67 FR 3143, January 23, 2002 A-427-820 Investigation Public Document IA/I/2/TRK/BCS MEMORANDUM TO: Faryar Shizad Assistant Secretary for Import Administration FROM: Richard W. Moreland Deputy Assistant Secretary for Import Administration SUBJECT: Issues and Decision Memorandum for the Final Determination of the Antidumping Duty Investigation of Stainless Steel Bar from France - October 1, 1999, through September 30, 2000 Summary We have analyzed the comments of the interested parties in the antidumping duty investigation of stainless steel bar from France. As a result of our analysis of these comments, we have made changes in the margin calculations as discussed in the "Margin Calculations" section of this memorandum. We recommend that you approve the positions we have developed in the "Discussion of the Issues" section of this memorandum. Below is the complete list of the issues in this investigation for which we received comments by parties: Issue 1: Whether to Collapse the Sales Prices and Production Costs of two Affiliates Across Countries Issue 2: Defining Foreign Like Product for Making Product Comparisons Issue 3: Ranking "Peeled and Descaled" as a Final Finish Characteristic Issue 4: Assigning Total Facts Available Issue 5: Calculating the Dumping Margin Issue 6: Level of Trade Issue 7: Home Market Expenses Reported In Lieu of Commissions Issue 8: Treatment of Movement and Selling Expenses Between Home Market Affiliates as Manufacturing Costs Issue 9: Home Market Warranty Expenses Issue 10: Treatment of UFS/U-SF's Restructuring Costs as Selling Expenses Issue 11: U.S. Credit Expenses for Consignment Sales Issue 12: Whether to Include Freight Revenue in Calculation Formulas Used to Report Certain U.S. Discounts and Expenses Issue 13: Treatment of the French Tax Provision Background On August 2, 2001, the Department of Commerce ("the Department") published the notice of preliminary determination of sales at less-than- fair value and postponement of final determination of the antidumping duty investigation of stainless steel bar ("SSB") from France. See Stainless Steel Bar from France: Notice of Preliminary Determination of Sales at Less-Than-Fair Value and Postponement of Final Determination, 66 FR 40201 (August 2, 2001) ("Preliminary Determination"). On August 28, 2001, Aubert & Duval, S.A. ("A&D") notified the Department that it was no longer participating in this investigation. Although the deadline for this determination was originally December 17, 2001, in order to accomodate certain verifications that were delayed because of the events of September 11, 2001, the Department tolled the final determination deadline in this and the concurrent SSB investigations until January 15, 2001. For purposes of this investigation, the term "stainless steel bar" includes articles of stainless steel in straight lengths that have been either hot-rolled, forged, turned, cold-drawn, cold-rolled or otherwise cold-finished, or ground, having a uniform solid cross section along their whole length in the shape of circles, segments of circles, ovals, rectangles (including squares), triangles, hexagons, octagons, or other convex polygons. Stainless steel bar includes cold-finished stainless steel bars that are turned or ground in straight lengths, whether produced from hot-rolled bar or from straightened and cut rod or wire, and reinforcing bars that have indentations, ribs, grooves, or other deformations produced during the rolling process. Except as specified above, the term does not include stainless steel semi-finished products, cut length flat-rolled products (i.e., cut length rolled products which if less than 4.75 mm in thickness have a width measuring at least 10 times the thickness, or if 4.75 mm or more in thickness having a width which exceeds 150 mm and measures at least twice the thickness), products that have been cut from stainless steel sheet, strip or plate, wire (i.e., cold- formed products in coils, of any uniform solid cross section along their whole length, which do not conform to the definition of flat-rolled products), and angles, shapes and sections. The stainless steel bar subject to this investigation is currently classifiable under subheadings 7222.11.00.05, 7222.11.00.50, 7222.19.00.05, 7222.19.00.50, 7222.20.00.05, 7222.20.00.45, 7222.20.00.75, and 7222.30.00.00 of the Harmonized Tariff Schedule of the United States ("HTSUS"). Although the HTSUS subheadings are provided for convenience and customs purposes, the written description of the scope of this investigation is dispositive. The period of investigation ("POI") is October 1, 1999, through September 30, 2000. We invited parties to comment on our preliminary determination. The petitioners and the respondent, Ugine-Savoie Imphy, S.A. ("U-SI"), submitted their case briefs on November 19, 2001. On November 27, 2001, the parties submitted their rebuttal briefs. A hearing was held on December 6, 2001. Margin Calculations We calculated constructed export price ("CEP") and normal value ("NV") using the same methodology stated in the preliminary determination, except as follows: We determined that although there are but two levels of trade ("LOT") in the home market, each of those LOTs were different from the U.S. LOT. Therefore, we granted a CEP offset (see Comment 6). We treated expenses incurred by a home market affiliate, which acted as a commission agent, as indirect selling expenses rather than direct selling expenses (see Comment 7). We recalculated home market warranty expenses on a customer-specific basis rather than on a general product basis and treated both the respondent and its affiliate's home market warranty expenses as direct selling expenses (see Comment 9). We treated the U.S. technical service expenses as an indirect rather than as a direct selling expense. Discussion of the Issues Comment 1: Whether to Collapse the Sales Prices and Production Costs of two Affiliates Across Countries The petitioners request that the Department collapse U-SI (the respondent in the SSB from France investigation and Trafilerie Bedini, Srl ("Bedini") (a respondent in the SSB from Italy investigation). The petitioners argue that the Department should collapse U-SI and Bedini into a single entity because they are affiliates and they satisfy the collapsing criteria in section 351.401(f) of the Department's regulations. The petitioners contend that U-SI and Bedini are significant SSB producers and would not have to "substantially retool" their manufacturing facilities to shift production. In addition, the petitioners allege that both U-SI and Bedini can produce SSB in significant volumes and have the ability to manipulate prices or production in an effort to evade the assessment of antidumping duties. Furthermore, the petitioners assert that since it is common for subject merchandise to be transferred between U-SI and Bedini and that all U.S. sales for both companies are made through their U.S. affiliate, Ugine Stainless and Alloys, Inc. ("US&A"), there is the chance that the origin of the subject merchandise could also be easily manipulated. Furthermore, the petitioners argue that because U-SI and Bedini both sell through US&A in the United States, the