Crowell & Moring, LLP to Dep’t of Commerce 51-52 (Dec. 5, 2005) (Admin.R.Doc. No. 160) (“NSK Supplemental Resp.”) (setting forth pages S-34 to S-35 of NSK’s response). It is also undisputed that the base salaries and the additional benefits expenses at issue were paid to these same employees. NSK’s claim raises the question of whether or not the additional benefits expenses were paid as part of the total compensation that the Japanese nationals received for the activities they conducted in the United States associated with the sale of subject merchandise to unaffiliated producers, as opposed to activities conducted elsewhere or for some other purpose. Commerce expressly found in the Decision Memorandum that “[although the [additional benefit] expensef] may mimic the parent company’s compensation structure it still represents full compensation associated with economic activity occurring in *1236 the United States.” Decision Mem. 75. NSK’s argument might have some merit were it supported by record evidence establishing that the additional benefits expenses pertained on the whole, or at least in part, to specifically-described economic activity other than the commercial activity in the United States that related to the sale of NSK’s subject merchandise to unaffiliated purchasers. But NSK points to no such record evidence, nor is the court able to locate such evidence on the record. Such evidence as is present on the record is sufficient, under the standard of review, to support a finding that the additional benefits expenses related to the same activities for which the base salaries were paid to the Japanese nationals in the United States. Accordingly, the court concludes that NSK has not met its burden on this claim and affirms the Department’s deduction of the additional benefits expenses from the prices paid to the unaffiliated purchasers for purposes of the determination of constructed export price.

E. The Department’s Reallocation of NTN’s Freight Expense Using Weight Instead of Value Arbitrarily Treated NTN Less Fairly than Other Respondents

Commerce makes deductions for freight expenses when determining export price and normal value. See 19 U.S.C. §§ 1677a(c)(2)(A), 1677b(a)(6)(B)(ii). In the AFBs 16 reviews, as it had in reviews for prior periods, Commerce made the freight deductions based on allocated freight costs rather than actual costs for freight in specific transactions. See Decision Mem. 47-56. Under the Department’s regulations, “[a]ny party seeking to report an expense or a price adjustment on an allocated basis must demonstrate to the Secretary’s satisfaction that the allocation is calculated on as specific a basis as is feasible, and must explain why the allocation methodology used does not cause inaccuracies or distortions.” 19 C.F.R. § 351.401(g)(2). In the AFBs 16 reviews, NTN allocated its freight costs to its sales according to the value of the subject merchandise. Decision Mem. 48. Commerce concluded that NTN had failed to demonstrate that its value-based freight expense allocation did not cause distortions. Id. at 48^19. Noting that NTN did not incur the freight expenses on the basis of value but instead incurred them on various other bases (such as volume, distance, or hours of truck usage), Commerce expressed a concern that allocation based on value shifts freight expense from dumped to nondumped sales. Id. at 48-50. Upon rejecting NTN’s allocation method, Commerce reallocated NTN’s freight costs according to weight by using record data that NTN provided on the shipping weight for certain of its models, and, for other models for which actual weight data were not available on the record, by calculating estimates of weight based on size, using NTN’s product data. See id. at 52-56.

NTN challenges on various grounds the Department’s reallocation of its freight expense, which it maintains was reasonable and did not distort the margin calculations. See NTN Mem. 11-17. NTN argues that Commerce should not have departed from its long-standing practice of using value to allocate NTN’s freight expense, which practice, according to NTN, Commerce followed “in every final determination it has issued regarding NTN since 1989.” NTN Mem. 11. According to NTN’s argument, “Commerce ... is required to follow its past administrative practices in cases where the underlying facts presented to the agency have not changed,” id. at 11-12, and that “none of the facts relating to how NTN ships merchandise and incurs freight expense have changed in this review from previous reviews.” Id. at 12. The court finds this argument unpersuasive. First, NTN over *1237 states the burden confronting an agency seeking to change an established practice or methodology. Commerce ordinarily may change a methodology provided it states its rationale for doing so and provided the stated rationale is reasonable. NMB Singapore Ltd. v. United States, 557 F.3d 1316, 1328 (Fed.Cir.2009). The reasonableness of the Department’s previous practice of approving a value-based freight cost allocation for NTN does not itself preclude Commerce from discontinuing the practice. Second, contrary to NTN’s contention that Commerce consistently has followed its long-standing practice of using value to allocate NTN’s freight expense, and its claimed reliance on that practice, Commerce rejected NTN’s reporting of freight expense according to a value-based allocation in at least one prior review, the twelfth administrative reviews of the orders (“AFBs 12”). Issues & Decision Mem. for the Admin. Reviews of Ball Bearings (other than tapered roller bearings) & Parts Thereof from France, Germany, Italy, Japan, & the United Kingdom — May 1, 2000, through April SO, 2001, at 75-79 (Aug 30, 2002). The issue of allocation of NTN’s freight expense resulted in litigation in which the Court of International Trade and the Court of Appeals affirmed the Department’s determination that NTN had failed to demonstrate that the value-based allocation method it sought to use in AFBs 12 was not distortive or inaccurate. See NSK Ltd. v. United States, 28 CIT 1535, 346 F.Supp.2d 1312 (2004) (“NSK I”); NSK Ltd. v. United States, 481 F.3d 1355, 1359-61 (Fed.Cir.2007) (“NSK II ”). Having been placed on notice in AFBs 12 that Commerce determined NTN’s earlier allocation of freight expenses based on value to be unacceptable, a determination that the courts later upheld, NTN cannot plausibly contend that it reasonably relied on an expectation that it would be permitted to allocate freight expense based on value in AFBs 16.

Nevertheless, the court concludes that Commerce did not act lawfully in resolving the freight expense allocation issue that was presented by the sales of NTN and those of similarly situated respondents in AFBs 16. The administrative record includes an internal Commerce memorandum disclosing that other respondents in AFBs 16, like NTN, did not incur freight expense on a weight basis but submitted freight expense allocations that were based on the value of the merchandise. See Mem. from Senior Int’l Trade Compliance Analyst, AD/CVD Operations, Office 5, and Inti Trade Compliance Analyst, AD/ CVD Operations, Office 5, to Office Dir., AD/CVD Operations, Office 5, at 5 (Mar. 2, 2006) (Admin.R.Doc. No. 217). The memorandum also discloses that Commerce declined to reject any of these other value-based allocations. Citing this internal memorandum, Commerce stated as follows in the Preliminary Results:

With respect to other respondents in these administrative reviews that used a value-based methodology to allocate freight expenses, we recognize that no longer accepting value-based freight-expense allocation methodologies is a significant change in practice. Moreover, we do not have all of the data (e.g., the per-unit weight of the bearings) we would need to reallocate these respondents’ freight expenses. Therefore, we have not reallocated other respondents’ freight expenses in the current reviews. For future reviews of these orders, we will not accept value-based methodologies for the allocation of inland freight or international freight expenses except in situations where the freight charges are, in fact, incurred on a value, not weight or volume, basis (e.g., marine insurance).

*1238 Prelim. Results, 71 Fed.Reg. at 12,173. Commerce did not change its position on the freight allocation issue in the Final Results. As demonstrated by the internal memorandum and the Preliminary Results, Commerce reached a determination (which it described as “a significant change in practice”) in the AFBs 16 reviews under which all value-based allocation of freight costs that were not charged according to value are distortive per se. Id.

NTN challenges the Department’s “distinct treatment that was applied only to NTN.” NTN Mem. 13. NTN takes issue with the Department’s conclusion that it could reallocate NTN’s freight cost from a value basis to a weight basis because the record contained data from NTN that allowed Commerce do so. Id. at 12 (stating that Commerce “used data that is [sic] incomplete and, in many cases, based on estimates not supported by the record.”).

With respect to the allocation of freight, "the governing regulations require that a party seeking a particular expense adjustment or allocation must demonstrate its correctness to the Secretary of Commerce's `satisfaction.' Under this standard, the decision whether to accept a proposed allocation lies primarily within Commerce's discretion." Koyo II, 551 F.3d at 1292. The Department had discretion to decline to apply its new position until a subsequent review, so long as it exercised that discretion in a way that was evenhanded and not arbitrary. The issue presented by Commerce's decision in this case is "whether Commerce abused its discretion .. ., that is, whether Commerce's decision was `based on a consideration of the relevant factors and whether there has been a clear error of judgment.'" Torrington Co. v. United States, 68 F.3d 1347, 1351 (Fed.Cir.1995) (quoting Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971), and addressing the issue of whether Commerce abused its discretion in the context of determining whether to conduct a verification). However, in the AFBs 16 reviews, Commerce acted arbitrarily. It adopted a blanket policy of deferring the application of its change in position, which it considered "a significant change in practice," Prelim. Results, 71 Fed.Reg. at 12,-173, under which all value-based allocations, except those where freight cost is actually incurred according to value, are deemed distortive per se. Id. New practice aside, Commerce selectively applied 19 C.F.R. § 351.401(g) to disallow NTN's freight expense allocation but not those of the other respondents, without a rational basis for the distinction. In the Final Results, according to the Decision Memorandum, Commerce rejected NTN's value-based allocation because it concluded that NTN had failed to demonstrate that NTN's allocation was non-distortive (which, under Commerce's new practice, no respondent could do), even though the other respondents also failed to make this demonstration. See Decision Mem. 48, 51-52.

"A long line of precedent has established that an agency action is arbitrary when the agency offer[s] insufficient reasons for treating similar situations differently." Transactive Corp. v. United States, 91 F.3d 232, 237 (D.C.Cir.1996); SKF USA Inc. v. United States, 263 F.3d 1369, 1382 (Fed.Cir.2001) ("SKF I") (quoting Transactive Corp., 91 F.3d at 237). "Deference to agency authority or expertise . . . `is not a license to . . . treat like cases differently." Airmark Corp. v. FAA, 758 F.2d 685, 691 (D.C.Cir.1985) (quoting United States v. Diapulse Corp., 748 F.2d 56, 62 (2d Cir.1984)) (describing the Federal Aviation Administration's application of five criteria, in considering petitions for exemption from regulations imposing noise standards, as "grossly in *1239 consistent and patently arbitrary” and stating that “[elementary even-handedness requires that if all five factors must be met by one petitioner, then all five factors must be met by the next.” Air-mark, 758 F.2d at 692.).

Defendant attempts to defend the Department’s decision to accept the value-based freight allocations of the other respondents, but not that of NTN, by arguing that “Commerce reasonably determined that it would not reject value-based allocations when the record lacked all of the necessary data (the per-unit weights of bearings), necessary to reallocate certain respondents’ freight expenses.” Def. Resp. 48. This argument does not overcome the problem posed by the arbitrary and disparate treatment. Once Commerce had determined that all the value-based freight expense allocations at issue in the AFBs 16 reviews, including that of NTN, were distortive and therefore impermissible under 19 C.F.R. § 351.401(g), Commerce was faced with the task of effectuating that determination on a basis that was fair to all respondents. It failed in this task.

