26 I.T.R.D. (BNA) 1412
Ct. Intl. Trade
2004

OPINION

TSOUCALAS, Senior Judge.

Plaintiffs, Timken U.S. Corporation 1 *1272 (“Timken”) and Timken Nadellager, GmbH 2 (“TNG”), move pursuant to US-CIT R. 56.2 for judgment upon the agency record challenging one aspect of the United States Department of Commerce’s (“Commerce”) determination entitled Final Results of Antidumping Duty Administrative Reviews and Revocation of Orders in Part on Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom (“Final Results”), 65 Fed.Reg. 49,219 (Aug. 11, 2000). Specifically, plaintiffs contend that Commerce erred by refusing to correct an error in reporting channels of distribution for TNG, an affiliated company of Timken and German producer of cylindrical roller bearings (“CRBs”). Commerce maintains that Timken failed to demonstrate that the error was clerical in nature and argues that the record supports Commerce’s finding that the alleged error was either a substantive issue or an error in judgment.

BACKGROUND

The administrative review at issue involves the period of review covering May 1, 1998, through April 30, 1999. 3 See Final Results, 65 Fed.Reg. at 49,219. Commerce published the preliminary results of the subject reviews on April 6, 2000. See Preliminary Results of Antidumping Duty Administrative Reviews, Partial Rescission of Administrative Reviews, and Notice of Intent to Revoke Orders in Part on Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom (“.Preliminary Results”), 65 Fed.Reg. 18,033. Commerce published the Final Results at issue on August 11, 2000.

JURISDICTION

The Court has jurisdiction over this matter pursuant to 19 U.S.C. § 1516a (2000) and 28 U.S.C. § 1581(c) (2000).

STANDARD OF REVIEW

The Court will uphold Commerce’s final determination in an antidumping administrative review unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.... ” 19 U.S.C.' § 1516a(b)(l)(B)(i) (1994); see NTN Bearing Corp. of Am. v. United States, 24 CIT 385, 389-90, 104 F.Supp.2d 110, 115-16 (2000) (detailing Court’s standard of review in antidumping proceedings).

*1273 DISCUSSION

A. Factual Background

During the tenth administrative review, Commerce instructed Timken to report the channels of distribution for TNG’s home market sales. See App. Timken’s Mem. Supp. R. 56.2 Mot. J. Upon Agency R. (“Timken’s App.”) Tab 4 at 1. Commerce indicated that “the information [was] necessary to make appropriate comparisons of sales at the same level of trade or to adjust normal value, if appropriate, when sales are compared at different levels of trade.” Def.’s Mem. Opp’n Pis.’ Mot. J. Upon Agency R. (“Def.’s Mem”) Ex. 1 at A-4 (emphasis omitted). In response to this request, Timken identified five distribution channels corresponding to customer categories for TNG’s home market sales, including: (1) factory to large original equipment manufacturers (“OEMs”) (“channel 1”); (2) factory to other OEMs (“channel 2”); (3) factory to distributors (“channel 3”); (4) TNG to OEM customers (“channel 4”); and (5) TNG to distributor customers (“channel 5”). See Timken’s App. Tab 3 at A-21 — A-24. Timken also described the sales process for each channel of distribution. See id.

Commerce eventually “conducted a sales verification” of Timken in February 2000, and subsequently issued an analysis memorandum for the Preliminary Results entitled Issues and Decision Memorandum for the Administrative Reviews of Anti-friction Bearings (other than tapered roller bearings) and parts thereof from France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kindom — May 1 1998, through April 30, 1999 (“Issues & Decision Mem.”). See id. Tab 10 at 1. In its decision memorandum, “Commerce explained that for the five channels of distribution reported by [TNG] in its [home market] sales listing, [a senior import compliance specialist] examined the selling activities, the point in the channel of distribution at which the selling activities occurred, and the types of customers that purchased the foreign like product from [TNG].” Def.’s Mem. at 5. Commerce determined that three channels of distribution were for home market sales by TNG, and the remaining two channels of distribution were for “resales” by TNG’s affiliated marketing entity. See Timken’s App. Tab 4 at 3. As a result, Commerce re-designated channel 1 as HM1 (home market 1), grouped channels 4 and 5 together and re-designated them HM2 (home market 2) and also grouped channels 2 and 3 together and redesignated them HM3 (home market 3). See id. Essentially, Commerce grouped channels 4 and 5 and channels 2 and 3 together upon a determination that the points in which selling activities occurred within these different channels of distribution were indistinguishable. See id.