expenses incurred by US&A should be calculated and reported in the same manner regardless if the expenses are associated with U-SI or Bedini bar. However, in the current investigations (1) U-SI and Bedini calculated identical expenses incurred by US&A using different allocation methodologies. The petitioners argue that US&A is a single company that maintains one set of accounting records and its use of different allocation methodologies for different respondents is suspect. The petitioners assert that this is an indication of the type of manipulation that is possible if U-SI and Bedini are not collapsed. Moreover, the petitioners argue collapsing U-SI and Bedini, and requiring the same expenses to be calculated identically, regardless of whether the bar originated from U-SI or Bedini would prevent any future distortion of US&A's expenses. In addition, the petitioners assert the Department should collapse all merchandise tolled by Bedini for U-SI into a single home market database. However, if the Department decides not to collapse U-SI and Bedini, it should at a minium insert the tolled sales into the French database given that U-SI is the principal producer of the underlying product, maintains title to the merchandise, and the finishing operations performed by Bedini in this case have not been shown to be significant. Finally, the petitioners suggest that although the issue of whether to collapse may not have been dealt with in a multinational situation in prior cases, this should not preclude the Department from deciding to collapse U-SI and Bedini because the Department clearly states that the decision of whether or not to collapse affiliated producers should be based solely on the conditions set forth in 19 C.F.R. 351.401(f). Accordingly, the petitioners propose that the Department should require U- SI and Bedini to submit their sales and cost databases on the record of respective French and Italian investigations. Alternatively, the petitioners request that the Department calculate separate margins for each respondent and then calculate a single weighted-average margin to be applied to both respondents. The respondent argues that if the Department collapses the sales prices and production costs of U-SI and Bedini and calculates a single dumping margin, this would be contrary to the plain language of the antidumping duty law. The respondent contends that the antidumping duty law operates with regard to merchandise from a particular country and not from multiple countries. (2) Furthermore, the respondent maintains that the plain language of the antidumping duty law and the Department's regulations reflect this one-country approach by requiring dumping petitions to "name the country in which the subject merchandise is manufactured or produced," define the "foreign like product" in terms of merchandise that was produced in the same country as the subject merchandise, and base normal value on "the price at which the foreign like product is first sold . . . in the exporting country." (3) The respondent maintains that collapsing affiliated manufacturers is only permissible under the Department's regulations in an antidumping proceeding, which, by its terms, can only be with regard to a single country. Since U-SI's and Bedini's sales and cost data relate to SSB manufactured in different countries, the respondent contends that these two companies cannot be collapsed under the antidumping duty law and treated as a single entity in the antidumping proceeding involving SSB from France. Department's Position: We disagree with the petitioners. In an antidumping duty investigation, the Department is charged with determining whether "a class or kind of foreign merchandise is being, or likely to be, sold in the United States at less than its fair value . . . ." Section 731 of Act. To offset dumping, an antidumping duty is imposed equal to "the amount by which normal value exceeds export price (or constructed export price) of the merchandise." Id. The Department, therefore, must make a comparison between export price or constructed export price and normal value. Section 773(a) of the Act. Normal value is defined as "the price at which the foreign like product is first sold (or, in the absence of a sale, offered for sale) for consumption in the exporting country . . . .". Section 773(a)(1)(B)(i) of the Act. Or, normal value may be based on the price of the "foreign like product" in a third country. Section 773(a)(1)(B)(ii) of the Act. In certain instances, normal value may be based on constructed value. Section 773(a)(4) of the Act. In determining constructed value, the statute instructs the Department to use the selling, general and administrative expenses incurred in connection with "the production and sale of a foreign like product . . . ". Section 773(e)(2)(A) of the Act. Thus, the inescapable conclusion is that the Department must look to the foreign like product when calculating normal value. Section 771(16) of the Act gives a number of alternative definitions of "foreign like product." In all of the alternative definitions, the foreign like product must be produced in the "same country" as the subject merchandise. "The term 'country' means a foreign country, a political subdivision, dependent territory, or possession of a foreign country, and, except for the purpose of antidumping proceedings, may include an association of 2 or more foreign countries, political subdivisions, dependent territories, or possessions of countries into a customs union outside the United States." Section 771(3) of the Act. Therefore, in an antidumping investigation, normal value can only be calculated by looking to the subject merchandise in one country. The regulations set forth the rules for collapsing. The regulations begin by stating, "In an antidumping proceeding under this part, the Secretary will treat two or more affiliated producers as a single entity where those producers have production facilities for similar or identical products that would not require substantial retooling of either facility in order to restructure manufacturing priorities and the Secretary concludes that there is a significant potential for the manipulation of price and production." 