Memorandum Opinion and Order
DiCarlo, Chief Judge:
National Steel Corporation, Armco Steel Company, L.E, Bethlehem Steel Corporation, Gulf States Steel, Inc. of Alabama, Inland Steel Industries, Inc., LTV Steel Company, Inc., Sharon Steel Corporation, U.S. Steel Group A Unit of USX Corporation, and WCI Steel, Inc. (Domestic Producers) and Hoogovens Groep B.V and N.VW (U.S.A.), Inc. (Hoogovens) challenge the final determination of the United States Department of Commerce concerning certain cold-rolled flat steel products from the Netherlands. Final Determinations of Sales at Less Than Fair Value: Certain Hot-Rolled Carbon Steel Flat Products and Certain Cold-Rolled Carbon Steel Flat Products From the Netherlands, 58 Fed. Reg. 37,199 (Dep’t Comm. 1993), amended by Antidumping Duty Order and Amendments to Final Determinations of Sales at Less Than Fair Value: Certain Hot-Rolled Carbon Steel Flat Products and Certain Cold-Rolled Carbon Steel Flat Products From the Netherlands, 58 Fed. Reg. 44,172 (Dep’t Comm. 1993). Plaintiffs and defendant-intervenors cross-move for Judgment Upon an Agency Record pursuant to USCIT R. 56.2. The court has jurisdiction pursuant to 19 U.S.C. § 1516a(a)(2) (1988) and 28 U.S.C. § 1581(c) (1988).
Background
Domestic Producers filed a petition requesting imposition of anti-dumping duties on carbon steel flat products from various countries. [1127]*1127The investigated products included cold-rolled carbon steel flat products from the Netherlands, at issue in this case. Certain Hot-Rolled, Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, Certain Corrosion-Resistant Carbon Steel Flat Products and Certain Cut-to-Length Carbon Steel Plate From Various Countries, 57 Fed. Reg. 33,488 (Dep’t Comm. 1992).
On August 19,1992, Commerce presented Hoogovens Groep B.V, the primary Dutch exporter of the subject merchandise, with its antidump-ing questionnaire. Hoogovens responded on September 10, 1992 and October 21, 1992. The responses of Hoogovens’ affiliates, Rafferty-Brown Steel Co., Inc. of Connecticut (RBC) and Rafferty-Brown Steel Co., Inc. of North Carolina, were filed on November 4,1992.
Commerce established December 21,1992 as the deadline for remedying deficiencies in the questionnaire responses, in order “to allow case analysts and other interested parties * * * time to review the submissions prior to verification.” Final Determ, 58 Fed. Reg. at 37,202. On that date, at the invitation of the Department, Hoogovens submitted a computer tape and narrative explanation correcting certain errors in its previous submissions. The Department published its preliminary determination on February 4, 1993. Certain Hot-Rolled Carbon Steel Flat Products and Certain Cold-Rolled Carbon Steel Flat Products From the Netherlands, 58 Fed. Reg. 7113 (Dep’t Comm. 1993) (prelim, determ.).
On February 10, 1993, in a letter to Hoogovens, the Department offered the foreign producers “an opportunity to clarify” information that had previously been submitted. (Pub. Doc. 109, Letter from Dep’t. ofCommerce to Powell, Goldstein, of Feb. 10,1993, at 1.) Commerce specifically instructed Hoogovens that it would not accept any new information after February 17,1993.
On February 22,1993, the first day of verification, Hoogovens notified Commerce that it had discovered a number of inaccuracies affecting a small number of the product characteristics submitted to Commerce for home market sales, regarding thickness groupings, product type, and tolerance. At verification, Hoogovens also disclosed its discovery of misreporting on a small number of U.S. sales with regard to the quality characteristic.
On April 28, 1993, Commerce asked Hoogovens to submit revised computerized sales listings with corrected product characteristic data. Hoogovens provided the data on May 10, 1993. Commerce then asked Hoogovens to resubmit the computerized data in slightly different formats, which it did on May 18, 1993. Due to time constraints on Commerce from various concurrent steel investigations, the Department decided not to accept the submitted corrections. Consequently, Commerce resorted to best information available (BIA) for the product characteristic data. In determining the appropriate BIA, Commerce concluded that the errors were limited in nature. Therefore, it used the weighted-average of the calculated positive dumping margins for each class of merchandise as BIA.