Referring to the Department’s new position of deeming distortive all value-based allocations of freight charges that were not actually paid according to value, defendant argues, further, that “[i]t was not necessary for Commerce to postpone applying this determination to NTN because Commerce verified that NTN possessed the necessary weight information to reallocate NTN’s freight expenses.” Def. Resp. 48-49; see also Decision Mem. 52 (“Furthermore, our decision to not re-calculate other respondents’ freight expenses in this review does not invalidate, in any way, our decision to implement such a change for NTN where we had readily available and verified data submitted by NTN that allowed us to calculate NTN’s freight expenses on the basis of weight reasonably.”). This argument also fails. It clearly was necessary for Commerce to postpone applying its new position to NTN because, under that new position, every respondent in the AFBs 16 reviews that used a value-based freight expense valuation failed the test of compliance with 19 C.F.R. § 351.401(g)(1). The Department’s reliance on its conclusion that NTN failed to demonstrate that its value-based allocation was not distortive for purposes of § 351.401(g)(1), see Decision Mem. 48, was pretextual in light of the agency’s announcing its new position, under which neither NTN nor anyone else could make such a demonstration. See Prelim. Results, 71 Fed.Reg. at 12,173. Commerce treated as inconsequential its own decision, apparent from the record, not to require the other respondents in the AFBs 16 reviews to make that demonstration. See id. Commerce attempted to justify its selective enforcement on a distinction as to whether it considered a party’s data adequate to enable Commerce to perform a reallocation, finding only NTN’s data suitable for this purpose, rather than on a distinction grounded in each party’s compliance with § 351.401(g). See id. This was an insufficient reason “for treating similar situations differently.” See Trans-active Corp., 91 F.3d at 237; SKF I, 263 F.3d at 1382. Commerce never determined that the allocations of the other respondents satisfied the requirement of 19 C.F.R. § 351.401(g)(1) and instead announced a new position under which these respondents, like NTN, necessarily would have failed to do so. The obligation to treat all respondents equally and fairly did not permit Commerce to overlook this failure and to reallocate NTN’s freight expenses.

As an incidental matter, the court observes that the Department’s justification for its arbitrary decision ignores the record fact that Commerce had weight data *1240 for only some of NTN’s models and resorted to its own methodology of estimating shipping weights for others of NTN’s models. Mem. from Financial Analyst, AD/ CVD Operations, Ojfice 5, to The File 7-9 (Mar. 2, 2006) (Admin.R.Doc. No. 222) (“NTN Prelim. Analysis Mem.”). Thus, Commerce did not have on the record a complete set of product weight data with which to reallocate the value-based freight expense calculations of any respondent in the reviews. See id.; Decision Mem. 53-54; Prelim. Results, 71 Fed.Reg. at 12,173 (“With respect to other respondents in these administrative reviews that used a value-based methodology to allocate freight expenses, ... we do not have all of the data (e.g., the per-unit weight of the bearings) we would need to reallocate these respondents’ freight expenses.”).

In summary, Commerce’s regulatory decision to postpone implementing its new position on value-based freight cost allocations was impermissibly arbitrary because it was applied selectively to all respondents other than NTN. The decision to reallocate only NTN’s freight expense was also impermissibly arbitrary because, under the Department’s new position, none of the other respondents in the AFBs 16 reviews who used a value-based freight allocation could have satisfied the test of compliance with 19 C.F.R. § 351.401(g)(1), and none were required even to attempt to make such a demonstration. Therefore, the Department’s decision to reallocate NTN’s freight cost must be set aside as contrary to law. See 19 U.S.C. § 1516a(b)(l)(B)(i).

F. Commerce Permissibly Recalculated NTN’s Home Market Packing Materials Expenses

NTN claims that Commerce unlawfully decided to recalculate NTN’s home market packing materials expenses and failed to support with substantial record evidence its redetermination of that expense. See NTN Mem. 30-35. Defendant responds that Commerce correctly determined that NTN’s home market packing expenses, as reported by NTN, was affected by distortions and supported that determination with substantial record evidence. Def. Resp. 38-43.

Commerce determined that NTN’s method of reporting data on expenses incurred for packing materials on home market sales was unsatisfactory in failing to account for differences in the packing materials expenses that were incurred in sales of foreign like products to different “customer categories,” i.e., original equipment manufacturers (“OEMs”), distributors, and after-market purchasers. Decision Mem. 81-82. Commerce concluded that NTN had reported its home market packing materials expenses according to “a single packing expense factor” rather than according to separate packing expense factors for these different customer categories. Id. at 82. Commerce considered the use of a single factor to cause distortions by failing to capture differences in expenses for packing materials inherent in packing requirements associated with the different customer categories. Id. Commerce noted that NTN indicated in its questionnaire responses that sales to OEMs are packed in bulk while sales to distributors and after-market purchasers are packed individually in boxes. Id. at 80. In response to its finding that NTN’s single packing expense factor caused distortions, Commerce recalculated NTN’s home-market packing materials expenses by using “as a proxy” customer-category-specific data that NTN submitted on the cost of repacking materials for its U.S. sales. Id. at 80. NTN claims that this recalculation was unlawful for various reasons. Id.

NTN first argues that in these reviews, NTN used the same method to report its *1241 home market packing expenses that it has used for many years, which method was based on previous decisions by Commerce that NTN should not allocate expenses for packing materials to account for differences among customer categories. NTN Mem. 31-82. NTN asserts that it has relied on the Department’s “established methodology,” id. at 31, and that Commerce should not be allowed to change that established methodology “without a reasoned explanation of its departure from previous, long-standing methods.” Id. at 32. NTN also implies that Commerce must provide “a compelling reason to use a ratio for packing material that varies by customer category....” Id. Defendant responds that NTN was on notice regarding the Department’s concerns and states that “Commerce raised concerns about NTN’s packing expense calculation methodology as early as the fourteenth administrative review when NTN calculated its home market packing expenses using a sales value allocation methodology, without keeping records of packing costs on a unit-by-unit basis.” 11 Def. Resp. 41 (citing Final Results POR 2003-04, 70 Fed.Reg. at 54,714); see AFBs 15 Decision Mem. 64-65 (finding that “it is appropriate to reallocate NTN’s packing expenses ... because packing in bulk costs less on a per-bearing basis than individual packing” and setting forth a methodology under which Commerce assigned sales of bulk-packaged merchandise a factor of one and sales of individually-packed merchandise a factor based on the number of bearings packed individually).

The court reviews a change in methodology for reasonableness. SKF II, 537 F.3d at 1377-78. As a general matter, Commerce is free to change its method of calculating home market packing expenses if it states its rationale for doing so and if that stated rationale is reasonable. NMB Singapore, 557 F.3d at 1328. In the Decision Memorandum, Commerce acknowledged that it “directed NTN not to report customer-specific packing expenses in an early administrative review of this order,” but explained that in the prior administrative review, as well as the instant review, “it ha[d] become apparent that such reporting is necessary because [Commerce] designate^] NTN’s levels of trade based on its reported customer categories.” Decision Mem. 82. Commerce further explained:

When we calculate the level-of-trade adjustment factors for NTN’s export-price sales, we compare weighted-average home-market prices for each level of trade, net of adjustments and expenses, including packing expenses. See 19 CFR 351.412(e). NTN derived and reported sale-specific packing expenses by multiplying the reported single packing-expense factor by gross unit prices. This results in distorted net prices among levels of trade because the average single packing-expense factor minimizes the disparity between packing costs for different customer categories. This in turn results in distorted level-of-trade adjustment factors which, in turn, results in distorted normal values after application of level-of-trade-adjustments for distributor, after-market, and OEM sales when we make comparisons at different levels of trade for NTN’s export-price sales.

*1242 In sum, the use of separate packing-expense factors by customer category is warranted because it captures differences in expenses for packing materials inherent in packing requirements with respect to different customer categories which, in turn, eliminates foreseeable distortions in our margin calculations.

Id.

Commerce is required to determine home market packing expenses and make level-of-trade adjustments when determining normal value and comparing it to export price. See 19 U.S.C. § 1677b(a)(l)(6), (7)(A). The court concludes that the Department’s objective of seeking to allocate packing expenses by a more accurate methodology, for purposes of the level-of-trade adjustments Commerce makes when comparing U.S. and home market sales, was reasonable. A single packing-expense ratio that applies to different customer categories must be considered less accurate than separate ratios where packing expense variances exist with respect to customer category. Additionally, although NTN claims that it relied on the Department’s previous methodology, NTN has not shown how any reliance it may have had was reasonable, given that Commerce made a similar adjustment in AFBs 15. Moreover, NTN has not shown specifically why any reliance it may have had on the previous methodology was detrimental reliance.

Next, NTN objects that “Commerce did not ask NTN to explain or comment on why it did not report its packing material ratios by customer category.” NTN Mem. 33. The court is unconvinced by this argument. At issue was not the question of why NTN did not report its packing material ratios by customer category — that much was clear from the record. Rather, the court must decide, first, whether the Department’s findings were supported by substantial record evidence, including the ultimate finding that NTN’s allocation of packing expenses according to sales value caused distortions. If so, the second question is whether it was reasonable for Commerce, in response to this finding, to recalculate the expenses based on packing materials factors derived from data on NTN’s sales in the U.S. market, which factors were customer-category-specific.

The administrative record, considered as a whole, supports the findings that NTN sold subject ball bearings in the United States and foreign like products in its home market in different customer categories, did not maintain data on the home market packing materials expenses according to those various categories and, as a result, allocated its home market packing materials expenses based on the sales value of the merchandise. Specifically, the administrative record demonstrates that NTN sold subject merchandise in the United States through three channels of distribution — through its affiliated U.S. sales company, to unaffiliated OEM customers directly, and to a U.S. manufacturing affiliate that incorporates subject merchandise as a subject component — and that the packing expenses vary according to the channel of distribution and customer type. Letter from Barnes, Richardson & Col-burn to Dep’t of Commerce 26-27, 31 (Sept. 26, 2005) (Admin.R.Doc. No. 89) (“NTN Questionnaire Resp.”) (at A-ll, A-12, and A-16 of the response). NTN itself set forth the example of repacking merchandise in individual boxes when sold to distributor and aftermarket customers compared with the bulk packing for sales to OEMs. Id. NTN’s questionnaire responses also show that NTN’s home market sales involved different customer categories. Id. at 29 (at A-14 of the response).