19 CFR 351.401(f). Thus, the regulations make clear that collapsing can only occur within "an antidumping proceeding". Because an antidumping proceeding only involves the subject merchandise of one country, this means that the Department cannot collapse producers across country lines under 19 CFR 351.401(f). The petitioners here have argued that the Department could calculate one margin for U-SI in France, and one margin for Bedini in Italy, and then calculate a weighted average of those margins to come up with one antidumping duty margin to apply to both U-SI and Bedini. The petitioners, however, have no statutory authority upon which to make this request. Indeed, if the Department were to do so, the Department would be violating its statutory mandate. In calculating such a weighted average of the margins, the Department would be injecting into the calculation of normal value the price of merchandise produced in a second country. While the Department has used the information from two companies to calculate a single weighted-average margin for those companies when it determined collapsing was appropriate, the Department has done so only within the confines of a single proceeding, which involved a single country. See, e.g., Final Results of Antidumping Duty Administrative Review: Gray Portland Cement and Clinker From Mexico, 66 FR 14889 (March 14, 2001). Finally with respect to the petitioner's request that the Department include in this proceeding the SSB which was sold by the Italian respondent Bedini in Italy but which was produced by U-SI, we do not consider this merchandise to be subject to the either proceeding because U- SI is not a respondent in the Italian SSB proceeding and Bedini is not a respondent in the French SSB proceeding. Comment 2: Defining Foreign Like Product for Making Product Comparison The respondent contends that the model matching methodology used in the preliminary determination was fundamentally flawed because it created artificial distinctions between comparable products which skewed the fair comparison between normal value and U.S. price. Specifically, the respondent argues that the Department undermined the accuracy of the dumping calculations by insisting that it report the foreign like product based on specific size ranges for cost purposes and on exact sizes for sales purposes. The respondent maintains that the antidumping duty law precludes the Department from using a different definition of the foreign like product for product matching purposes than what it uses for determining costs. In support of its argument, the respondent cites to a decision made in United States Court of Appeals for the Federal Circuit in which the Court recently rejected the Department's use of inconsistent foreign like product definitions (SKF USA Inc. v. United States, 263 F. 3d 1369, 1382 (CIT 2001) ("SKF")). The respondent contends that there is no explanation or reasoning to justify the Department's use of an inconsistent foreign like product definition for sales product matching purposes, especially because the Department recognizes in this case that minimal size variations within a specified range have no price effect. Moreover, the respondent argues that by adopting an inconsistent like product definition for product matching purposes, the Department also breached its obligation to make a fair comparison (i.e., comparing the weighted-average prices of "comparable merchandise in the U.S. and home markets). The respondent further argues that sections 351.414(d)(2) and 351.414(c) of the Department's regulations instruct it when applying the average-to-average method to identify those sales of subject merchandise to the United States that are comparable, and to include such sales in an "averaging group." Moreover, the respondent maintains that the averaging group is to consist of "merchandise that is identical or virtually identical in all physical characteristics." The respondent contends that the Department has already identified the comparable merchandise for purposes of the proper averaging group, by way of its adoption of size ranges for cost reporting purposes. By using size ranges, the respondent states that the Department has treated SSB products within the specified size ranges as identical or virtually identical and recognized that minuscule size variations within a range are not meaningful physical differences and do not affect price. Moreover, the respondent argues that an analysis of the data submitted to the Department shows that, if all home market sales of comparable products (i.e., otherwise wholly identical products within a size range) were utilized for matching purposes, a significant percent of viable home market sales would match to a comparable U.S. product. However, the respondent states, in the preliminary determination, the Department's approach threw out a significant portion of possible matches by restricting comparisons to exact size, and only a small percentage of U- SI's home market sales were used in the preliminary determination. The respondent maintains that product matching by exact size resulted in the artificial division of individual products, which the Department has recognized as identical for cost purposes. Additionally, the respondent argues that the skewed results of the Department's exact size matching approach is even more apparent when examining certain control numbers. The respondent maintains that requiring certain control numbers to be reported by exact size artificially dissected products otherwise identical in all respects within a size range, with no real distinctions. Moreover, the respondent alleges that some of the size differences were as small as one- thousandth of a millimeter. Finally, the respondent contends that the Department's exact size approach contradicts its own model matching criteria. Specifically, the respondent argues that the Department's approach restricted the scope of comparable products to which a product might have been matched. For example, the respondent states that control numbers (which were assigned exact sizes) were matched to products outside their applicable size range, thereby turning the Department's product matching on its head by matching to a different product even though sales of comparable merchandise as defined by the Department were available. For the final determination, the respondent requests that the Department discard exact size matching and make comparisons by the size ranges which it has established as representing comparable merchandise. The respondent contends that only in this way can the fair comparison mandated by the statute be achieved. The petitioners maintain that the Department did not err in the preliminary determination by using a different definition of foreign like product for its model matching methodology. Moreover, the petitioners argue that the respondent's reliance on a Federal Circuit opinion handed down in SKF is misplaced. Specifically, the petitioners argues that in SKF, the Court considered an appeal by respondents as to whether the Department erred in defining the normal value on sales of identical antifriction bearings ("AFBs") and sales of the same AFB family, while defining the constructed value profit on all AFB products (regardless if these products were identical or in the same family). The petitioners maintain that in citing the complexity of the statute and the deference provided by the Court to the Department, the Court remanded the issue back to the Department for additional explanation of the factual settings for the calculations at issue, and to explain exactly how those calculations are made. The petitioners contend that as an initial point, the SKF appeal does not offer any guidance by the Court as to how the Department must interpret and apply the definition of foreign like product. In addition, the petitioners argue that the Court in the SKF appeal did not address the issue raised by the respondent in its case brief at to whether the Department erred in its price-to-price comparisons vis-a-vis its below cost test. Therefore, the petitioners contend that little if any weight can be given to the Federal Circuit opinion