[1128]*1128The Department also conducted verification of Hoogovens’ exporter sales price (ESP) data at Hoogovens’ affiliate, RBC, on April 7-8,1993. During that verification, Hoogovens attempted to submit new factual information regarding omitted ESP transactions.
The number of omitted transactions was substantial. (Pub. Doc. 142, Conf. Doc. 52, RBC Verification Report of Apr. 27,1993 at 2). Commerce had advised RBC that the new information could not be accepted because it was submitted after the deadline. RBC attempted to resubmit the same information on April 9, 1993. Commerce did not accept the submission, and applied the highest non-aberrant margin as BIA.
In its final determination, Commerce made an adjustment to the United States Price to account for the Dutch value-added tax and eliminate the false dumping margin created by the tax. Value-added taxes (VAT) are assessed only on sales within the Netherlands, but not on products exported to the United States.
In making the adjustment, the Department increased the United States Price by the absolute amount of tax actually incurred on sales of comparison Dutch merchandise. The Domestic Producers contest this as contrary to the methodology required by 19 U.S.C. § 1677a(d)(l)(C) (1988). Rather, the Domestic Producers allege that, under the statute, Commerce should have multiplied the price of the merchandise exported to the United States by the foreign market tax rate and added this amount to the United States Price.
Hoogovens now contests Commerce’s (1) resort to BIA for the unreported ESP transactions; (2) selection of the highest non-aberrant margin as BIA; and (3) application of that margin. The Domestic Producers challenge Commerce’s (1) characterization of Hoogovens’ misreported home market sales data as limited in nature and resultant application of the weighted average of the positive calculated dumping margins as BIA; and (2) calculation of the adjustment to U.S. price for tax forgiven upon exportation of the steel products to the United States, and failure to measure the amount of the tax passed through to home market customers.1
Discussion
This court shall uphold Commerce’s final determination in an anti-dumping duty investigation unless that determination is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B). Substantial evidence is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Universal Camera Corp. v. NLRB, 340 U.S. 474, 477 (1951) (quoting Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 (1938)).
[1129]*11291. Unreported ESP Transactions:
a. Application of BIA:
Respondent Hoogovens contests Commerce’s resort to BIA for the omitted data for RBC’s ESP transactions. According to Hoogovens, the omissions can not be categorized as unreported transactions. Instead, due to inadvertent computer errors, multiple invoices for single coils of metal were treated as a single transaction. Therefore, Hoogovens argues, “the missing data did not relate to shipments pursuant to unreported sales orders or to previously-unreported customers or through previously-unreported sales channels.” (Resp’t Conf. Mem. Supp. J. Agcy. R. at 15.)
Hoogovens also contends that the omissions were limited in nature, noting that the RBC sales represented only [ ] percent of Hoogovens’ total U.S. sales. (Resp’t Conf. Mem. Supp. J. Agcy. R. at 14.) Therefore, Hoogovens argues, the additional administrative burden placed on Commerce or unfairness to petitioners in accepting and verifying the data would have been minimal.
Finally, Hoogovens alleges that Commerce invited Hoogovens to submit new information in its letter of February 12, 1993. According to Hoogovens, Commerce’s request (authorizing Hoogovens to bring minor inadvertent errors to its attention at verification) “must be interpreted as indicating a willingness on the Department’s part to accept minor changes * * * and as extending the Department’s regulations on the timeliness of questionnaire responses * * (Resp’t Conf. Mem. Supp. J. Agcy. R. at 17.)
This court does not agree. Hoogovens failed to meet clear deadlines for submitting new information. Once such deadlines have passed, whether Commerce accepts the late submissions is within its discretion. See 19 C.F.R. § 353.31(b) (1994).
Section 353.31(a), title 19, Code of Federal Regulations specifically provides that submissions of factual information for consideration are to be submitted no later than seven days before commencement of verification. 19 C.F.R. § 353.31(a) (1994). Commerce warned Hoogovens that it would not accept any new data after February 17,1994. Nonetheless, Hoogovens attempted to submit data after verification had begun.