*1243 The record indicates that1 in a supplemental questionnaire, Commerce made the following request:

based on the information you provided in your response to section A of the Department’s questionnaire, it is apparent that NTN packs merchandise for sale to OEMs in bulk and individually for sales to distributors and after-market purchasers. Please demonstrate that an allocation of packing expenses based on sales value does not cause distortion especially in light of the fact that margin calculation may result in the comparisons of same models at different levels of trade designated by customer category.

Letter from Dep’t of Commerce to Barnes, Richardson & Colburn 19 (Oct. 20, 2005) (Admin.R.Doc. No. 105) (“NTN Supplemental Questionnaire ”) (at page 17, ¶ 82 of the questionnaire). NTN responded that

The ITA has recognized on many previous occasions that NTN does not track individual packing expenses. Therefore, NTN has adopted the only available methodology: to report packing expense based on relative sales value. Additionally, individual distributor and aftermarket prices are generally higher than OEM prices and, under NTN’s methodology, a greater proportion of the packing expenses will be allocated to these purchasers. We believe, therefore, that the methodology is not distortive; rather, it accurately reflects packing expenses in general.

Letter from Barnes, Richardson & Col-burn to Dep’t of Commerce 67-68 (Nov. 17, 2005) (Admin.R.Doc. No. 139) (“NTN Supplemental Resp.”) (at B-14 to B-15 of the response). Commerce followed up on the issue with NTN in the notice prior to verification:

Demonstrate how you allocated packing expenses. Provide all source documentation. Show how the packed form reported for the preselected sales relates to customers’ packing requirements and to the per-unit expenses. Provide documentation supporting your calculation of packing labor for domestic sales. Contrast HM [ie., home market] packing costs with export packing costs.

Letter from Dep’t of Commerce to NTN Corp. 12 (Nov. 23, 2005) (Admin.R.Doc. No. 145) (“NTN Verification Notice ”) (at page 10 of the notice). In its report on the verification of NTN’s home-market and export-price sales, Commerce “found that NTN’s calculation of packing expenses results in essentially the same packing ratios for distributors and OEMs.” Letter from Senior Int’l Trade Specialist, AD/CVD Operations, Office 5, to File 11 (Jan. 4, 2006) (Admin.R.Doc. No. 184) (“NTN Verification Report of Home-Market & ExporttPrice Sales ”). This finding, adopted by the Final Results, clearly was supported by substantial record evidence, which NTN itself provided in its questionnaire responses. See NTN Supplemental Resp. 67-68 (at B-14 to B-15 of the response). The finding was the basis for the conclusion by Commerce that NTN’s reporting of its home market packing expenses causes distortions. That finding, also, is supported by the record evidence showing that NTN’s packing methods were not uniform across the different customer categories. Although NTN points out that its method allocated more packing materials expenses to the distributor and aftermarket sales, which generally were of higher value than the OEM sales, see NTN Mem. 32-33 (citing NTN Supplemental Resp. 67-68 (at B-14 to B-15 of the response)), this argument fails to account for the distortion that necessarily would result upon the Department’s making a level-of-trade adjustment.

*1244 The remaining question is whether the recalculation Commerce performed was reasonable. NTN argues that Commerce must use the actual home market data because it is accurate and the best evidence available, particularly when compared to the information related to U.S. packing expenses upon which Commerce relied. NTN Mem. 33-35. NTN contends that the Department's adjustment is unrelated to the costs incurred, ignores evidence on the record, and distorts NTN's packing expenses. Id. at 35.

In the Decision Memorandum, Commerce explained that "the record does not have the necessary data to calculate home-market packing-expense factors for each customer category" and that Commerce "made that statement because, although NTN provided home-market packing-expense factors by customer category, [Commerce] found that [NTN] had derived such factors by allocating the total cost center-specific packing-materials and packing-labor expenses based on the ratio of customer category-specific sales value to the total sales value of each cost center." Decision Mem. 82-83. Commerce found that "each of [NTN's] cost centers is dedicated to capturing sales information applicable to the various customer categories" but that "these cost centers do not capture the expense information in the same manner." Id. at 83. Commerce explained that "the packing-expense factor for distributors for a particular cost center, as reported by NTN, does not capture actual packing expenses incurred by this cost center even though it does capture information concerning the value of sales to distributors." Id. Commerce restated its finding that "the home-market customer category-specific factors NTN reported are virtually identical in values on a cost center-specific level or on a company-wide level. . . ." Id.

Despite the shortcomings that Cornmerce found to exist in NTN's allocation method, NTN frames the issue as a choice between its home market packing cost data and its packing cost data pertaining to its U.S. sales. This formulation overlooks the problem that resulted for level-of-trade adjustments from NTN's having reported its home market packing materials expenses according to a single packing expense factor rather than actual packing cost data specific to the different customer categories. It was reasonable for Commerce to attempt to address this problem through a recalculation. Although NTN objects to the Department's use of U.S. packing data to accomplish the recalculation, the mere fact that the data used to perform the adjustment pertained to U.S. packing costs, not home market packing costs, does not support a conclusion that the Department's recalculation was impermissible. It is not apparent how use of any other data on the record would have solved the distortion problem that Commerce identified.

For these various reasons, the court affirms the Department's recalculation of NTN's home market packing materials expenses in the Final Results.

G. Commerce Lawfully Disallowed Certain Downward Price Adjustments that Resulted from NTN'S Allocation of Discounts to Home Market Customers

NTN, when reporting its hom€ market sales to Commerce, made dowm ward adjustments to the original prices is sales of foreign like products to certain customers to account for discounts that it granted retroactively to those customers. At verification, Commerce found a discrepancy between the adjusted prices as reported to Commerce and the actual method by which NTN granted the discounts, concluding that the downward adjustments were made to all sales of a particular customer during the fiscal year even *1245 though the discounts actually applied only to certain models for certain time periods. 12 Commerce disallowed NTN’s downward adjustments, concluding that “NTN’s allocation of the discounts in question results in the assignment of price adjustments to sales which cannotbe [sic ] said to have been reasonably affected by such adjustments,” Decision Mem. 67, and “results in a price adjustment that is not reasonably attributable to sales of the foreign like product, as required by 19 CFR 351.401(c).” Id. at 67-68. NTN contends, on several grounds, that “Commerce erroneously disallowed NTN’s home market discounts because NTN provided sufficient evidence to support its allocation.” 13 NTN Mem. 35; see NTN Reply 13-15. Defendant responds that “Commerce correctly rejected NTN’s price adjustment methodology because the methodology inaccurately allocates price adjustments from sales that were actually adjusted to sales that did not receive adjustments.” Def. Resp. 29. Defendant further argues that “[additionally, NTN has not adequately explained why its methodology is not distortive.” Id. (relying on 19 C.F.R. § 351.401(g)).

As a preliminary matter, NTN raises a procedural issue, arguing that because Commerce accepted NTN’s allocation for the tenth through fifteenth administrative reviews and verified the data used in the methodology in the tenth and thirteenth reviews, Commerce should have explained its reasons for changing its policy. NTN Mem. 35, 39. As the Court of Appeals has held, “Commerce’s acceptance of an allocation methodology in a previous review does not relieve a party of its burden of demonstrating the methodology is non-distortive in the current review.” NSK III, 510 F.3d at 1381. Nonetheless, if Commerce changes a methodology to which it adhered in previous proceedings, Commerce must “provide[] a sufficient, reasoned analysis explaining why a change is necessary.” NMB Singapore, 557 F.3d at 1328 (“Commerce, like any agency, is due deference from the courts in certain matters entrusted to it by Congress. Once Commerce establishes a course of action, however, Commerce is obliged to follow it until Commerce provides a sufficient, reasoned analysis explaining why a change is necessary.” (citation omitted)); NSK III, 510 F.3d at 1381.

In these reviews, the Department’s finding at verification that NTN’s allocation method allocated price adjustments to sales that were not actually subject to these adjustments was the basis for the decision NTN challenges. The Court of Appeals addressed an analogous situation in NSK III, concluding that although *1246 Commerce had accepted a respondent’s method of allocating lump-sum billing adjustments in previous reviews, the Department’s observation of clear evidence of a substantial distortion was a sufficient basis for Commerce to change its position and reject the respondent’s allocation methodology. NSK III, 510 F.3d at 1381-82. Here also, the Department’s finding that an allocation method caused distortions and the explanation of the Department’s reasoning in the Decision Memorandum provide adequate support for the Department’s determination to change its position on the acceptability of NSK’s allocation methodology.

NTN disputes the Department’s statement in the Decision Memorandum that price adjustments must be product-specific and time-specific. NTN Mem. 36-37; see Decision Mem. 67 (in which Commerce stated that in determining an antidumping duty margin “we do not use customer-specific aggregate prices” and that “[w]e use product-specific and time-specific prices in our margin calculation”). NTN contends that “[a]ny product-specific or time-specific reporting must, necessarily, be based on a customer-specific buildup” and states, as an example, that “NTN does not simply grant a discount for a certain product for any and all sales of such product made in Japan during the period of review. On the contrary, the discount is granted based on sales of that product to a particular customer.” NTN Mem. 37. NTN further states that “the discount policy is a post hoc adjustment that is related to the total sales of a certain product that were made to a certain customer during the fiscal year,” id. at 38, and concludes that “[o]nly when sales to a particular customer are aggregated can a product-specific, or time-specific, attribution be accomplished.” Id. at 37.

To be recognized by Commerce, a price adjustment must be reasonably attributed to the subject merchandise or the foreign like product. 19 C.F.R. § 351.401(c). Under 19 C.F.R. § 351.401(g)(1), Commerce “may consider allocated ... price adjustments when transaction-specific reporting is not feasible.” 19 C.F.R. § 351.401(g)(1). A respondent, however, must satisfy Commerce that “the allocation method used does not cause inaccuracies or distortions.” Id. Specifically, the requesting party “must demonstrate to the Secretary’s satisfaction that the allocation is calculated on as specific a basis as is feasible, and must explain why the allocation methodology used does not cause inaccuracies or distortions.” Id. at § 351.401(g)(2).

In the Decision Memorandum, Commerce “determined that NTN reported the discounts differently than how it granted such adjustments.” Decision Mem. 67. Commerce explained that it “determined that, in its attempt to derive a factor for reporting the discounts in question, NTN divided the total value of discounts granted (i.e., the numerator) by the total value of certain sales (i.e., the denominator)” and found, at verification, “that the total value of sales NTN used in the denominator contained sales ... [that] did not bear any relationship to the discounts NTN had captured in the numerator.” Id. Commerce concluded that “the numerator and denominator were not stated on a rational basis. In other words, NTN’s methodology is distortive because NTN allocates adjustments from sales that actually had an adjustment to sales that did not have an adjustment.” Id. Substantial record evidence supports the Department’s finding that the allocation methodology was distortive. NTN does not refute the factual finding that was central to the Department’s rejection of NTN’s allocation method — that NTN applied the aggregate discounts to sales of all models from a specific customer for the fiscal year when the dis *1247 counts were afforded only on certain models or for specific time periods.