In determining the dumping margin, Commerce used constructed value (“CV”) when no sales existed of “an identical or similar model sold in the home market or when the identical or similar model was disregarded as below cost.” Id. at 7. 4 “Profit was calculated by multiplying the level-of-trade-specific weighted-average profit rate calculated on home market sales made in the ordinary course of trade by the cost of production ... of the model.” Id. Based on this analysis, Commerce calculated a dumping margin of 61.60 percent for CRBs produced by TNG. See id. at 1; Final Results, 65 Fed.Reg. at 49,221; *1274 Preliminary Results, 65 FecLReg. at 18,-041.

B. Timken’s Contentions

Timken states that in responding to Commerce’s request for information, Timken relied on “customer names to classify its home market sales according to distribution channel.” Timken’s Mem. Supp. R. 56.2 Mot. J. Upon Agency R. (“Timken’s Mem.”) at 6. As a result, a certain number of transactions were unintentionally reported in channel 1 instead of channel 2 or channel 3. Ultimately, these transactions were grouped into HM1 instead of HM3. See id. at 7-13.

According to Timken, these misclassifi-eations can be grouped into three broad fact patterns. First, Timken inadvertently reported a certain number of transactions as sales to large OEMs instead of classifying them as sales to small OEMs. See id. at 7. Second, Timken unintentionally reported a certain number of transactions as sales for the production of large original equipment instead of sales for use as replacement parts. See id. at 8. Third, Timken mistakenly reported a certain number of transactions as sales to large OEMs instead of labeling them as “units sold as samples, [that] were delivered to the customer’s prototype center.” Id. at 10 (emphasis omitted). “Although Commerce verified [Timken’s] home market sales, [Timken argues that Commerce] did not pursue the classification o[f] individual transactions into their appropriate distribution channels at verification.” Id. at 11 (citations omitted).

Timken claims that Commerce’s subsequent calculation of an abnormally high CV profit rate resulted from these misclassified transactions. See id. at 11-12. According to Timken, these transactions only covered a minimal percentage “of all units sold in the [level of trade (‘LOT’)] (as a percentage of sales quantity reported in the home market sales listing).” Id. at 11. Timken maintains that the remaining sales in the LOT were “disregarded” because Commerce found them to have been made at prices below cost of production (“COP”). See id. Timken further claims that the calculation of an abnormal CV profit rate caused Commerce to compute an inaccurate dumping margin. See id. at 11-12.

During the administrative review, Timken identified these mistakes and requested that Commerce either correct Timken’s “inadvertent errors by reclassifying certain home market sales, or ... in the alternative, combine all home market LOTs in the CV-profit calculation and use that rate for home market LOT.” Id. at 12 (emphasis omitted). Timken supplied supporting documentation and claims that Commerce did not indicate that such information was inadequate, nor did Commerce request additional supporting evidence. See id. at 12-13.

In the Issues & Decision Mem., Commerce rejected Timken’s arguments that the errors should be corrected because Timken did not show that these errors were clerical in nature. See Timken’s App. Tab 10 at 50-52. Timken contends that Commerce is mandated to correct these errors because they were unintentional and because “Commerce’s goal in administrative reviews is to determine margins ‘as accurately as possible.’ ” Timken’s Mem. at 17 (citing Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir.1990)). Timken argues that the present situation involving mis-classi-fied channels of distribution is similar to the mis-categorized sales situation identified in NTN Bearing Corp. v. United States, 74 F.3d 1204, 1206-09 (Fed.Cir.1995). See Timken’s Mem. at 17-18, 26-31. Timken points out that in NTN, the United States Court of Appeals for the Federal Circuit (“CAFC”) held that “Com *1275 merce should correct inadvertent ‘clerical’ errors made by respondents to avoid manifestly unjust results, even if the errors are discovered subsequent to the deadline for submitting information, and even if the error is not obvious from [the] record at the time.” Id. at 18 (citing NTN Bearing, 74 F.Sd at 1206-09) (emphasis added). Timken also points out that precedent has cautioned Commerce “not to draw distinctions between ‘substantive’ and ‘clerical’ errors in an overly narrow manner.” Id. (citing World Finer Foods, Inc. v. United States, 24 CIT 541, 550 (2000)).