The purpose behind permitting Commerce to resort to BIA is to induce respondents to provide Commerce with requested information in a timely, complete, and accurate manner so that the Department may determine current margins within statutory deadlines. Rhone Poulenc, Inc. v. United States, 8 Fed. Cir. (T) 61, 67-68, 899 F.2d 1185, 1191 (1990); see also 19 U.S.C. § 1677e (1988); 19 U.S.C. § 1673d(a)(1) (1988). BIA, therefore, may be viewed as “an investigative tool, which that agency may wield as an informal club over recalcitrant parties or persons whose failure to cooperate may work against their best interest.” Atlantic Sugar, Ltd. v. United States, 2 Fed. Cir. (T) 130, 134, 744 F.2d 1556, 1560 (1984). This authority is especially important in light of the [1130]*1130agency’slack of subpoena power. Rhone Poulenc, 8 Fed. Cir. (T) at 67-68, 899 F.2d at 1191.
Moreover, although the omitted information made up a small portion of Hoogovens’ U.S. sales, it was significant in terms of its proportion of Hoogovens’ U.S. affiliate’s sales. Calculation of ESP dumping margins is more time-consuming and complicated than for purchase price transactions. Commerce did not have the time or resources to incorporate the belated submissions into its calculations. Nor was Commerce required to do so. 19U.S.C. § 1677e.Asthis court has stated, “[i]tisfortheITAto conduct its antidumping investigations the way it sees fit, not the way an interested party seeks to have it conducted.” N.A.R., S.p.A. v. United States, 14 CIT 409, 415, 741 F. Supp. 936, 941 (1990); see also Olympic Adhesives, Inc. v. United States, 8 Fed. Cir. (T) 69, 76, 899 F.2d 1565, 1571 (1990).
Finally, Hoogovens was on notice as to the effect of the late submissions. Commerce concedes that it had granted Hoogovens the opportunity to clarify previous submissions. (Gov’t Conf. Mem. at 18.) Nonetheless, this was not an invitation to submit entirely new factual data. As the February 12th letter provided, in pertinent part, “ [p] lease note that verification is not intended to be an opportunity for submission of new factual information.” (Pub. Doc. 113, Letter from Gary Tav-erman, Director, Office of Antidumping Investigations, to counsel for Hoogovens of Feb. 12,1994, at 2.) Therefore, Commerce acted properly in resorting to BIA.
b. Selection of the Highest Non-Aberrant Margin:
Hoogovens also contests Commerce’s selection of the highest non-aberrant dumping margin as BIA for the unreported ESP transactions. Hoogovens contends that, because it was “cooperative,” Commerce was unreasonable in applying a punitive dumping margin. Hoogovens notes that its failure to report was due to computer error, rather than an intentional act or gross negligence, and that once the errors were discovered, Hoogovens promptly informed Commerce.
Both the statute, see 19 U.S.C. § 1677e, and the implementing regulation, see 19 C.F.R. 353.37 (1994), establishes mandates as to when Commerce must resort to BIA for a respondent’s failure to submit timely information. However, little indication is provided as to what constitutes the appropriate BIA. Krupp Stahl A.G. v. United States, 17 CIT 450, 453, 822 F. Supp. 789, 792 (1993).
As a general rule, the Supreme Court has noted that an agency must either conform itself to its prior decisions or explain the reasons for its departure. See Secretary of Agriculture v. United States, 347 U.S. 645, 652-53 (1954). However, it appears the Federal Circuit has given Commerce broad discretion to change its methodology without explanation. Allied-Signal Aerospace Co. v. United States, 12 Fed. Cir. (T)_,_, 28 F.3d 1188, 1191 (1994). The Court need not address this question, as [1131]*1131it finds selection of the highest non-aberrant dumping margin as BIA consistent with Commerce’s prior practice as denoted below.
The application of BIA can be “total” or “partial.” Commerce applies total BIA when a respondent has failed to submit information in a timely manner, or when part of the submitted data is sufficiently flawed, so that the response as a whole is rendered unusable. See Rhone Poulenc, Inc. v. United States, 13 CIT 218, 224, 710 F. Supp. 341, 346 (1989), aff’d 8 Fed. Cir. (T) 61, 899 F.2d 1185 (1990), Persico Pizzamiglio, S.A. v. United States, 18 CIT 299, Slip Op. 94-61 at 17-19 (Apr. 14, 1994); Chinsung Indus. Co. v. United States, 13 CIT 103, 105-06, 705 F. Supp. 598, 600-01 (1989). For total BIA, the margins used in determining the overall rate are derived from sources other than the data submitted by the respondent in the final investigation.