in SKF. Furthermore, the petitioners argue that the Department's definition of control number for the sales and costs enable the Department to uphold its statutory responsibility to calculate the most accurate dumping margin possible, because those definitions appropriately reflect the unique nature of the sales and production of SSB. The petitioners contend that in producing SSB, the cost will generally vary by ranges of size, not by specific size, because a factory will only retool if it is planning to change from one size range to another size range. Therefore, the petitioners assert that the cost should reflect the same fact pattern. In contrast, the petitioners assert that the selling price of each product reflects the actual size ordered. In similar fashion, the petitioners conclude that the model matching methodology used for defining the sales should reflect this market condition. The petitioners argue that it was necessary for the Department to gather cost data in a manner that reflects the actual cost of the product, and to define the product sold as it is priced in the stainless steel market. Department's Position: We disagree with the respondent that the Department has used different definitions of foreign like product in the same proceeding and that we should make the price comparisons based on the size product groupings used for the cost of production analysis. Section 771(16)(A) of the Act defines "foreign like product" as merchandise that is "identical in physical characteristics." The Department reviewed comments by parties to this investigation prior to defining six physical characteristics of stainless steel bar that it considered appropriate for model matching criteria. Section 777A(d)(1)A)(i) of the Act specifies that the Department compare "the weighted average of the normal values to the weighted average of the CEPs for comparable merchandise." In the Preliminary Determination, the Department's price averaging groups were defined by the six physical characteristics reported by the respondent, although some modifications may have been made by the Department on a case-by-case basis. It is significant to note that CEP and NV prices were defined by the exact same physical characteristics reported by the respondent, meaning that the same definition of foreign like product was used for both price averaging groups. With respect to the respondent's argument that different definitions were used to define the "size" physical characteristic for price averaging and cost reporting, we note that the preamble to the Department's regulations (61 FR 7339) states that "the Department's practice is to calculate costs consistent with the model matching criteria it develops at the outset of an investigation or review, after having received the views of the interested parties. The product categories developed in such fashion generally account for significant differences in actual costs affecting price. The Department intends to continue this practice because it prevents any manipulation of the cost analysis through changes in internal product classifications." Thus, while we take the model match (product definition) criteria into account in determining the appropriate methodology for reporting cost, this does not mean that in every case we can or must calculate a unique cost for every unique product (CONNUM) identified by our product definition criteria. Nor does it mean that if we accept or adopt a cost methodology that yields the same cost for two products, we must consider those two products to be a single product (CONNUM), or a single foreign like product for model matching purposes. Finally, we disagree with the specific argument that the Department's use of size ranges for cost reporting purposes results in or is based on the use of a different definition of foreign like product. For purposes of reporting costs, the Department instructed the respondent to follow the 3 size ranges proposed by the Department, with the significant caveat that if the company's normal books and records differentiate the production costs associated with size using slightly different ranges, and the company wished to report cost using company-specific ranges, then they should contact the Department. See Memorandum to Susan Kuhbach and Louis Apple entitled, "Selection of Model Matching Criteria for Purposes of Issuing the Antidumping Duty Questionnaire," dated February 20, 2001. The Department initially created the 3 size ranges for reporting costs in recognition of its experience in stainless steel bar cases that a respondent may not keep its cost records by specific sizes and that costs generally do not vary within specific ranges of sizes. The Department also recognized that companies that do record their costs by internally unique size ranges may use cost ranges different than those proposed by the Department, thus the caveat that such companies "should contact the Department." In other SSB cases, certain companies did indeed contact the Department on this issue, which led the Department to revise its size reporting requirements for those companies with respect to cost: The original cost questionnaire requested that parties contact the Department if they planned to depart from the designated size ranges. Therefore, for clarification and consistency between companies we request the following. Your reported per-unit cost of production ("COPs") and constructed value must account for cost differences between shapes (e.g., rounds, flats, rectangles, etc.). For size differences within shapes, you must follow the size categories from your normal books and records (i.e., inventory valuation or existing cost accounting records). If you do not differentiate costs by size in your normal books and records then you must use the ranges identified by the Department in the original questionnaire. See e.g., Letter dated May 24, 2001, to U-SI from Irene Darzenta Tzafolias Thus, for cost reporting purposes, the respondent was advised to follow the size categories maintained in its ordinary books and records. The record in this investigation shows that the respondent complied with these instructions by reporting its costs based the Department's size ranges. Other respondents in the SSB investigations complied with these instructions by reporting their costs based on exact sizes, company- specific size ranges or the Department's size ranges. Unless adjustments to a particular respondent's cost size ranges were warranted based on the results of verification or other information on the record, it is reasonable to conclude that the respondents (including U-SI) have reported their costs for specific sizes on as specific a basis as possible for that company, whether that is exact size, company size range, or the Department's alternative size range. Thus, where a respondent reported costs by size range, the cost for any specific size range corresponds to the exact cost for any specific size within that range, and we have not applied different definitions of foreign like product in defining price averaging groups and in calculating cost. Comment 3: Ranking "Peeled and Descaled" as a Final Finish Characteristic The petitioners contend that the