While Timken admits that the relevant statute, legislative history and agency regulations do not directly address the issue of inadvertent errors committed by respondents, Timken argues that each supports the proposition that unintentional errors should be corrected. See id. at 22-26. Timken also raises issue with Commerce’s position that errors of judgment are distinguishable because such a finding is “inconsistent with the caveat articulated in World Finer Foods that ‘[w]here the line is difficult to draw between permissible ministerial or clerical error correction and impermissible factual or methodological changes,’ ” Commerce should classify such error as clerical. Id. at 27 (quoting World Finer Foods, 24 CIT at 550).

Finally, Timken applies the test established by Commerce in Final Results of Antidumping Duty Administrative Reviews of Certain Fresh Cut Flowers From Columbia (“Columbian Flowers”), 61 Fed. Reg. 42,838, 42,834 (Aug. 19, 1996), and cited by Commerce in the Final Results, to the facts of this case and argues that its errors “are analogous to the types of errors Commerce determined to be ‘clerical’ errors and which were corrected in [Co-lumbian] Flowers. ” Timken’s Mem. at 30. Accordingly, Timken maintains that Commerce improperly refused to correct the error identified by Timken in the subject review.

C. Analysis

The antidumping statute requires Commerce to calculate dumping margins as accurately as possible in each administrative review. See Fujian Mach. & Equip. Imp. & Exp. Corp. v. United States, 25 CIT-,-, 178 F.Supp.2d 1305, 1322 (2001) (citing Rhone Poulenc, 899 F.2d at 1191). “[Ajntidumping laws are not punitive in nature, but are designed to remedy the inequities caused by unfair trade practices.” Allied Tube & Conduit Corp. v. United States, 24 CIT 1357, 1370, 127 F.Supp.2d 207, 218 (2000); see NTN, 74 F.3d at 1208 (stating that “the antidumping laws are remedial not punitive” (citing Chaparral Steel Co. v. United States, 901 F.2d 1097, 1103-04 (Fed.Cir.1990))). Although antidumping laws “afford the domestic manufacturer strong protection against dumping,” Commerce is still required to “make a fair and equitable valuation, which may [ultimately] reduce the antidumping margin.” Smith-Corona Group v. United States, 713 F.2d 1568, 1576 (Fed.Cir.1983), cert. denied, 465 U.S. 1022, 104 S.Ct. 1274, 79 L.Ed.2d 679 (1984). These two competing purposes seem to conflict with each other. See American Permac, Inc. v. United States, 12 CIT 1134, 1137, 703 F.Supp. 97, 100 (1988). On the one hand, Commerce is commissioned to protect the domestic industry from unfair trade practices and on the other, Commerce is responsible for promoting free trade. See id. In application, however, the “two purposes of the statute complement, rather than conflict with each other.” Id.

When applying these notions to the issue at bar, the Court recognizes that Commerce often establishes policies to ensure the consistent procedural application of an- *1276 tidumping laws. See Allied Tube, 24 CIT at 1370, 127 F.Supp.2d at 218-19 (stating that “[flair and equitable margins are calculated when the administering authorities are consistent in their procedural application of the law”). In the past, Commerce corrected a respondent’s own clerical errors only if Commerce “could assess from information already on the record that an error ha[d] been made, that the error [was] obvious from the record, and that the correction [was] accurate.” Def.’s Mem. at 26. As a result of the CAFC’s holding in NTN, 5 however, Commerce reevaluated its policy for correcting clerical errors of respondents and developed the six-part Columbian Flowers Test. 6 In its Issues & Decision Mem., Commerce explained that Timken failed to satisfy the first and second criteria of this test. See Timken’s App. Tab 10 at 50-52.

The first prong of the Columbian Flowers Test states that Commerce accepts respondent’s corrections if the error in question is demonstrated to be clerical. Clerical errors have been defined as mistakes “made by a clerk or other subordinate, upon whom devolved no duty to exercise judgment, in writing or copying the figures or in exercising his intention.” PPG Indus., Inc. v. United States, 7 CIT 118, 124 (1984) (citations omitted). Inad-vertencies, on the other hand, have been described as “an oversight or involuntary accident, or the result of inattention or carelessness.” Id. (citing C.J. Tower & Sons of Buffalo, Inc. v. United States, 68 Oust. Ct. 17, 22, 336 F.Supp. 1395, 1399 (1972), aff'd, 61 C.C.P.A. 90, 499 F.2d 1277, C.A.D. 1129 (1974)).