When Commerce resorts to total BIA, it has adopted a two-tiered methodology for determining the rate to be applied. Allied-Signal Aerospace Co. v. United States, 11 Fed. Cir. (T)_ 996 F.2d 1185, 1190-91 (1993); Certain Hot-Rolled Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, Certain Corrosion-Resistant Carbon Steel Flat Products and Certain Cut-to-Length Carbon Steel Plate From Germany, 58 Fed. Reg. 37,136,37,137 (Dep’t Comm. 1993). The degree of cooperation determines which tier is used. The Department assigns lower margins to those respondents who cooperate in the investigation; more adverse margins are assigned to those who are uncooperative. See e.g., Certain Hot-Rolled Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, Certain Corrosion-Resistant Carbon Steel Flat Products, and Certain Cut-to-Length Carbon Steel Plate From France, 58 Fed. Reg. 37,125, 37,129 (Dep’t Comm. 1993); Certain Cold-Rolled Carbon Steel Flat Products From Argentina, 58 Fed. Reg. 7066,7070 (Dep’t Comm. 1993) (prelim, determ.); Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof from the Federal Republic of Germany, 54 Fed. Reg. 18,992, 19,033 (Dep’t Comm. 1989).
Commerce applies partial BIA when only part of the submitted information is deficient. The adversity of the information used as partial BIA depends on the level of sufficiency of the information provided. It is noteworthy that Commerce does not consider the respondent’s level of cooperation when applying partial BIA. Steel Flat Products From France, 58 Fed. Reg. at 37,129; see also Certain Helical Spring Lock Washers From the People’s Republic of China, 58 Fed. Reg. 48,833, 48,839 (Dep’t Comm. 1993) (applying highest non-aberrant margin despite respondent’s cooperation and identification of omissions).
When errors in the information submitted constitutes a failure to provide the necessary data, Commerce applies a more adverse dumping margin as partial BIA. Certain Cut-to-Length Carbon Steel Plate From Finland, 58 Fed. Reg. 37,122, 37,124 (Dep’t Comm. 1993) (applying adverse BIA standards when errors are “systemic in nature”); see also Certain Stainless Steel Wire Rods From France, 58 Fed. Reg. 68,865, [1132]*113268,869 (Dep’t Comm. 1993) (applying “adverse” partial BIA where omitted sales information from limited period constituted “significant portion” of total U.S. sales). This margin constitutes “the higher of: (a) [t]he highest non-aberr[ant] transaction margin calculated for that firm from among the sales of the same class or kind of merchandise where we were able to calculate a margin or (b) the average petition rate for the same class or kind of merchandise from the same country of origin.” Carbon Steel Flat Products From Argentina, 58 Fed. Reg. at 7070.
When only a minor adjustment in the data is involved or when there is an inadvertent gap in the record, Commerce applies a less adverse BIA or a neutral surrogate. Ad Hoc Comm. of AZ-NM-TX-FL Producers of Gray Portland Cement v. United States, 18 CIT 906, 915, Slip Op. 94-151 at 20 n. 22 (Sept. 26, 1994).
Commerce’s actions in applying the highest non-aberrant dumping margin were supported by substantial evidence. Although Hoogovens was cooperative, the quality and completeness of the data were the critical characteristics for determining whether the misreporting rose to such a level as to constitute a failure to provide any data at all. Commerce applied a lower partial BIA when Hoogovens submitted minimally-flawed data for the purchase price transactions. For the ESP transactions, however, Hoogovens failed to supply sufficient data in a timely manner to calculate a dumping margin. Therefore, selection of the highest non-aberrant margin as BIA was proper.
c. Application of the Highest Non-Aberrant Margin:
Hoogovens challenges Commerce’s application of the highest non-aberrant margin in this case and argues that the selected rate of 157.51 percent lacks a nexus with the data from which it is taken. The court finds further explanation as to the basis for Commerce’s determination is needed.