respondent's hierarchical coding for its final finish operations is incorrect even though its reported coding is in accordance with the Department's questionnaire instructions. The petitioners reiterate their alternative proposal expressed prior to the preliminary determination that for matching purposes the Department re- code the reported final finishes by grouping together products that have undergone similar degrees of finishing operations into the following four groups: (1) basic finishing operations; (2) intermediate levels of finishing; (3) basic cold finishing; and (4) more expensive cold finishing operations. For example, the petitioners urge the Department to code and group products together that underwent basic finishing operations (i.e., unannealed, pickled, etc) separately from products that underwent intermediate levels of finishing (i.e., rough-turned, peeled and descaled, etc.) and from products that either underwent basic cold finishing (i.e., cold-drawn) or more expensive cold finishing operations (i.e., turned and burnished). In addition, the petitioners contend that because the peeled and descaled operation is an intermediate level of finishing, it should be coded as one but ranked after the rough-turned operation in terms of the hierarchical coding employed for this model matching characteristic. The respondent agrees that the peeled and descaled finish should be assigned a separate final finish code but disagree with the petitioners that it should be ranked after the rough-turned finishing operation in the hierarchical order. Instead, the respondent requests that the Department rank the peeled and descaled operation before the rough-turned operation. The respondent provides no further explanation in support of its request. Department's Position: We disagree with the petitioners that we should be altering the hierarchical ranking order utilized for the final finish model matching characteristic in this case. Based on the data contained in this record, we are not pursuaded that the petitioners' proposed changes mentioned above are appropriate to incorporate in the final determination. Specifically, we are unable to determine whether reordering the final finish characteristic in the manner proposed by the petitioners yields more accurate results than the method we employed in the preliminary determination. Therefore, we have continued to employ in the final determination the hierarchical ranking order and coding system used in the preliminary determination for the final finish product characteristic. In addition, although we recognize that the peeled and descaled operation is a unique finish operation and have treated it as one in the final determination, we do not find a sufficient basis for re-ranking this characteristic in the hierarchical coding order. Comment 4: Assigning Total Facts Available The petitioners argue that the Department should resort to total facts available with respect to U-SI because the respondent failed to provide certain unreported home market sales because of a computer programming error until verification and the total quantity and value of these unreported sales is significant in terms of the total reported quantity and value of home market sales. In addition, the petitioners maintain that by declaring these sales as a minor correction at the start of verification, the respondent failed to notify the Department of these missing sales in a timely manner or as soon as they were discovered by the respondent and thus failed to cooperate to the best of its ability. In addition to this error, the petitioners maintain that the respondent made several other errors in its sales reporting (i.e., finish and grade coding errors, unreported and/or improper inclusion of sales of non- subject merchandise by its home market affiliate resulting from country-of- origin errors, and omission of sales based on improper date of sale treatment) which in their totality bring into question the reliability of the respondent's data and compel the Department to resort to total facts available with respect to the respondent. At a minimum, the petitioners request the Department to resort to partial facts available with respect to the missing and/or erroneous data problems noted above given the seriousness of these errors. Alternatively, the petitioners contend in its December 7, 2001, submission, filed in accordance with 19 CFR 351.301, that the home market sales database submitted by the respondent on November 27, 2001, at the request of the Department, contains new factual information and should not be used by the Department in this determination. Specifically, the petitioners maintain that the respondent added numerous expense, level-of- trade, and customer data fields of information (including the date of payment) applicable to the unreported sales which were not reflected as minor corrections at the start of verification. As a result, the petitioners argue that the home market sales database filed by the respondent on November 27, 2001, should not be used by the Department. Moreover, the petitioners allege that the total quantity and value of unreported sales provided by the respondent at verification differs from the total quantity and value corresponding to those sales in the November 27, 2001, home market sales database. Therefore, the petitioners argue that the data the respondent did provide for the unreported sales is incomplete, unreliable, and unuseable for the final determination. In support of its argument, the petitioner cites to Tung Mung Development Co., Ltd. and Yieh United Steel Corp. v. United States, Slip Op. 01-83 (July 3, 2001) (Court of International Trade) ("Tung Mung"). The respondent maintains that it has fully cooperated with every aspect of the Department's investigation and has acted to the best of its ability to comply with all the Department's requests for information before and after verification. Specifically, the respondent claims that the unreported sales (which resulted from a computer programming error) at issue were discovered just prior to verification and were brought to the Department's attention at the first opportunity to do so (i.e., the beginning of verification). Moreover, the respondent maintains that the Department thoroughly examined the programming error and tested the accuracy of the data through sampling at verification. With respect to the sales inclusion/exclusion corrections mentioned by the petitioners of which U-SI informed the Department at the start of verification, the respondent contends that it has provided the Department with accurate and reliable data which was also examined by the Department at verification. Moreover, with respect to not reporting certain sales based on its date of sale methodology, the respondent maintains that the Department also examined this issue at verification and found that its use of invoice date as the date of sale for purposes of reporting all of its home market sales was appropriate. Therefore, the respondent claims that the petitioners' arguments for resorting to total or partial facts available are without merit. In its December 7, and 11, 2001 rebuttal submissions, which address the inclusion