In its supporting brief, Timken argues that Commerce has applied a broader definition of “clerical errors” in past reviews. In fact, Commerce applied the Columbian Flowers Test and accepted the inclusion of a non-subject merchandise sale in the dumping margin as a clerical error in its Issues and Decision Memorandum for the Final Results of the Antidumping Duty Administrative Review and Final Determination Not to Revoke in Part for Canned Pineapple Fruit from Thailand, *1277 2000 WL 1880665 at Cmt. 6 (Dec. 18, 2000). Commerce also found an error in coding the date of sale for one quarter of a customer’s contracts to be clerical in Final Results of Antidumping Duty Administrative Review of Certain Cold-rolled Carbon Steel Flat Products from the Netherlands, 64 Fed.Reg. 11,825 (Mar. 10, 1999). However, in neither of these cases, nor in other holdings referenced by Timken, see Timken’s Mem. at 31, does the respondent commit the same error in two prior reviews. In the present case, Timken admits that it misclassified home market sales of the same subject merchandise in the eighth and ninth administrative reviews, but argues the misclassifications did not result in any meaningful changes to the dumping margins. See supra text accompanying note 2. Nonetheless, this Court agrees with Commerce’s determination that the error at issue was not clerical. Moreover, the Court does not accept Timken’s interpretation of the law that Commerce should correct the error simply because it was inadvertent, see Timken’s Reply Br. at 2, since such a finding would broaden the holding of NTN.

Commerce proffers a harsh interpretation of the facts of this case. Commerce states that

in preparing the data for channels of distribution, as in prior reviews, [Timken] relied on the customer names to classify its home market sales according to distribution channel. Thereby, [Timken] exercised judgment, based upon its examination of its own documents and procedures, that sales to particular customers properly belonged in Channel 1 as sales to “large” OEMs, because of the various distribution and sales activities that were incurred with respect to the sales to these “large” OEMs. [Timken’s] coding of the sales at issue as sales to “large OEMs,” thus, was neither a ministerial nor clerical error, but a deliberate and intentional decision, based upon its actions in prior reviews.

Def.’s Mem. at 28-29 (emphasis in original). Commerce compares the facts of this case with those in Hambro Auto. Corp. v. United States, 66 C.C.P.A. 113, 117-20, 603 F.2d 850, 853-55, C.A.D. 1231 (1979), and argues that the Court should render the misclassifications an error in judgment or a mistake of law. See Def.’s Mem. at 29. Commerce’s reliance on Hombro is misplaced.

Hombro concerned an importer’s challenge to a denial of protests filed after the government refused the importer’s request for reliquidation of certain entries. See Hambro, 66 C.C.P.A. at 117, 603 F.2d at 850. The alleged errors involved statutory values that were calculated by subtracting (rather than adding) cost of export divisions from home market cost figures. See Hambro, 66 C.C.P.A. at 188-19, 603 F.2d at 854. In Hombro, the “errors” were deemed a mistake of law because the importer was fully aware of its general expenses and profits, but believed the legal consequences of such values to be different than they were. See Hambro, 66 C.C.P.A. at 117-20, 603 F.2d at 853-55. Unlike the importer in Hombro, Timken did not realize that it misclassified certain home market sales, nor was Timken cognizant of the legal consequences of its error until the dumping margins were calculated in the Preliminary Results. See Timken’s Mem. at 17-36.

A complete review of the confidential material of this case reveals a situation where rigid compliance with the Colum-bian Flowers Test would render a grossly erroneous dumping margin. Timken’s misclassified transactions covered only a minuscule percentage of all units sold in the LOT. See Timken’s Mem. at 7-10 (proprietary version). The remaining sales in the LOT were disregarded in Commerce’s *1278 profit calculation because they were to be sold at prices below the COP. See id. Ultimately, - Commerce calculated an extremely high CV profit rate and a dumping margin of 61.60 percent based only upon a few misclassified sales. See Timken’s App. Tabs 2, 4, 7 (proprietary version). Timken contends that it “had no knowledge or reason to believe” that its reliance on customer names would result in any misclassi-fications. Timken’s Mem. at 6. Commerce argues that since Timken used the same reporting method in the eighth and ninth reviews, Timken’s action in the tenth review was indeed calculated. See Def.’s Mem. at 28-29. Since the classification of Timken’s home market sales was never an issue addressed in the eighth and ninth administrative reviews, this Court cannot reach the conclusion that Timken intentionally misclassified transactions to attain a desired dumping margin.