Section 1677e allows Commerce to presume that the highest margin may constitute the best information available of the current margins. 19U.S.C. § 1677e. Although the ultimate purpose ofBIA is not to punish, BIA is intended to be adverse, see Pulton Chain Co. v. United States, 17 CIT 1136, 1139, Slip Op. 93-202 at 7 (October 18, 1993), and requires the use of adverse assumptions. See Rhone Poulenc, 8 Fed. Cir. (T) at 67-68, 899 F.2d at 1190-91.
Nonetheless, Commerce must be reasonable in its application of its chosen methodology. Commerce’s actions may become unreasonable in nature if “the agency * * * [has] * * * rejected] low margin information in favor of high margin information that was demonstrably less probative of current conditions.” Rhone Poulenc, 8 Fed. Cir. (T) at 67, 899 F.2d at 1190 (emphasis added). A rational relationship must exist between the “data chosen and the matter to which they are to apply.” Manifattura Emmepi S.p.A. v. United States, 16 CIT 619, 624, 799 F. Supp. 110, 115 (1992)
[1133]*1133Commerce contends that it chose from margins on individual sales as high as [ ] percent in selecting the 157.51 percent rate. (Gov’t Conf. Mem. at 28.) According to Commerce, it “gave Hoogovens the benefit of the doubt * * * that the highest sales were atypical,” and selected a margin much lower than the highest figure calculated. Id. at 30. Commerce used this percentage to establish an overall dumping margin of 20.19 percent.
Commerce’s arguments are fallacious. That Commerce could have chosen a higher rate does not necessarily mean that the rate selected was not “aberrant. ” Further, Commerce has not provided the court with any explanation of the term “non-aberrant margin,” nor a method for distinguishing a non-aberrant margin from an aberrant margin.
Webster’s Dictionary defines “aberrant” as “deviating from the usual or natural type” or “abnormal.” Webster’s Third New International Dictionary 3 (1965). Under this definition, Commerce’s selected margin could be considered aberrant. A significant portion of Hoogovens sales had margins well below the selected rate of 157.51 percent. Only 17 transactions out of “[ ] sales reported by Hoogovens had margins equal to or greater than” the selected percentage. (Resp’t Conf. Mem. Supp. J. Agcy. R. at 30.) These sales accounted for only a minuscule percentage by volume of Hoogovens’ reported U.S. sales.
The court has found no evidence in the record showing the particular margins selected by Commerce bore a rational relationship to Hoogo-vens’ sales. Further explanation as to the basis for Commerce’s determination is needed. Therefore, this court remands the issue to Commerce for redetermination. On remand, commerce is instructed to articulate standards for determining when a rate is “aberrant,” and select a BIA margin indicative of Hoogovens’ sales.
2. Misreported Characteristics:
Domestic Producers allege Commerce improperly determined that Hoogovens’ misreporting of its home-market sales was limited. Rather, the Domestic Producers contend the misreporting was substantial and that Commerce should have utilized the highest-non-aberrant margin as BIA in accordance with prior practice, as set forth in Carbon Steel Flat Products From Argentina, 58 Fed. Reg. at 7070.
Although Hoogovens’ misreporting affected a significant percentage of margin calculations, the errors themselves were minor. The misreporting affected product characteristics used by Commerce in the model-matching hierarchy. The affected characteristics, arranged in order of importance, are “type,” “quality,” “thickness group,” and “tolerance,” ranked in the second, third, fifth, and eleventh positions, respectively. The majority of the errors involved the miscalculating of thickness by eight millionths of an inch, the result of a-programming error in conversion of millimeters to inches. This error placed a number of sales in the wrong thickness group. The misreported thickness also put some products in the wrong type category. The miscalculation for the type category [1134]*1134was 213 millionths of an inch. The quality category was also affected, due to the misreporting of one sale to one customer. Finally, Hoogovens misreported about 35% or the sales as to the “tolerance” characteristic.