NTN alleges that its allocation method was based on the most specific way that NTN could collect data due to the determination of, and distribution of, such discounts. NTN Mem. 38-39. However, this allegation does not find adequate support in the record of the administrative reviews. In the Final Results, Commerce rejected NTN’s assertion that it was not feasible for NTN to report its discounts any more specifically, Decision Mem. 68, and NTN did not supply a convincing reason, either to Commerce or in its later submissions to the court, why records kept in the ordinary course of business and establishing the actual discounts could not have been used, manually if necessary, to make more accurate adjustments than those resulting from NTN’s broad allocation method. As the regulations provide, the burden was on NTN to demonstrate that the allocation method does not cause inaccuracies or distortions. See 19 C.F.R. § 351.401(g). In determining whether the burden has been met, Commerce looks at “the records maintained by the party in question in the ordinary course of its business, as well as such factors as the normal accounting practices in the country and industry in question and the number of sales made by the party during the period of ... review.” 19 C.F.R. § 351.401(g)(3). Although Commerce “reeognize[d] the potential existence of limitations associated with NTN’s electronic record-keeping” it also observed that “NTN did not make an argument as to whether it could have accomplished the proper reporting of discounts in question from its records manually.” Decision Mem. 68. Commerce explained:

NTN granted the discounts in question to specific customers for purchases of specific products during specific periods of time. Surely NTN maintains the records in the ordinary course of its business, such as invoices, that are necessary to supplement its reporting. In addition, the number of sales for which the discounts were granted is insignificant in relation to NTN’s univérse of reported home-market sales. Furthermore, NTN is silent as to why it applied its allocated-discounts factor to reported home-market sales which were not eligible for such discounts (i.e., sales to the same customer of products for which NTN did not grant the discounts and sales to the same customer of products for which NTN did grant the discounts but outside the period for which it granted the discounts).

Id. The Department’s determination that NTN did not meet the burden imposed by 19 C.F.R. § 351.401(g) was supported by substantial evidence on the record.

In summary, the record evidence supports Commerce’s overall finding that NTN’s allocation method allocated discounts to sales ineligible to receive such discounts. The discrepancies that Commerce found at verification to exist between the home market prices reported by NTN and the prices that would have resulted from prices adjusted to reflect the actual discounts necessarily would have distorted the determination of normal value for specific like products and, accordingly, the dumping margins on sales of subject merchandise. The Department’s regulations specifically allow the Secretary to reject price adjustments that cannot be shown to be reasonably attributable to the foreign like product, 19 C.F.R. § 351.401(c), because they have been allocated by a method that has not been shown not to cause inaccuracies or distortions, id. § 351.401(g). The court, therefore, concludes that Commerce acted within its authority under the applicable regulations to disallow NTN’s downward price adjustments in the AFBs 16 reviews.

*1248 H. The Department’s Use of Facts Otherwise Available and Adverse Inferences in Response to Discovering Na-chi’s Errors in Reporting Physical Bearing Characteristics Was Contrary to Law

To implement its model-match methodology, Commerce requested that respondents report individual physical characteristics for each bearing model in their U.S. and home market sales databases. Ñachi Mem. 5. In response, Ñachi reported physical characteristics for 2,084 individual models of bearings. Id. at 6. During verification, Commerce reviewed a sample of the data that Ñachi reported on the physical characteristics of its bearings, examining information corresponding to forty of the 2,084 models reported. Id.; Mem. from Senior Case Analyst, AD/CVD Enforcement & Case Analyst, AD/CVD Enforcement, to The File 8-5 (Feb. 9, 2006) (Admin.R.Doc. No. 208) {“Ñachi Verification Mem.”). Upon verification, Commerce found nineteen errors in Nachi’s reporting of physical characteristics, which nineteen errors affected sixteen of the forty models reviewed. Ñachi Verification Mem. 4. Based on the verification, Commerce reached a finding that “Ñachi reported incorrect physical characteristics for 16 of the 40 models we examined at verification.” Final Results, 71 Fed.Reg. at 40,066. Citing “the significant number of errors and the pervasive nature of those errors,” Commerce characterized Nachi’s reporting errors as “systemic in nature.” Decision Mem. 34. Commerce concluded that “[b]eeause the errors in the reported physical characteristics were systemic in nature and can distort the model match ... we cannot be sure that the similar matches we find using Nachi’s reported physical characteristics are actually matches of the most similar models.” Id. at 36. “Moreover, we cannot be sure, where we find no similar matches using Nachi’s reported physical characteristics, that no similar matches actually exist because Ñachi reported the characteristics of those potential matches incorrectly.” Id.

Based on its finding of systemic reporting errors, Commerce, pursuant to 19 U.S.C. § 1677e(a), reached a general conclusion that it must use facts otherwise available in calculating the antidumping duty margin for all of Nachi’s U.S. sales that lacked an identical match. See Decision Mem. 36 (stating that “we determine that we must use the facts available for any of Nachi’s U.S. sales for which we did not find a contemporaneous identical match.”). In response to Nachi’s comments on the Department’s intention to proceed in this way, Commerce determined that it was appropriate to make three exceptions. Id. at 38-39. First, Commerce did not use facts otherwise available where the Department’s model-matching methodology resulted in a match to a model in the same bearing series as the U.S. model, reasoning that models in the same bearing series are essentially the same bearing with minor modifications and that, as a result, the model selected is in fact the most similar model. Id. at 38. Second, Commerce did not invoke facts otherwise available for U.S. sales of ball bearing parts that were further manufactured in the United States because, for parts sales, the Department’s methodology does not attempt to find a similar match if it finds no identical match. Id. Finally, Commerce did not use facts otherwise available for Nachi’s U.S. sales of models that did not have the same bearing design, load direction, number of rows, and precision grade as any models sold in the home market. Id. at 39. Commerce reasoned that because Ñachi did not make substantial errors in reporting these particular four characteristics, the errors affecting these characteristics were not systemic in nature. Id. For Nachi’s U.S. sales of models that did not have the same bearing *1249 design, load direction, number of rows, and precision grade as any models sold in the home market, Commerce based normal value on constructed value. Id.

Overall, Commerce resorted to facts otherwise available for 203 comparisons between constructed export price and normal value, out of a total of 7,385 such comparisons. Ñachi Mem. 22. The result was that Commerce resorted to facts otherwise available for 2.7% of Naehi’s U.S. sales. Id.

In the Final Results, Commerce further found that “Ñachi did not act to the best of its ability in reporting its physical characteristics because Ñachi had the correct data available to it.” Final Results, 71 Fed.Reg. at 40,066. Based on the latter finding, Commerce determined under 19 U.S.C. § 1677e(b) that “it is appropriate to use adverse inferences in addressing the errors in the characteristics Ñachi reported.” Id. Commerce selected as “adverse facts available” the highest margin determined for Ñachi in any previous proceeding, i.e., the 48.69% rate Commerce determined for Ñachi in Final Determinations of Sales at Less Than Fair Value; Anti-friction Bearings (Other Than Tapered Roller Bearings) & Parts Thereof From Japan, 54 Fed.Reg. 19,101 (May 3, 1989) (“Japan Final Determination 1987-1998”). See Final Results, 71 Fed.Reg. at 40,066. As an adverse inference under § 1677e(b), Commerce applied that rate to all of Nachi’s U.S. sales for which Commerce found no identical match, and to which none of the three exceptions discussed above applied. Id.; see also Decision Mem. 38-39. In the Final Results, Commerce assigned Ñachi a weighted average dumping margin of 16.02%. Final Results, 71 Fed.Reg. at 40,066.

Ñachi argues that “[t]he Department’s choice to use adverse facts available as a result of several errors in Nachi’s reporting was improper.” Reply Br. of Pis. Na-chi-Fujikoshi Corp., Ñachi Am., Inc. & Ñachi Tech., Inc. in Supp. of Rule 56.2 Mot. for J. on the Agency R. 7 (“Ñachi Reply”). Ñachi acknowledges that Commerce was justified in invoking the “facts otherwise available” provision of 19 U.S.C. § 1677e(a) with respect to the matching of “similar” bearing models when the dumping calculation was affected by Nachi’s errors. Ñachi Mem. 11. Ñachi submits, however, that Commerce acted contrary to law, and without support in record evidence, when it used “adverse inferences,” pursuant to 19 U.S.C. § 1677e(b), in selecting from among the “facts otherwise available.” Id. Ñachi further argues that the 48.69% rate Commerce selected by Commerce as “adverse facts available” is contrary to law because it is unduly punitive and bears no relationship to Nachi’s pricing and cost structure in this current review. Id. at 12.

1. The Department’s Use of Facts Otherwise Available Was Impermissibly Broad

Under subsection (a)(1) of § 1677e, Commerce uses facts otherwise available when “necessary information is not available on the record.” 19 U.S.C. § 1677e(a)(l). In addition, the statute directs Commerce generally to use facts otherwise available in the circumstances identified in any of the four subparagraphs of subsection (a)(2) of the section, which circumstances include, as provided in subparagraph (B), where the requested information is not submitted within the applicable time period, and the circumstances in sub-paragraph (D), where the requested information is provided but cannot be verified. 14 *1250 19 U.S.C. § 1677e(a)(2). If one of the four conditions is met, Commerce shall, subject to § 1677m(d) (which addresses deficient submissions), “use the facts otherwise available in reaching the applicable determination.” 19 U.S.C. § 1677e(a). In the Decision Memorandum, Commerce invoked subparagraphs (B) and (D) in support of its decision to use “facts otherwise available.” Decision Mem. 33; Final Results, 71 Fed.Reg. at 40,066.

Subparagraph (B) of § 1677e(a)(2) does not apply to the incorrect data that Ñachi reported, for which Commerce did not make a finding of untimeliness, although subparagraph (B) would have applied to corrected information, had Ñachi later attempted to submit any, because the deadline for submission of questionnaire responses had already passed. Subparagraph (D) calls for the use of facts otherwise available as a substitute for information that “cannot be verified as provided in section 1677m(i) of this title.” 15 Commerce found that certain of the reported information on physical bearing characteristics was incorrect because it was inconsistent with information in Nachi’s technical drawings and catalogs, Decision Mem. 34, a finding that Ñachi does not contest in this litigation.