of additional fields of information contained in its November 27, 2001, home market sales database submission, the respondent states that the additional fields of information do not constitute new factual information because they were furnished to the Department in accordance with the Department's November 13, 2001, letter. Specifically, the respondent maintains that the Department requested in that letter that it provide a home market sales database which incorporated the results noted in the October 25, 2001, sales verification report ("sales verification report") and that these additional fields were covered by the Department's request. For example, for the additional expense fields, the respondent contends that the data contained in those fields of information are derived using expense-specific calculation formulas which the Department examined at verification. Therefore, the respondent contends that the data provided for the expense fields are already on the record. With respect to the remaining fields of information pertaining to the customer, level of trade, and dates of payment and shipment, the respondent claims that these fields were necessary in order to provide a complete record for each transaction based on the Department's November 13, 2001, letter (which instructed it to include the omitted sales in a home market sales database which incorporated the results from verification). With respect to the difference in total quantity and value referred to by the petitioners, the respondent explains that the difference in quantity results from a slight omission it made with respect to specific billing adjustment amounts associated with certain sales included in the listing provided at verification which it later corrected in the November 27, 2001, home market sales database provided to the Department. Department's Position: We agree with the respondent. At verification, U-SI promptly alerted the Department that it had made a programming error in determining which home market sales made by its home market affiliate should have been included in the home market sales database. The Department thoroughly examined the nature of this error at verification and was able to ascertain the nature and magnitude of the error based on the respondent's detailed explanation, and based on the Department's testing of respondent's sales data during the completeness portion of the verification (see U-SI exhibits 7A through 7G and pages 18 to 21 of the sales verification report). In addition, the respondent provided the Department at verification with a listing of all sales affected by the programming error which the Department sampled at verification by examining sales invoices and country-of-origin documentation. Based on our verification findings and an examination of the respondent's list of corrections, we requested the respondent to provide on November 27, 2001, revised sales and cost databases which incorporated the findings listed in the verification report (see the Department's November 13, 2001, letter to respondent, the sales verification report, and the Department's November 6, 2001, cost verification report). The petitioners are correct in that the fields of information the respondent provided the Department at verification for these affected sales numbered only 12, but these fields represented the most essential fields of information for purposes of examining the sales transactions and tracing the data from the sales listing to data contained in the sales invoices. Expense fields were not included in the sales listing provided to the Department at verification. However, most of those expense fields (with the exception of credit, warranty, and inventory carrying expenses) were derived by applying allocation formulas noted in the respondent's response which the Department either examined at verification or elected not to examine at verification. Therefore, the Department also examined the expense fields at issue by virtue of examining the allocation formulas used to derive the per-unit amounts in the fields of information contained in the November 27, 2001, filing. For the remaining three expense fields of data (i.e., credit, warranty, and inventory carrying expenses), and additional fields of data included in the November 27, 2001, filing, but not contained in the sales listing provided at verification (i.e., product code, customer code, shipment date, level of trade), we consider these fields of data related to the sales transactions and to the data fields the Department requested U-SI to provide in its November 13, 2001, letter. Moreover, since we are using the sales at issue in our analysis, the additional fields of information are necessary in order to consider all of the data which is relevant for those sales transactions at issue. Therefore, we have also used the data provided for those other fields in our analysis. For certain sales at issue and for other transactions as well, we traced the data contained in these additional fields of information to corresponding data contained in source documentation (i.e., sales invoices) (see U-SI exhibits 7F, and 9A through 9I of the sales verification report). In very few instances did we find errors in the data reported for the sales transactions we examined at verification. In addition, verification is not intended to examine all data related to a particular sales transaction in order determine the reliability of each piece of data associated with a given sales transaction. Unlike the situation in Tung Mung, the Department has useable data on the record of this proceeding for the sales at issue and is able to incorporate this data into its analysis for the final determination. Moreover, unlike Tung Mung, the Department has examined the usable data through sampling techniques employed at verification and considers the data contained in the extra fields of information for the sales at issue to be reliable based on its examination of similar data for other sales transactions sampled at verification. With regard to the other errors mentioned by the petitioners with respect to the respondent's sales reporting (i.e., finish and grade coding errors, unreported and/or improper inclusion of sales of non-subject merchandise by its home market affiliate resulting from country-of-origin errors), these errors were thoroughly examined by the Department at verification and the Department was able to ascertain through utilization of various verification techniques that the errors were isolated and minor in nature such that they did not as a whole bring into question the reliability of the respondent's reported home market sales (see pages 16 and 19 of the sales verification report). With respect to certain sales not reported based on U-SI's date of sale methodology, the Department finds that U-SI properly excluded those sales from the home market sales database based on the fact that the blanket order covering those affected sales only specified a base price for certain products and not the quantity or alloy surcharge. Therefore, the Department finds that U-SI was justified in reporting any sales covered by that blanket order based on the date of invoice rather than the