Commerce has discretion to determine whether, under BIA, the discrepancies or misreportings are significant enough to amount to a failure to provide the requested information. Commerce has taken the position that selection of a severely adverse BIA may be improper when only a minor adjustment in the data is involved or there is an inadvertent gap in the record, but otherwise the respondent has substantially complied with the questionnaire. See Ad Hoc Comm., 18 CIT at 915, Slip. Op. 94-151 at 20 n. 22; Asociacion Colombiana de Exportadores de Flores v. United States, 13 CIT 13, 16, 704 F. Supp. 1114, 1117 (1989) aff’d 8 Fed. Cir. (T) 97, 901 F.2d 1089, cert. denied, 498 U.S. 848 (1990), see also Brass Sheet and Strip From West Germany, 55 Fed. Reg. 28,264, 28,265 (Dep’t Comm. 1990) (prelim, results of antidumping duty admin, review); Replacement Parts for Self-Propelled Bituminous Paving Equipment From Canada, 58 Fed. Reg. 15,481, 15,482-83 (Dep’t Comm. 1993) (final results of admin, review of antidumping finding); Porcelainon-Steel Cooking Ware From Mexico, 58 Fed. Reg. 43,327, 43,329 (Dep’t Comm. 1993) (prelim, results of antidumping duty admin, review). On the other hand, Commerce has consistently applied more adverse BIA when errors are significant or systemic in nature or where there is ^pattern of misclassification. Carbon Steel Plate From Finland, 58 Fed. Reg. at 37,124; Certain Hot-Rolled Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, and Certain Corrosion-Resistant Carbon Steel Flat Products From Japan, 58 Fed. Reg. 37,154, 37,161 (Dep’t Comm. 1993); Steel Wire Rods From France, 58 Fed. Reg. at 68,869.
This is not the case here. Hoogovens did not omit data, but only provided inaccurate information, which in most instances was due to a computer conversion error, nor were the errors systematic in nature. Although the errors affected three U.S. product control numbers (CON-NUMUs) accounting for [ ] percent of the U.S. cold-rolled transactions, (Pet’rs Conf. Mem. Supp. J. Agcy. R. at 14), and one home market product control number affecting [ ] percent of U.S. cold-rolled sales, id. at 14-15, Commerce could have reasonably concluded that the errors did not amount to a failure to provide information, in light of how they occurred, and more importantly, in light of their significance.
Finally, Domestic Producers claim that Commerce’s determination that Hoogovens’ misreporting was limited in nature “unfairly rewards” Hoogovens, and undermines the effectiveness and purpose of BIA as an incentive for full disclosure and cooperation. However, Hoogovens fully cooperated in the investigation. Therefore, there was no behavior to deter with the use of highest non-aberrant BIA. There is substantial evidence in the record to support Commerce’s conclusion that the misreporting was limited in nature. Therefore, Commerce did not err in applying the weighted-average margin as BIA.
[1135]*11353. Calculation of Adjustment to U.S. Price for VAT:
The Domestic Producers contest the methodology Commerce applied in adjusting the United States Price for VAT not collected or rebated due to exportation. Commerce has joined in the Domestic Producer’s request for remand. (Gov’t Mem. at 10-11.) Commerce’s decision is based on its acknowledgment that recent decisions of this court have overruled Commerce’s former methodology. See Federal-Mogul Corp. v. United States, 17 CIT 1093, 834 F. Supp. 1391 (1993), appeal docketed, Nos. 94-1097, 94-1194 (Fed. Cir. Aug. 5, 1994); see also Torrington Co. v. United States, 18 CIT 1012, 1016, Slip-Op. 94-167 at 8 (Oct. 20, 1994); Timken Co. v. United States, 18 CIT 942, 944-45, Slip-Op 94-157 at 5-6 (Oct. 7, 1994); Ad Hoc Comm., 18 CIT at 911-12, Slip. Op. 94-151 at 13; Zenith Elecs. Corp. v. United States, 18 CIT 882, 883, Slip Op. 94-148 at 3-5 (Sept. 21, 1994); Independent Radionic Workers v. United States, 18 CIT 851, 852-53 Slip-Op 94-144 at 3-4 (Sept. 16, 1994); Avesta Sheffield, Inc. v. United States, 17 CIT 1212, 1218, 838 F. Supp. 608, 614-15 (1993); Federal-Mogul Corp. v. United States, 17 CIT 88, 96, 813 F. Supp. 856, 863-65 (1993)
Hoogovens contends that the methodology applied by Commerce in this case would also be proper and consistent with case law. According to Hoogovens, the alternative methodology advocated by the Domestic Producers and Commerce creates a distortion in the dumping margin, termed “the multiplier effect,” which unfairly penalizes foreign producers for economic policies outside of their control. (Resp’t Conf. Mem. Opp. J. Agcy. R. at 31.) This result, Hoogovens’ argues, would therefore constitute a violation of Article VI of GATT, which requires signatory states to offset tax differences when calculating dumping duties. (Resp’t Conf. Mem. Opp. J. Agcy. R. at 34-35.)