Commerce did not explain in the Final Results why it considered it reasonable to construe subparagraph (D) of 19 U.S.C. § 1677e(a)(2) to allow it to proceed as it did in the subject review. Nevertheless, the court discerns in the Final Results an implied construction of the statute, which it reviews according to the deference required by Chevron, 467 U.S. at 842-44, 104 S.Ct. 2778. See Pesquera Mares, 266 F.3d at 1379-82. Under Chevron, the court first considers “whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.” Chevron, 467 U.S. at 842-43, 104 S.Ct. 2778. If, however, “the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.” Id. at 843,104 S.Ct. 2778.

Read according to plain meaning, sub-paragraph (D) allows Commerce to substitute facts otherwise available for specific record information provided that the information “cannot be verified.” 19 U.S.C. § 1677e(a)(2)(D). According to this plain meaning, the provision obviously would be satisfied if a party refuses to allow verification of information it has reported to Commerce or fails to maintain or produce the records that are needed to verify that information. Because the word “verify” is generally understood to mean “to prove to be true” or “to confirm or establish the authenticity or existence of,” the court concludes that the provision, when read according to its plain meaning, also encompasses situations in which specific reported information, when examined according to the supporting records, is shown to be incorrect. Webster’s Third New Interna *1251 tional Dictionary 2543 (1993) (defs. 2 & 5); 19 Oxford English Dictionary 540 (2d ed.1989) (defining “verify,” in the fourth definition, as “[t]o ascertain or test the accuracy of (something), esp. by examination or by comparison with known data, an original, or some standard.... ”). In invoking § 1677e(a)(2)(D), however, Commerce went much further. First, it treated all the reported information on physical bearing characteristics that Ñachi reported, and that Commerce actually examined (except to the extent one of the three aforementioned exceptions applied), as information that could not be verified, including the information that Commerce actually found to be correct. See Decision Mem. 33-35. Second, Commerce treated as unverifiable all of Nachi’s reported information on physical bearing characteristics that it did not examine during verification (again, other than in instances in which one of the three exceptions applied). Therefore, for purposes of Chevron, 467 U.S. at 843, 104 S.Ct. 2778, the “precise question at issue” is whether it was permissible for Commerce impliedly to construe 19 U.S.C. § 1677e(a)(2)(D) so broadly as to encompass specific information that Commerce determined upon examination to be correct. A second question is whether Commerce could treat as unverifiable a second body of reported information that Commerce did not examine but that, based on findings stemming from its examination of the first body of information, assumed to contain errors.

As to the first question, the court concludes that the Department’s statutory construction is contrary to the clearly expressed intent of Congress. In § 1677e(a)(2), Congress referred to specific information that “has been requested by the administering authority” and that “cannot be verified.” 19 U.S.C. § 1677e(a)(2)(A), (D) (emphasis added). Specific reported information that Commerce did examine, and find to be correct, is the opposite of information that “cannot” be verified; it is information that was verified. Although Commerce found some physical characteristics information that Ñachi reported at the same time to be unverifiable, that finding did not apply to the same “information.” This inherent contradiction is sufficient to demonstrate that the Department’s implied construction of subparagraph (D) and, specifically, of the term “cannot be verified,” contravenes the congressional intent. Even though it is not necessary to consider the question further, the court observes that other provisions in the statute confirm Congress’s intent. In 19 U.S.C. § 1677m(e), Congress addressed the situation in which reported information “can be verified” but nevertheless “does not meet all the applicable requirements established by the administering authority,” directing that Commerce use the information if certain requirements are met. See 19 U.S.C. § 1677m(e)(2). The specific information on physical bearing characteristics that Ñachi reported and that Commerce confirmed as correct must be considered to meet all applicable requirements, as Commerce did not find to the contrary. Because Commerce, according to § 1677m(e)(2), would have been required to use this information even if the information were found not to meet all applicable requirements, it would be nonsensical to construe the statute to allow Commerce to reject it, in favor of facts otherwise available, even though it was information that could be, and was, verified, and that otherwise met all applicable requirements. The court concludes, therefore, that Congress explicitly conditioned the application of § 1677e(a)(2)(D) on a finding of fact that the particular information at issue be found to be unverifiable. In summary, Commerce acted unlawfully in impliedly construing § 1677e(a)(2)(D) to allow it to reject as unverifiable specific information that it actually verified.

*1252 With respect to the second question, the issue presented is whether Commerce may treat as information that “cannot be verified” information that a party made available for examination but that Commerce, for reasons of its own, made no attempt to examine. The court does not find in the statute an expression of congressional intent on this precise question. Congress gave no indication of foreclosing an application of § 1677e(a)(2)(B) simply because Commerce chose not to conduct an actual verification procedure with respect to the specific information involved. Nor does the court, under the second step of the analysis required in Chevron, find unreasonable an implied construction under which the mere fact that no examination was conducted does not necessarily foreclose an application of subparagraph (D). Instead, the general terms in which the provision is stated indicates that Congress did not attempt to anticipate the myriad circumstances in which information may be found to be unverifiable and instead left it to Commerce to decide what those factual circumstances might be. There may be circumstances in which Commerce conceivably could reach a legitimate finding that information cannot be verified, even though no actual verification procedure ensued. The answer in any specific circumstance may depend partly on the reasons for the Department’s decision not to conduct the examination. The question is properly analyzed as a question of substantial evidence for the factual finding of unverifiability. Here, Commerce made an express or implied finding, or inference, that all of Nachi’s reported information on physical bearing characteristics that Commerce chose not to examine, except where one of the three exceptions applied, could not be verified. The issue presented in this case is whether substantial evidence on the record supported that finding or inference. The court concludes that it does not.

There is no evidence of record that could support an actual finding of fact that errors exist in the unexamined information on physical characteristics that Ñachi reported to Commerce. Although the unexamined reported information is on the administrative record, the business records of Ñachi that would be needed to verify that information are not on the record. The record discloses that the examined information contained errors at a significant level. The fact that these errors occurred is undisputed. Commerce could draw an inference from this evidence that the unexamined information is likely to contain errors as well. Commerce drew just such an inference. See Decision Mem. 35 (“Given that we found additional errors for individual models within each additional 10 models we selected, however, we do not have any reason to believe that we would not have continued to find errors had we enlarged the sample size.”). But the record evidence is not sufficient under the substantial evidence standard to support a finding or an inference under 19 U.S.C. § 1677e(a)(2)(D) that all of the unexamined information (except where one of the three exceptions applied) was incorrect and therefore unverifiable. Because the Department’s decision to substitute facts otherwise available for all of this information necessarily required such a finding of fact, the Department’s decision to make the substitution under § 1677e(a)(2)(D) was contrary to law and must be set aside. 16

*1253 Nor is there record evidence indicating that Ñachi did anything that would have prevented Commerce from subjecting the unexamined information to verification, had Commerce chosen to do so. The only evidence is to the contrary. In the Decision Memorandum, Commerce cited time constraints, not Nachi’s failure to cooperate, as the reason for its not conducting further examination of Nachi’s data. Decision Mem. 35 (“We allotted one week to verify Nachi’s home-market sales.... [I]t would not be possible for us to examine the characteristics for all of the models a respondent reports in its sales database during the limited time for conducting verification of the data.”). 17

In summary, subparagraph (D) of § 1677e(a)(2) was applicable on this administrative record to the specific information that Commerce had found to be unverifiable because it was inconsistent with Na-chi’s technical drawings and catalogs. The procedure of subparagraph (D) was not available as to the body of reported information that Commerce actually verified. The Department’s finding or inference of unverifiability for the larger body of information (except where one of the exceptions applied) that Commerce made no attempt to verify was not supported by substantial evidence on the record. Commerce also erred in concluding that subparagraph (B) of § 1677e(a)(2) was applicable to any of these two categories of information, for which Commerce never made a finding of untimely submission.

In the Decision Memorandum, Commerce cited Micron Tech. v. United States, 117 F.3d 1386, 1396 (Fed.Cir.1997), for the principle that its methodology of conducting a spot check is reasonable. Micron did not involve an application by Commerce of 19 U.S.C. § 1677e(a)(2) in its present form but instead was concerned with a claim by the petitioner, the plaintiff in that case, that Commerce had conducted an inadequate verification of the cost of production data of two respondents that Commerce used in determining constructed value. Micron, 117 F.3d at 1388-89. The Court of Appeals, applying an abuse of discretion standard to verification procedures employed by Commerce, held that Commerce acted reasonably in conducting a spot check of the respondents’ financial data and did not thereby violate 19 U.S.C. § 1677e(b), which is now superseded but then required Commerce to use “best information available” if it was unable to verify the accuracy of submitted information. Micron, 117 F.3d at 1394-97. The case does not stand for the principle that Commerce, under the current statute, is empowered to substitute facts otherwise available for information it actually verified or unexamined information for which the record contains insufficient evidence to support a finding of unverifiability.

On remand, Commerce must revise its analysis under 19 U.S.C. § 1677e(a)(2)(D) and redetermine Nachi’s dumping margin accordingly. As required by this statutory provision, Commerce may substitute facts otherwise available only for the portion of Nachi’s reported information on physical characteristics that is the subject of a valid finding of unverifiability, i.e., a finding that *1254 is supported by substantial evidence on the record.

2. The Department’s Finding that Ñachi Failed to Cooperate by Not Acting to the Best of Its Ability to Comply with the Information Request Is SuppoHed by Substantial Record Evidence

Ñachi also contests the finding Commerce made in support of the determination to use “adverse inferences” in selecting substitute data for unverifiable data, i.e., the finding that Ñachi failed to cooperate by not acting the best of its ability to fulfill the Department’s information request. Ñachi Mem. 19. According to Ñachi, this finding is unreasonable and lacks a stated connection between the facts found and the choice made. Id. at 20. While acknowledging that some of the information submitted to Commerce was incorrect, Ñachi argues that Commerce properly cannot make a finding that Ñachi failed to cooperate to the best of its abilities simply because the correct information existed on Nachi’s books. Id. “Because such a justification would exist in every instance in which any error was made, it would be tantamount to requiring a perfect submission in order to avoid adverse facts available....” Id. at 11. Relying upon the Court of International Trade’s decision in Fujian Machinery and Equipment ImpoH & ExpoH Corp. v. United States, 25 CIT 1150, 1177, 178 F.Supp.2d 1305, 1334 (2001), and upon 19 U.S.C. §§ 1677e and 1677m, Ñachi contends that it is unreasonable for Commerce to expect that no reporting errors would occur and that a review could be “ ‘completely errorless.’ ” Nachi Mem. 20-22 (quoting Fujian, 25 CIT at 1177, 178 F.Supp.2d at 1334). According to Ñachi, Commerce must instead look at the “the totality of a respondent’s cooperation in an antidumping proceeding.” Id. at 11. Ñachi argues that the application of adverse inferences is not warranted with respect to its entries because Nachi’s reporting errors were inadvertent and only affected 2.7% of constructed export price comparisons, because the proceeding was especially complex, because this is the first review in which Ñachi has been required to report its information pursuant to the Department’s new model-match methodology, and because Ñachi fully cooperated with the Department’s requests for information. Id. at 12, 19, 24-26. Ñachi requests that the court instruct Commerce, “for the 2.7 percent of CEP comparisons affected by Na-chi’s reporting errors,” to “use as facts available the weighted-average of all possible normal values, including the normal value calculated based upon constructed value.” Id. at 26.