The Court, alternatively, must balance the interest of protecting Commerce’s authority to create and implement policy to protect the domestic industry from unfair trade practices, and its correlating obligation “to calculate the most accurate dumping margins possible.” Shandong Huarong Gen. Corp. v. United States, 25 CIT -, -, 159 F.Supp.2d 714, 727 (2001), aff'd, 60 Fed.Appx. 797, 2003 U.SApp. LEXIS 466 (Fed.Cir. Jan. 10, 2003), reh’g denied en banc, 60 Fed.Appx. 797, 798, 2003 U.S.App. LEXIS 6759, *1 (Fed.Cir. Mar. 18, 2003). Commerce directed Timken to provide a description of its channels of distribution and sales process and clarified that the information is “intended to provide [Commerce] with the information necessary to make appropriate comparisons of sales at the same level of trade or to adjust [NV,] if appropriate, when sales are compared at different levels of trade.” Timken’s App. Tab 10 at 50. Commerce also cautioned Timken that the information was of “critical importance.” These instructions are intended to solicit accurate information that Commerce must utilize to calculate the antidumping margins. Cf. Acciai Speciali Terni, S.P.A v. United States, 25 CIT -, -, 142 F.Supp.2d 969, 982 (2001) (stating that “[i]t is respondent’s obligation to supply Commerce with accurate information” (citations omitted)). Had Timken followed these instructions, perhaps this case would not be before this Court. The Court cautions Timken to pay closer attention to the manner in which it classifies and reports future home market sales transactions. Commerce, with its limited resources, cannot be expected “to serve as a surrogate to guarantee the correctness of submissions.” Id. (quoting Yamaha Motor Co. v. United States, 19 CIT 1349, 1359, 910 F.Supp. 679, 687 (1995)). Although the Court recognizes that the burden falls on a respondent to provide accurate information, it is unreasonable to believe that a respondent can do so in each and every review without committing occasional errors. See Shandong, 25 CIT at -, 159 F.Supp.2d at 727 (holding that a restriction of Commerce’s power to correct ministerial errors would undermine its “underlying obligation to calculate the most accurate dumping margins possible”).

Upon publication of the estimated dumping margin in the Preliminary Results, Timken “reexamined its submission and discovered the inadvertent channel of distribution classification errors (regarding] home market LOT 1).” Timken’s Mem. at 12. Timken thereafter submitted, documentation clarifying the correct channel of distribution classifications. See id. Commerce contends that such documentation is not reliable and, therefore, Timken’s clerical error claims also fail to satisfy the second prong of the Columbian Flowers Test. Commerce determined that Timken’s purchase orders, invoices, and notes (some handwritten) were unreliable because “certain record evidence conflicted] with” the *1279 supplemental information. Timken’s App. Tab 10 at 52. In its supporting brief, Commerce states:

The conflict, of course, existed between [Timken’s] original description of distribution channel 1 ... which included prototype and sample sales ... and [Timken’s] request re-categorization of the prototype and sample sales at issue as sales to “other” OEMs or sales to “distributors.” There was also conflict between Commerce’s conclusions in its verification report and its preliminary sales analysis memorandum that [Timken’s] representations of its distribution channels, including channel 1, were consistent with its response and that [Timken] had reported the customer category and channel of distribution fields accurately in its sales databases. Commerce simply found “no information on the record that specifically precludes the transactions in question from being categorized as sales to large OEMs.”

Def.’s Mem. at 37 (quoting Timken’s App. Tab 10 at 52). The Court, however, cannot discern what record evidence Commerce is referring to. In this case, had Timken properly classified the transactions at issue in its response to Commerce’s questionnaire, it would have categorized the sales as distribution channel 2 or 3 because the customers did not buy the units for use in producing large original equipment. See Timken’s App. Tabs 7 & 15 (proprietary version). Thus, any information submitted by Timken to correct the mis-classifications would conflict with the original data supplied by Timken. In the interest of implementing the overarching principle of the antidumping statute, that is, to determine dumping margins as accurately as possible, see Fujian, 25 CIT at -, 178 F.Supp.2d at 1322, the Court remands this case to Commerce to further investigate the claims raised during the administrative proceeding. Any other finding would render a punitive result in contravention to Allied Tube, 24 CIT at 1370, 127 F.Supp.2d at 218.