Hoogovens suggests that two cases, Zenith Elecs. Corp. v. United States, 11 Fed. Cir. (T)_, 988 F.2d 1573 (1993) and Hyster Co. v. United States, 18 CIT 119, 848 F. Supp. 178 (1994), permit elimination of the multiplier effect to achieve tax neutrality. A closer examination of these cases reveals a contrary conclusion.
The Court of Appeals for the Federal Circuit stated in Zenith that tax neutrality was clearly contemplated and rejected by Congress. 11 Fed. Cir. (T) at_, 988 F.2d at 1582. The Court emphasized that increases in dumping margins caused by the multiplier effect only occur where there is pre-existing dumping. Zenith, 11 Fed. Cir. (T) at_, 988 F.2d at 1582. Thus, the multiplier effect does not artificially create new margins, although it may inflate existing ones. Zenith, 11 Fed. Cir. (T) at _, 988 F.2d at 1580-81. This effect is not inconsistent with GATT, which only bars the creation of new duties, see Federal-Mogul, 17 CIT at 98, 813 F. Supp. at 865, nor is it necessarily unfair. Zenith, 11 Fed. Cir. (T) at_, 988 F.2d at 1581. As the Federal Circuit noted, “[i]f a foreign manufacturer does not export its wares at less than fair value, it will not suffer disadvantage from the operation of [19 U.S.C. § 1677a(d) (1) (C)].” Zenith, 11 Fed. Cir. (T) at_, 988 F.2d at 1581-82. Although [1136]*1136Hoogovens argues that footnote 4 of Zenith permits an adjustment for the multiplier effect, adherence to the dicta in this footnote would be inconsistent with Zenith’s basic holding.
Hoogovens also contends that the court in Hyster “agreed that the statute authorizes the methodology suggested in footnote 4 by the Zenith Court.” (Resp’t Conf. Mem. Supp. J. Agcy. R. at 32.) Hoogovens misinterprets Hyster. Although the court in Hyster noted the suggested methodology in footnote 4 oí Zenith, it did not endorse it. Rather, Hyster merely “directed] Commerce to consider any further adjustments to [United States Price] consistent with Zenith and [section 1677a(d)(l)(C)].” 848 F. Supp. at 184. Any attempt by Commerce to eliminate the multiplier effect is inconsistent with both the statute and its legislative history. See Zenith, 11 Fed. Cir. (T) at_, 988 F.2d at 1582.
The provision in question, section 1677a(d)(l)(C), directs Commerce to increase the United States Price by “the amount of any taxes imposed in the country of exportation directly upon the exported merchandise * * * which have not been collected, by reason of the exportation of the merchandise to the United States * * *. 19U.S.C. 1677a(d)(l)(C). Thus, the statute prescribes that proper adjustment for VAT is the amount of taxes that would have been imposed on the exported merchandise if the exports had remained in the home market, rather than the amount of taxes actually imposed on the home-market products. Zenith, 11 Fed. Cir. (T) at_, 988 F.2d at 1580.
This court declines to depart from the mandates of the statute and the decisions of this court and of the Court of Appeals for the Federal Circuit. Therefore, the court remands the tax adjustment to Commerce for recalculation.
Conclusion
In accordance with the foregoing opinion, Commerce’s final determination is affirmed in part and remanded in part. This court denies Domestic Producers’ motion to alter Commerce’s selection of BIA for the misreported sales. It denies Hoogovens’ motion to direct Commerce to accept the untimely ESP data and to set aside Commerce’s selection of the highest non-aberrant dumping margin as BIA for those transactions. The court remands the final determination with directions to Commerce to provide clear criteria to judge the highest non-aberrant dumping margin standard. Commerce is further instructed to recalculate the adjustment to U.S. price for tax forgiven upon exportation of the steel products to the United States, in conformity with the methodology articulated in prior opinions of this court.