Defendant responds that Commerce properly drew an adverse inference because Ñachi failed to cooperate to the best of its ability in providing accurate physical characteristics. Def. Resp. 53. Defendant sets forth the principle that the standard of cooperation imposed by 19 U.S.C. § 1677e(b) assumes that respondents are familiar with the process and will “ ‘take reasonable steps to keep and maintain full and complete records.’ ” Id. (quoting Nippon Steel Corp. v. United States, 337 F.3d 1373, 1382 (Fed.Cir.2003)). In defendant’s view,

[bjecause Commerce discovered errors each time it attempted to examine the reported physical characteristics, and because Commerce discovered that Ñachi could have used its own drawings and catalogs to provide the correct information, Commerce correctly determined that the incorrect reporting was not an isolated mistake but rather a systemic occurrence resulting from Nachi’s failure to act to the best of its ability.

Id. Defendant further argues that the revised model-match methodology did not change the actual product characteristics that must be reported, and that Nachi’s *1255 alleged difficulties with the additional reporting requirements of the new methodology are not adequately explained. Id. at 54.

Under 19 U.S.C. § 1677e(b), if Commerce “finds that an interested party has failed to cooperate by not acting to the best of its ability to comply with a request for information,” Commerce, “in reaching the applicable determination ... may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available.” 19 U.S.C. § 1677e(b). The statutory obligation to act to the best of one’s ability in complying with an agency’s request for information requires the respondent to do the “maximum” it is able to do, which includes putting forth its maximum efforts to obtain the requested information from its records. Nippon Steel Corp., 337 F.3d at 1382-83.

Commerce found, and Ñachi does not dispute, that specific data describing the physical characteristics of some of Nachi’s models were not reported correctly. Decision Mem. 36; Ñachi Mem. 11, 19. Still, Ñachi argues, and argues correctly, that the statute does not require a completely errorless review. See Nippon Steel Corp., 337 F.3d at 1382 (stating that “the standard does not require perfection and recognizes that mistakes sometimes occur”). Nevertheless, the evidence of record in the review demonstrates that Ñachi provided Commerce incorrect information on physical bearing characteristics at a frequency (nineteen errors affecting sixteen out of forty models reviewed) that cannot be described as negligible or inconsequential. The record evidence consists of the documentation Commerce produced to report on the verification process, in which documentation it is reported that information in Nachi’s technical drawings and catalogs was inconsistent with Nachi’s questionnaire responses. Decision Mem. 36. Commerce found that Ñachi “had the correct data available to it in the technical-specification sheets it maintains in its normal course of business and in its catalog,” id., and used the “technical specifications and catalog descriptions provided by Ñachi at verification to compare them for accuracy against the reported physical characteristics.” Id. at 37. Commerce further found that “Ñachi has not proffered an adequate or reasonable justification for why it could not have reported the correct characteristics.” Id. at 36. Nachi’s argument that this was the first review in which Ñachi was required to report its information pursuant to the Department’s new model-match methodology is unavailing. As Commerce pointed out, Commerce has required “Ñachi, and all respondents, to provide such information in every review since the first review.” Decision Mem. at 36-37 (citing Ball Bearings & Parts Thereof From France, Germany, Italy, & Singapore: Prelim. Results of Antidumping Duty Admin. Reviews, Partial Rescission of Admin. Reviews, & Notice of Intent To Revoke Order In Part, 68 Fed. Reg. 6404, 6408 (Feb. 7, 2003)). The correct reporting of information on physical characteristics of bearings was also important under the Department’s prior model-match methodology. For these various reasons, the court is unable to agree with Ñachi that the record lacks substantial evidence to support the Department’s finding that Ñachi failed to cooperate by not acting to the best of its ability to comply with the Department’s request for information on the physical characteristics of Nachi’s bearings. The court concludes that Nachi’s reporting of these data fell short of the standard established by 19 U.S.C. § 1677e(b) as interpreted by the Court of Appeals in Nippon Steel Corp., 337 F.3d at 1382.

*1256 S. The Court Affirms the Department’s Choice of the 48.69% Rate

Ñachi claims that Commerce erred in applying the 48.69% rate as an adverse inference. Ñachi Mem. 12. Ñachi argues that this seventeen-year-old rate bears no relationship to Nachi’s pricing and cost structure in the current review and objects that the rate chosen is “five times the rate found for Ñachi in the most recent review in which Ñachi participated prior to the 2004-2005 review.” Id. at 13.

In explaining the choice of the 48.69% rate, the Final Results state that “[a]s adverse facts available, we have selected the highest margin we have determined for Ñachi in any previous segment of this proceeding....” Final Results, 71 Fed. Reg. at 40,066. In the Decision Memorandum, Commerce expressed a finding that this rate is corroborated by the fact that a substantial number of Nachi’s transactions for which Commerce determined positive margins in the subject review had margins higher than the 48.69% rate. Decision Mem. 37.

The Court of Appeals has stated that “[i]t is clear from Congress’s imposition of the corroboration requirement in 19 U.S.C. § 1677e(c) that it intended for an adverse facts available rate to be a reasonably accurate estimate of the respondent’s actual rate, albeit with some built-in increase intended as a deterrent to non-compliance.” F.lli De Ceceo Di Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000) (“De Ceceo”). Congress imposed the corroboration requirement “to prevent the petition rate (or other adverse inference rate), when unreasonable, from prevailing and to block any temptation by Commerce to overreach reality in seeking to maximize deterrence.” Id. The Court of Appeals, while recognizing the breadth of the Department’s discretion under 19 U.S.C. § 1677e(b), also cautioned that “Commerce’s discretion in these matters, however, is not unbounded.” Id. As the Court of Appeals reiterated,

“Congress could not have intended for Commerce’s discretion to include the ability to select unreasonably high rates with no relationship to the respondent’s actual dumping margin. Obviously a higher adverse margin creates a stronger deterrent, but Congress tempered deterrent value with the corroboration requirement. It could only have done so to prevent the petition rate (or other adverse inference rate), when unreasonable, from prevailing and to block any temptation by Commerce to overreach reality in seeking to maximize deterrence.”

Ta Chen Stainless Steel Pipe, Inc. v. United States, 298 F.3d 1330, 1340 (Fed.Cir.2002) (quoting De Ceceo, 216 F.3d at 1032).

The 48.69% rate was a rate Commerce actually assigned to Ñachi, but it occurred in 1989 in the final less-than-fair-value determination resulting from the antidumping duty investigation. Japan Final Determination 1987-1998, 54 Fed.Reg. at 19,108. Because it is so remote in time, the rate might be seen as having only an attenuated relationship to Nachi’s actual dumping margin in the subject reviews. Nevertheless, Commerce’s reliance on record evidence of the number of transactions in the current reviews that had individual margins exceeding the 48.69% rate must be considered sufficient to satisfy the corroboration requirement of 19 U.S.C. § 1677e(c).

Timken directs the court’s attention to PAM, S.p.A. v. United States, 582 F.3d 1336 (Fed.Cir.2009), in which the Court of Appeals recently held to be corroborated a rate of 45.49% that Commerce chose as an adverse inference for a respondent that had failed to report all of its home market sales in the sixth administrative review of *1257 an antidumping duty order. PAM, S.p.A. v. United States, 582 F.3d 1336, 1340-41 (Fed.Cir.2009). In support of a finding of corroboration in PAM, Commerce relied on 29 sales by the same respondent in the fourth review of the order (ie., two reviews prior to- the review under consideration) with margins above 45.49%, which sales represented only 0.5% of the respondent’s total sales of subject merchandise during the period of the fourth administrative review. Id. at 1340. The Court of Appeals in PAM rejected the respondent’s argument that these sales constituted “outlier data,” concluding that the respondent’s position was rejected in Ta Chen, 298 F.3d at 1339. Id.

PAM does not hold that corroboration of a specific rate chosen as an adverse inference necessarily is established in any case in which 0.5% or more of individual sales of the non-cooperating respondent have dumping margins above that rate. To the contrary, the Court of Appeals implicitly recognized that the question of whether the record evidence relied on for corroboration meets the substantial evidence standard must be considered according to the record as a whole. See PAM, 582 F.3d at 1340 (“The burden imposed by substantial evidence review may not be heavy, but it is not ephemeral. There must be at least enough evidence to allow reasonable minds to differ.”). Additionally, a finding of corroboration is not dispositive of the question of whether Commerce has discretion to apply a particular rate as an adverse inference; that question, too, must be decided according to the relevant factual circumstances as shown by the entire record. Once Commerce satisfies the statutory corroboration requirement that the statute imposes on the use of secondary information, Commerce is entitled to exercise a significant, albeit still not unlimited, measure of discretion in selecting, from among the facts otherwise available, those facts that will provide “a proper deterrent to non-cooperation.” De Ceceo, 216 F.3d at 1032.

In PAM, the Court of Appeals was confronted with a circumstance in which PAM, the party against whom the adverse inference was used, was found to have withheld requested information on sales that “combined amounted to about two-thirds of PAM’s total domestic sales.” PAM, 582 F.3d at 1338. The Court of Appeals reasoned that “Commerce’s discretion in applying an AFA margin is particularly great when a respondent is uncooperative by failing to provide or withholding information.” Id. at 1340. In this case, Ñachi did not withhold information. The finding of non-cooperation instead was based on a failure to exercise due care in reporting the information contained in the business records.

With respect to support for the corroboration finding in this case, the court’s examination of the record evidence supports the finding in the Decision Memorandum that a “substantial number” of Nachi’s sales in the AFBs 16 reviews had margins exceeding 48.69%. Decision Mem. 37; see Ñachi Mem. 8-9. The court concludes, therefore, that substantial record evidence supports the Department’s finding of corroboration for its chosen adverse inference rate of 48.69%.