date of the blanket order. With respect to the slight difference in net quantity and value between what was included in the listing provided at verification and what was included in the November 27, 2001, submission, the Department finds the difference has been sufficiently explained by the respondent in its December 7, 2001, submission, and does not merit calling into question the accuracy of the data contained in the November 27, 2001, submission. Therefore, we are using all the data contained in the November 27, 2001, submission in our analysis for the final determination. Finally, we recognize that both the petitioners and the respondent filed submissions on the issue of whether the Department should use the sales at issue after they submitted their case briefs and after the respondent submitted the revised database per the Department's request (which was due about one week after the case briefs were due) (see November 19, 2001, case briefs and November 27, 2001, revised database submission). In accordance with 19 CFR 351.301, the Department normally does not allow parties to submit arguments after the deadline for submitting case briefs. However, in this case, because the revised database was due to the Department after the case briefs were filed, this was the first opportunity the petitioners had to comment on the updated information. Therefore, we have allowed the petitioners' December 7, 2001 submission to rebut the use of respondent's data to remain on the record pursuant to 19 CFR 351.301(c)(1) and responded to their argument in this memorandum. Comment 5: Calculating the Dumping Margin The respondent alleged that in the preliminary determination the Department inappropriately disregarded negative margin sales in calculating its dumping margin, and this unfairly and inaccurately inflated its overall dumping margin. In calculating the respondent's dumping margin for the preliminary determination, the Department assigned a zero to the transactions in which the U.S. price exceeded normal value. The respondent argues that this practice violates U.S. and international law, which requires that the Department must make a "fair comparison" between the export price or constructed export price and normal value. Moreover, the respondent argues that a recent WTO Appellate Body decision (4) found that the practice of "zeroing," under which export sales made at prices greater than normal value are assigned a zero dumping margin, violates Article 2.4 of the WTO Antidumping Agreement. Furthermore, the respondent argues that the Department's position in Final Determination of Sales at Less Than Fair Value: Certain Hot-Rolled Carbon Steel Flat Products From the Netherlands, 66 FR 50408 (October 3, 2001) ("Hot-Rolled from the Netherlands") in which it dismissed the applicability of the Appellate Body's Bed Linen decision is incorrect, and should not be followed. Contrary to the Department's position in Hot-Rolled from the Netherlands that (1) the "zeroing methodology is "required by U.S. law" (citing section 771(35)(A) of the Act); (2) even though the zeroing methodology produces unfair results, it is nevertheless "reasonable" especially in an investigation; and (3) U.S. law takes precedence over any "potentially conflicting obligations under the Uruguay Round Agreements", the respondent argues that no provision of U.S. law requires the Department to ignore negative margin sales. Further, the respondent argues that no methodology that fails to effectuate a "fair comparison" pursuant to section 773(a) of the Act can be considered a reasonable methodology, and where U.S. and international law obligations do not conflict it is established that U.S. law is to be read in harmony with international obligations. For these reasons, the respondent requests that for the final determination the Department recalculate its margin to account for negative margin sales. The petitioners argue that the Department correctly disregarded negative margin sales in the preliminary determination. The petitioners contend that the Department explicitly found that the WTO's Bed Linen from India decision does not apply to U.S. proceedings. See, e.g., Certain Preserved Mushrooms from India: Final Results of Antidumping Duty Administrative Review, 66 FR 42507 (August 13, 2001), and accompanying Issues and Decision Memorandum at Comment 16 ("Mushrooms from India"). As the Department noted in Mushrooms from India, the Bed Linen from India Panel and Appellate Decision concerned a dispute between the EU and India. The Department has no WTO obligations to act. Therefore, The Department has continued the practice of using zero where the NV does not exceed the export price or CEP in our calculations of overall margins for the final results. The petitioners argue that the respondent's argument should be similarly rejected in this investigation as well. Department's Position: We disagree with the respondent. As we have discussed in prior cases, our methodology is consistent with our statutory obligations under the Act. See, e.g., Hot-Rolled from the Netherlands, at FR 50408 (October 3, 2001), and accompanying Issues and Decision Memorandum, at Comment 1. First, sales that did not fall below normal value are included in the weighted- average margin calculation as sales with no dumping margin. The value of such sales is included in the denominator of the weighted-average margin along with the value of dumped sales. We do not, however, allow sales that did not fall below normal value to cancel out dumping found on other sales. Second, the Act requires that the Department employ this methodology. Section 771(35)(A) of the Act defines "dumping margin" as "the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise." Section 771(35)(B) of the Act defines "weighted-average dumping margin" as "the percentage determined by dividing the aggregate dumping margins determined for a specific exporter or producer by the aggregate export prices and constructed export prices of such exporter or producer." These sections, taken together, direct the Department to aggregate all individual dumping margins, each of which is determined by the amount by which normal value exceeds export price or constructed export price, and to divide this amount by the value of all sales. The directive to determine the "aggregate dumping margins" in section 771(35)(B) makes clear that the singular "dumping margin" in section 771(35)(A) applies on a comparison-specific level, and does not itself apply on an aggregate basis. At no stage in this process is the amount by which EP or CEP exceeds normal value on sales that did not fall below normal value permitted to cancel out the dumping margins found on other sales. This does not mean, however, that sales that did not fall below normal value are ignored in calculating the weighted-average rate. It is important to note that the weighted-average margin will reflect any "non-dumped" merchandise examined during the investigation, the value of such sales is included in the denominator of the dumping rate, while no dumping amount for "non-dumped" merchandise is included in the numerator. Thus, a greater amount of "non-dumped" merchandise results in a