CONCLUSION

The error committed by Timken is not clerical. This case could have been avoided had Timken followed the instructions of Commerce and classified its sales transactions properly. Nonetheless, the Court must consider the overarching principle of the antidumping statute in rendering a decision on this issue. It is undisputed that Commerce’s goal in administrative reviews is to determine antidumping margins “as accurately as possible.” Rhone Pou-lenc, 899 F.2d at 1191. In this case, an erroneous dumping margin was calculated as a result of a few misclassified transactions reported by Timken. These errors were identified upon publication of the Preliminary Results, and Commerce was provided with supporting documentation. Since the Court finds the facts of this case to be distinct from prior case law, the Court remands this case to Commerce for further investigation and to make any corrections necessary to attain the most accurate antidumping margin.

Notes

1

. This action was originally brought by The Torrington Company and Torrington Nadel-lager GmbH in September 2000. See Summons ¶ 1. The Torrington Company was acquired by the Timken Company on February 18, 2003, and is now known as Timken U.S. Corporation. Timken’s German affiliate is now known as Timken Nadellager, GmbH ("TNG”). See Disclosure of Corporate Affiliations & Fin. Interest at 1 (filed with this Court on Feb. 3, 2004). Timken appeared as a respondent in the subject reviews before *1272 Commerce and appears before this Court in the same capacity.

2

. This action challenges the final results of the tenth administrative review of antifriction bearings from Germany, which cover CRBs produced by TNG. Cylindrical roller bearings produced by TNG were also subject to the eighth and ninth administrative reviews in which Commerce determined de minimis dumping margins for the subject product. See Pinal Results of Antidumping Duty Administrative Reviews on Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Sweden, and the United Kingdom, 64 Fed.Reg. 35,590, 35,591 (July 1, 1999); Final Results of Antidumping Duty Administrative Reviews on Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, Germany, Italy, Japan, Romania, Singapore, Sweden, and the United Kingdom, 63 Fed.Reg. 33,320, 33,321 (June 18, 1998) (collectively "eighth and ninth administrative reviews”).

3

. Since the administrative review at issue was initiated after December 31, 1994, the applicable law is the antidumping statute as amended by the Uruguay Round Agreements Act ("URAA”), Pub.L. No. 103-465, 108 Stat. 4809 (1994) (effective January 1, 1995). See Torrington Co. v. United States, 68 F.3d 1347, 1352 (Fed.Cir.1995) (citing URAA § 291(a)(2), (b) (noting effective date of URAA amendments)).

4

. Under post-URAA law, pursuant to 19 U.S.C. §§ 1677b(a)(l) and 1677(16) (1994), Commerce must first look to identical merchandise in matching the United States model to the comparable home market model. If a determination cannot be satisfactorily made using identical merchandise, Commerce must look to like merchandise-initially under the second category and, if that is not available, under the third category.

5

. NTN involved the inadvertent use of a code for high precision bearing tolerances rather than the standard precision bearing tolerances and the listing of four sales to foreign customers as sales to domestic customers. See NTN, 74 F.3d at 1205. Commerce rejected the respondent’s request for correction upon a finding that "the errors were not obvious from the record and that the deadline for submitting new information had expired.” Id. The CAFC held, however, that the "requirement that the record disclose the error essentially preclude[d] correction of clerical errors made by a respondent. Where clerical personnel of a respondent transpose code numbers, the existing administrative record will not disclose that such error occurred.” Id. at 1208. The CAFC also held that "while it may be a reasonable exercise of [an agency] to restrict the correction of its own clerical errors to those obvious from the record, the same rule applied to a respondent's errors becomes arbitrary." Id.

6

. In Columbian Flowers, Commerce stated that it would

accept corrections of clerical errors under the following conditions: (l)[t]he error in question must be demonstrated to be a clerical error, not a methodological error, an error in judgment, or a substantive error; (2) [Commerce] must be satisfied that the corrective documentation provided in support of the clerical error allegation is reliable; (3) the respondent must have availed itself of the earliest reasonable opportunity to correct the error; (4) the clerical error allegation, and any corrective documentation, must be submitted to [Commerce] no later than the due date for the respondent's administrative case brief; (5) the clerical error must not entail a substantial revision of tire response; and (6) the respondent's corrective documentation must not contradict information previously determined to be accurate at verification.

Columbian Flowers ("Columbian Flowers Test"), 61 Fed.Reg. at 42,834.