Based on the entire record in this case, the court concludes, further, that the use of the 48.69% rate does not exceed the scope of Commerce’s discretion under 19 U.S.C. § 1677e(b). Commerce discovered that Nachi’s reporting errors were present in a significant percentage of the transactions that Commerce examined at verification. It is essential to the proper administration of the statute that respondents participating in an antidumping proceeding endeavor to provide Commerce correct information about the sales of their merchandise. See PAM, 582 F.3d at 1339 *1258 (“Congress has made very clear the importance of accurate and complete reporting of home market sales.”)- In reviewing the exercise of Commerce’s discretion under § 1677e(b), a court must be mindful that Commerce “is in the best position, based on its expert knowledge of the market and the individual respondent, to select adverse facts that will create a proper deterrent to non-cooperation with its investigations and assure a reasonable margin.” De Ceceo, 216 F.3d at 1032. The rate Commerce selected was, as Ñachi points out, five times the rate Commerce assigned to Ñachi in its most recent review. Ñachi Mem. 13. Nevertheless, the court cannot conclude that the choice of rate was unreasonable or an abuse of discretion under all the circumstances present here, including the circumstance that the party’s efforts to report the physical data correctly did not even come close to satisfying the standard for cooperation that 19 U.S.C. § 1677e(b) establishes.

I. Commerce Did Not Provide a Full and Adequate Rationale for its Decision to Use Japanese Interest Rates to Calculate a Portion of the Adjustment for Imputed Interest Candying Costs when Determining Constructed Export Prices for NTN and Ñachi

In determining constructed export price, Commerce deducts from the price paid in the first unaffiliated resale (the “starting price”), as a selling expense under 19 U.S.C. § 1677a(d)(l), an amount representing inventory carrying cost of the subject merchandise in the United States, to the extent the inventory carrying cost relates to that resale. See 19 U.S.C. § 1677a(d)(l). To quantify the deduction, Commerce determines an applicable interest rate. In the administrative review, Commerce used Japanese interest rates when calculating certain U.S. inventory carrying costs with respect to two respondents, Ñachi and NTN. Decision Mem. 40-41. Commerce found as a fact that in both instances, the Japanese foreign parent, by extending credit terms to the U.S. affiliate, assumed the burden of the U.S. inventory carrying costs for the portion of time that the merchandise was held in the inventory of the U.S. affiliate and concluded that “it is that party’s short-term interest rate that we should use in calculating inventory carrying costs” for that time period. Id. Timken claims that Commerce acted contrary to law, and, specifically, contrary to its own regulations, in using Japanese interest rates to calculate the inventory carrying costs. Timken U.S. Corporation’s Mem. in Supp. of its Rule 56.2 Mot. for J. on the Agency R. 7 (“Timken Mem.”). Timken maintains that Commerce instead should have used interest rates prevailing in the United States. Id. at 24. NTN and Ñachi urge the court to reject Timken’s claim. Resp. of Pis.’ NTN Corp., NTN Bearing Corp. of Am., Am. NTN Bearing Mfg. Corp., NTN-BCA Corp., NTN-Bower Corp., & NTN Drive-shaft, Inc. to the Rule 56.2 Mot. of Def.Intervenor Timken U.S. Corp. 2 (“NTN Resp.”); Resp. Br. of Pis. Nachi-Fujikoshi Corp., Ñachi Am., Inc., & Ñachi Technology, Inc. in Opp’n to Def.-Intervenor Timken U.S. Corp.’s Mot. for J. on the Agency R. 4 (“Ñachi Resp.”). NTN argues that Commerce correctly determined in the Final Results that “the party that extends the payment terms to its U.S. affiliate carries the financial burden of financing the inventory.” NTN Resp. 7; see Ñachi Resp. 4. NTN and Ñachi also argue that the Department’s determination to apply the Japanese interest rate rather than the U.S. interest rate reflected the commercial and financial reality of NTN Japan’s financing of the inventory cost. NTN Resp. 10; see Ñachi Resp. 4.

Timken’s first argument is that the Department’s choice of Japanese interest rates improperly relied on the payment *1259 terms between the producer and the affiliated reseller in calculating constructed export price using the sales between the affiliated reseller and the unaffiliated purchaser. Timken Mem. 9-11. Timken construes 19 U.S.C. § 1677a(d) to prohibit Commerce from relying on “affiliated party data” for this purpose, arguing that

to the extent the Japanese interest rates are lower, and to the extent the payment terms extend beyond the date of importation — thus reducing the period calculated at U.S. interest rates, Commerce has permitted the affiliated parties to reduce the adjustment and inflate the U.S. price, and, to that extent, has failed in its duty to prevent those parties from competing unfairly in the U.S. market.

Id. at 11. According to Timken, this result is contrary to the intent of the statute and to the decision of the Court of Appeals in AK Steel Corp. v. United States, 226 F.3d 1361, 1367 (Fed.Cir.2000). Timken Mem. 11.

The court finds nothing in § 1677a(d) that prohibits Commerce, expressly or impliedly, from using credit terms existing between the foreign parents and the U.S. affiliates in reaching its findings that, in the case of both Ñachi and NTN, the foreign parent bore the burden of inventory carrying costs subsequent to importation of the subject merchandise. See Decision Mem. 40-41. Commerce concluded that, in calculating the inventory carrying cost, it was reasonable to use the short term interest rates to which the parties actually bearing the inventory costs — ie., the foreign affiliates — were subject. See 19 U.S.C. § 1677a(d). Timken fails to point to any specific language in § 1677a(d) that the court could construe as posing a barrier to the Department’s reaching this conclusion. The subsection requires Commerce to reduce the starting price by “the amount of’ the selling expense without specifying how Commerce is to determine that amount. Id. The Department’s quantifying of the inventory carrying cost according to the interest rate to which the party bearing the cost was subject appears to the court to be reasonable.

The credit terms constituted substantial record evidence that supported the Department’s findings. See Decision Mem. 41 (stating that “payment terms demonstrate that NTN bears the cost of carrying the merchandise for a portion of the time that the merchandise is in inventory”). The court is aware of no rule or principle that would prohibit Commerce from using, as evidence in support of those findings, credit terms that were negotiated between the related parties and that are disclosed by record evidence. In arguing to the contrary, Timken advocates as a general rule or principle that “[t]he cost to be measured to determine an adjustment to U.S. price for inventory carrying cost is the cost of holding inventory, not the cost of extending credit, and so the proper credit rate is that of the company holding the inventory.” Timken U.S. Corp.’s Reply Br. 2 (“Timken Reply”). The court is not convinced by this argument that it must adopt the general rule or principle that Timken advocates. Moreover, the distinction Timken attempts to draw would disregard the significance of the factual findings that the foreign parents, not their U.S. affiliates, bore the actual inventory carrying costs at issue. See Decision Mem. 41. According to those findings, the “cost of holding inventory” was “the cost of extending credit,” ie., it was the credit cost as borne by the foreign parents. Timken Reply 2.

The Court of Appeals in AK Steel Corp., 226 F.3d 1361, does not establish a rule or principle that prohibits Commerce from using credit terms existing between the foreign parents and the U.S. affiliates in reaching its findings that, in the case of *1260 both Ñachi and NTN, the foreign parent bore the burden of inventory carrying costs subsequent to importation of the subject merchandise. Timken cites to general language in the case explaining that the purpose of the statutory distinction between constructed export price transactions and export price transactions is to isolate an arm’s-length transaction, and that the purpose of making deductions from the starting price to arrive at a constructed export price “ ‘is to prevent foreign producers from competing unfairly in the United States market by inflating the U.S. Price with amounts spent by the U.S. affiliate on marketing and selling the products in the United States.’ ” Timken Mem. 10 (quoting AK Steel Corp., 226 F.3d at 1367). Nothing in the decision of the Court of Appeals establishes that Commerce acted contrary to law in relying on the credit terms for its findings.

NTN, Ñachi, and the defendant United States rely on certain cases in which courts previously have affirmed the Department’s use of foreign interest rates to quantify the cost of carrying U.S. inventory where that cost is borne by the foreign affiliate. NTN Resp. 6 (citing LMI-La Metalli Industriale, S.p.A. v. United States, 912 F.2d 455, 461 (Fed.Cir.1990); Thai Pineapple Pub. Co. v. United States, 20 CIT 1312, 1329-30, 946 F.Supp. 11, 26-27 (1996); Timken Co. v. United States, 18 CIT 619, 625-26, 858 F.Supp. 206, 212-13 (1994)); Nachi Resp. 5 (citing Timken Co. v. United States, 18 CIT 942, 948-49, 865 F.Supp. 881, 886-87 (1994)); see Def. Resp. 60-62. Timken argues that the pri- or cases are distinguishable in arising before the enactment of the Uruguay Round Agreements Act (“URAA”) and specifically before Commerce promulgated the current 19 C.F.R. § 351.402(b) (1997) as an implementing regulation. Timken Mem. 12-17, 20-21. Timken argues, further, that the Department’s use of Japanese interest rates is inconsistent with the amended statute and the regulation. Id. However, the enactment of 19 U.S.C. § 1677a(d) by the URAA does not lend support to Timken’s argument. As discussed above, nothing in § 1677a(d) expressly or impliedly disallows the methodology Commerce employed here to quantify the inventory carrying cost.

The court is unable to agree with Timken’s argument that the use of Japanese interest rates violates 19 C.F.R. § 351.402(b). The regulation provides that “ ‘[i]n establishing constructed export price under section 772(d) of the Act [19 U.S.C. § 1677a(d) ], the Secretary will make adjustments for expenses associated with commercial activities in the United States that relate to the sale to an unaffiliated purchaser, no matter where or when paid ’ ” and further provides that “ ‘[t]he Secretary will not make an adjustment for any expense that is related solely to the sale to an affiliated importer in United States, although the Secretary may make an adjustment to normal value for such expenses under section 773(a)(6)(C)(iii) of the Act [19 U.S.C. § 1677b(a)(6)(C)(iii) ].’ ” Timken Mem. 13 (quoting 19 C.F.R. § 351.402(b)). In taking issue with the Department’s relying on the payment terms between the foreign sellers and their U.S. affiliates in support of the decision to use the Japanese interest rates, Timken argues that “[h]ere, the payment terms which the affiliate (reseller) and its parent (producer) agreed on incontrovertibly relate ‘solel/ to the Japanese parent’s sale of bearing products to its affiliated importer in the United States.” Id. This characterization of the payment terms is inconsistent with the Department’s finding that the Japanese foreign parents, by extending credit terms to the U.S. affiliates, assumed the burden of the U.S. inventory carrying costs for the portion of time that the merchandise was held in the inventory of the U.S. affiliate. With respect to *1261 Timken’s argument based on 19 C.F.R. § 851.402(b), the regulatory provision at issue is addressed to whether a particular expense is related to the sale of the subject merchandise from the foreign seller to the unaffiliated U.S. importer or, alternatively, to the resale to an unaffiliated purchaser. 19 C.F.R. § 351.402(b). Under the regulation, only if the expense relates to the sale to the unaffiliated purchaser will Commerce make a deduction from the starting price when determining constructed export price. Id. In providing that the Secretary “will make certain of the adjustments” for expenses associated with commercial activities in the United States that relate to the sale to an unaffiliated purchaser, the regulation governs whether the Secretary will make the adjustment, not how the Secretary will make it. Id. In this case, the expense involved is the carrying cost of subject merchandise from the time of importation until the time payment is due to the affiliated foreign seller. See Decision Mem. 4(M1, 43. Commerce found, according to substantial record evidence, that the expenses were associated with commercial activities in the United States that related to the sales to the unaffiliated purchasers.

Timken also argues that the Department’s reliance on the payment terms as record evidence contradicts the logic and purpose of calculating the carrying costs. Timken Mem. 17. Timken submits that the issue of whether the parent or the affiliate bears the actual expense associated with the carrying costs is not material to the determination of those carrying costs because Commerce does not use actual costs but instead estimates cost according to objective factors such as the period of time that credit is extended, the period of time merchandise remains in inventory, and the cost of borrowing money in the market. Id. at 17-18. Timken quotes the Department’s explanation set forth in the final results of the first administrative review, in which Commerce explains its reasons for calculating an estimated cost based on objective criteria rather than using actual costs. Id. at 18 (quoting Antifriction Bearings (Other Than Tapered Roller Bearings) & Parts Thereof From the Federal Republic of Germany; Final Results of Antidumping Duty Admin. Review, 56 Fed.Reg. 31,692, 31,727 (July 11, 1991)). The Department’s explanation, which occurred in the context of the Department’s rejection of the petitioner’s argument that it is appropriate to apply adjustments for inventory carrying cost only for the United States price, not the foreign market value, concerned the issue of whether it is appropriate to impute a cost where it is not possible to identify actual cost. It did not speak directly to the question of which interest rate is appropriate in this case.

Finally, Timken argues that in using payment terms to determine the period for which the Japanese interest rates apply, Commerce departed from the evidentiary standard it applied in previous proceedings. In support of this argument, Timken relies on the Department’s decisions in Certain Welded Stainless Steel Pipe From Taiwan: Final Results of Antidumping Duty Admin. Review & Determination To Revoke Order In Part, 65 Fed.Reg. 39,367 (June 26, 2000) (“Welded Steel Pipe from Taiwan Final Results ”) (incorporating Issues & Decision Mem. for the Admin. Review of Certain Welded Stainless Steel Pipe from Taiwan: Dec. 1, 1997 through Nov. SO, 1998, at Comment 1 (June 19, 2000) (“Welded Steel Pipe from Taiwan Decision Mem.”)), and Notice of Final Results of Antidumping Duty Admin. Review & Notice of Final Results of Anti-dumping Duty Changed Circumstances Review: Certain Softwood Lumber Products From Canada, 69 Fed.Reg. 75,921 (Dec. 20, 2004) (“Softwood Lumber from Canada Final Results ”) (incorporating Is *1262 sues & Decision Mem. for the Final Results of the Antidumping Duty Admin. Review of Certain Softwood Lumber Products From Canada 141-42 (Dec. 13, 2004) (at Comment 39) (“Softwood Lumber from Canada Decision Mem.”)). Timken Mem. 21-24; Timken Reply 5-6. Timken submits that payment terms requiring payment by a specific date do not establish that payment occurred on that date, and that Commerce therefore should have applied U.S. interest rates for part of the period in question due to an alleged lack of proof of actual payments by the U.S. affiliate to the foreign parent on, and not before, the due dates. Timken Reply 5-6. Defendant responds that payment terms constitute sufficient record evidence to support the determination that the U.S. affiliates benefitted from the Japanese parents’ ability to borrow money in Japan. Def. Resp. 61 (citing Thai Pineapple, 20 CIT at 1329-30, 946 F.Supp. at 26-27).

In Welded Steel Pipe from Taiwan Decision Mem., Commerce stated that:

It is the Department’s practice to use the short-term borrowing rate in the currency in which the cost of the inventory is incurred by the entity that bears the cost of producing or acquiring such inventory. The Department deviates from this practice only in instances where there is clear evidence that an entity other than the one holding the merchandise in inventory absorbs the full cost of financing the cost of the merchandise during the time that the merchandise is held in inventory. In this case, Ta Chen [the foreign producer] provided no clear record evidence that it, and not TCI [the U.S. affiliate], incurs the cost of the merchandise in inventory during the entire period of credit days given to TCI. Since there is no record evidence of when payment is made by TCI, the Department cannot assume that TCI does not pay Ta Chen until the last day of the payment period.

Welded Steel Pipe from Taiwan Decision Mem., at Comment 2; Welded Steel Pipe from Taiwan Final Results, 65 Fed.Reg. 39,367-68. In Certain Softwood Lumber Products from Canada, Commerce stated that it agreed with the position of respondent Weyerhaeuser, who commented that the correct rate for use in calculating inventory carrying costs “is the short-term borrowing rate of the company that holds title to the subject merchandise and invoices the ultimate customer.” Softwood Lumber from Canada Decision Mem. 141. In light of the language of these two administrative decisions, the court cannot rule out the possibility that Commerce, subsequent to the decisions in LMI, 912 F.2d at 461, Thai Pineapple, 20 CIT at 1329-30, 946 F.Supp. at 26-27, and Timken, 18 CIT at 625-26, 858 F.Supp. at 212-13 (on which Commerce relies in the Decision Memorandum, Decision Mem. 41), has adopted a practice or established methodology from which it departed in some way in reaching the decision to use Japanese interest rates in the AFBs 16 reviews. This specific issue is not addressed in the Decision Memorandum. Although citing LMI and Thai Pineapple for the principle that courts have affirmed the use of foreign interest rates to value carrying costs for inventory in the United States, defendant does not address in its response to Timken’s motion for judgment upon the agency record the specific issue of whether the decision to use Japanese interest rates in the reviews was a departure from a practice or established methodology, and if so, whether Commerce knowingly departed from that practice or methodology and provided adequate reasoning for doing so. See Def. Resp. 60-61. Accordingly, the court is directing that Commerce include in its remand redetermination an analysis responding to Timken’s argument concerning a departure from an alleged practice or methodology *1263 and reconsider accordingly its decision to use Japanese interest rates when calculating the subject U.S. inventory carrying costs with respect to Ñachi and NTN.

III. Conclusion

For the reasons discussed in the foregoing, the court will affirm in part, and remand in part, the Final Results, and will set aside as unlawful certain decisions and determinations in the Final Results.

ORDER

Upon consideration of all papers and proceedings herein, it is hereby

ORDERED that the Rule 56.2 motions for judgment upon the agency record of JTEKT and NSK, be, and hereby are, denied; it is further

ORDERED that the Rule 56.2 motions for judgment upon the agency record of NBP, NTN, Ñachi be, and hereby are, granted in part and denied in part; it is further

ORDERED that the Rule 56.2 motion of Timken be, and hereby is, granted to the extent that the Department is directed on remand to reconsider its decision to allow use of Japanese interest rates to calculate a portion of the adjustment for imputed interest carrying costs when calculating constructed export prices for NTN and Ñachi and to include in its remand redetermination an analysis responding to the issue Timken raised concerning departure from a practice or established methodology concerning choice of interest rate, as provided herein; it is further

ORDERED that the final determination of the International Trade Administration, United States Department of Commerce (“Commerce” or the “Department”), published as Ball Bearings & Parts Thereof from France, Germany, Italy, Japan, & the United Kingdom: Final Results of Antidumping Duty Administrative Reviews, 71 Fed.Reg. 40,064, 40,065 (July 14, 2006) (the “Final Results”), be, and hereby is, affirmed in part and set aside in part as contrary to law, and remanded to the Department as provided herein; it is further

ORDERED that the following decisions and determinations by Commerce in the Final Results, for the reasons that are set forth in the following respective subparts of Part II of this Opinion and Order, be, and hereby are, affirmed: (A) the decision to apply the Department’s zeroing procedure; (B) the decision to discontinue the use of the previous model-match methodology and the decisions Commerce reached in applying the Department’s revised model-match methodology, except the decision to reject NPB’s proposal to expand the choice of sampling months and the decision to reject NTN’s proposal to incorporate into the model-match methodology additional design-type categories for specific types of ball bearings; (C) the decision to treat JTEKT and its affiliate as a single entity; (D) the decision to deduct certain additional benefits expenses when determining the constructed export price of NSK’s subject merchandise; (F) the recalculated determination of NTN’s home market packing materials expenses; (G) the decision to disallow certain downward price adjustments resulting from NTN’s allocation of discounts to home market customers; and (H) the factual determination that Ñachi failed to cooperate by not acting to the best of its ability to comply with the Department’s request for certain information on physical bearing characteristics and the Department’s selection of the 48.69% rate; it is further

ORDERED that the following decisions and determinations by Commerce in the Final Results, for the reasons that are set forth in the respective sections of Part II of this Opinion and Order, be, and hereby are, set aside as contrary to law: (E) the Department’s redetermination of NTN’s *1264 freight expense; and (H) the Department’s decision to substitute facts otherwise available for information that Ñachi submitted on physical bearing characteristics, except for the specific information that Commerce determined during its verification procedure to be incorrect; it is further

ORDERED that Commerce, on remand, shall reconsider its decision to reject NPB’s proposal to expand the choice of sampling months and NTN’s proposal to incorporate into the model-match methodology additional design-type categories, as reviewed in subparts (B.3) and (B.4), respectively, of Part II of this Opinion and Order; shall redetermine NTN’s freight expense using a method that is consistent with the Department’s treatment of the freight expense of other respondents in the administrative reviews that are the subject of this action, as set forth in subpart (E) of Part II of this Opinion and Order; shall redetermine the application of facts otherwise available for information that Ñachi submitted on physical bearing characteristics as discussed in subpart (H) of Part II of this Opinion and Order; and shall reconsider its decision to allow use of Japanese interest rates to calculate a portion of the adjustment for imputed interest carrying costs when calculating constructed export prices for NTN and Ñachi as discussed in subpart (I) of Part II of this Opinion and Order, and in doing so, shall provide an analysis responding to Timken’s argument concerning a departure from an alleged practice or methodology; it is further

ORDERED that Commerce shall redetermine the weighted-average dumping margins of plaintiffs, as appropriate, in complying with this Opinion and Order; it is further

ORDERED that Commerce shall have ninety (90) days from the date of this Opinion and Order in which to file its redetermination upon remand, which shall comply with all directives in this Opinion and Order; it is further

ORDERED that plaintiffs and defendant-intervenor shall have thirty (30) days from the filing of the remand redetermination in which to file comments thereon.