lower weighted-average margin. This is, furthermore, a reasonable means of establishing duty deposits in investigations, and assessing duties in reviews. In an investigation such as the present case, the deposit rate calculated must reflect the fact that the Customs Service is not in a position to know which entries of merchandise entered after the imposition of a dumping order are dumped and which are not. By spreading the estimated liability for dumped sales across all investigated sales, the weighted-average dumping margin allows the Customs Service to apply this rate to all merchandise entered after an order goes into effect. Finally, with respect to the respondent's WTO-specific arguments, we believe U.S. law is consistent with our WTO obligations. Moreover, the Bed Linen from India Panel and Appellate Body decisions concerned a dispute between the European Union and India. We have no obligation under U.S. law to act on this decision. Comment 6: Level of Trade The petitioners argue that the Department should find that one level of trade ("LOT") exists in the home market and that this LOT is comparable to the LOT in the United States. Specifically, the petitioners requested that for the final determination the Department find that ex-works sales (Channel 2 sales) are also at the same LOT as ex-inventory sales (Channel 3 and 4 sales). The petitioners contend that to distinguish ex-works sales (Channel 2) from ex-inventory sales (Channel 3 and 4) is nothing more than an artificial segregation of the manufacturing process and this does not constitute or support finding a difference in LOT between the two channels. Furthermore, the petitioners contend that all of the ex-works and ex-inventory sales are in reality mill direct sales to the customer and the place of manufacture does not (and cannot) affect or advance the LOT. Moreover, the petitioners contend that because the Department has treated certain selling and movement expenses between U-SI and U-SI's affiliate, Ugine France Service/Ugine-Savoie France ("UFS/U-SF"), as manufacturing costs rather than as selling expenses, the Department is essentially measuring net home market price for ex-inventory sales (Channel 3 and 4) from UFS/U-SF's door. Hence, the petitioners contend that the Department should treat these ex-inventory sales as mill direct sales based on the rationale that the selling functions performed by both U-SI and UFS/U-SF should be combined in the same manner as are the manufacturing expenses for purposes of determining the total manufacturing cost of SSB which UFS/U-SF further processed prior to sale to the unaffiliated customer. For the reasons mentioned above, the petitioners request that the Department find that there is a single LOT in the home market and that this LOT is equal to the LOT in the United States. The respondent argues that in the preliminary determination, the Department erred in finding that U-SI's home market ex-works sales (Channel 2) and its CEP sales were made at the same LOT, and that UFS/U- SF's ex-inventory sales of standard SSB (Channel 3) and UFS/U-SF A.P's sales of specialized SSB (Channel 4) were made at the same LOT. The respondent contends that the Department's preliminary LOT analysis was fundamentally flawed. Instead of specifically considering and evaluating all the selling functions and expenses reported, the Department generalized the respondent's 24 specific selling functions into four broad categories. The respondent argues that by grouping distinct selling functions together the Department diluted and masked the significant differences that exist in the selling activities performed in the home market LOTs and those selling activities performed for CEP sales. The respondent maintains that its home market affiliated reseller, UFS/U-SF, undertakes significantly more selling functions in the home market by selling at more remote stages of distribution compared to U-SI's selling functions for its CEP sales to US&A. In addition, the respondent argues that the petitioners' assertion that one LOT exists in the home market ignores the distinct stages of distribution represented in the respondent's home market distribution network by treating the product U-SI sold to UFS/U-SF through channels of distribution as semi-finished product and considering the further processing services provided by UFS/U-SF as manufacturing, rather than selling, in nature. Furthermore, the respondent argues that in order to achieve a fair comparison between CEP and normal value as required by the antidumping duty law, the Department must make a LOT adjustment to normal value. The respondent maintains that this LOT adjustment is necessary to account for the significant extra layers of selling activities performed and significant extra expenses incurred for home market sales vis-a-vis CEP sales. Therefore, for the final determination, the respondent requests that the Department find three distinct levels of trade in the home market (i.e., direct ex-works sales (Channel 2), ex-inventory sales of standard SSB (Channel 3), and ex-inventory sales of specialized SSB (Channel 4)), each of which is at a more advanced stage of distribution than the respondent's CEP sales. Department's Position: We agree in part with the respondent. For the final determination, we have continued to find that two levels of trade exist in the home market but have determined that each of those levels of trade are distinct LOTs which are different from the LOT which exists in the U.S. market. We examined whether a LOT adjustment was appropriate by comparing the CEP sales to home market sales at a different LOT. Since the data available did not provide an appropriate basis to determine a LOT adjustment, we are unable to make this adjustment. However, since the LOTs in France are a more advance stage than the LOT of the CEP sales, a CEP offset is appropriate. As we stated in the preliminary determination, in accordance with section 773(a)(1)(B) of the Act, to the extent practicable, we determine NV based on sales in the comparison market at the same LOT as the CEP transaction. The NV LOT is that of the starting-price sales in the comparison market. For CEP, the U.S. LOT is also the level of the starting-price sale, which is usually from the exporter to the U.S. affiliate. To determine whether NV sales are at a different LOT than CEP, we examine stages in the marketing process and selling functions (5) along the chain of distribution between the exporter and the U.S. affiliated reseller. If the comparison-market sales are at a different LOT and the difference affects price comparability, as manifested in a pattern of consistent price differences between the sales on which NV is based and comparison-market sales at the LOT of the CEP transaction, we make a LOT adjustment pursuant to section 773(a)(7)(A) of the Act. Finally, for CEP sales only, if a NV LOT is more remote from the factory than the CEP LOT and we are unable to make a LOT adjustment, the Department shall grant a CEP offset, as provided in section 773(a)(7)(B) of the Act (see Notice of Final Determination of Sales at Less Than Fair Value: Certain Cut-to-Length Carbon Steel Plate from South Africa, 62 FR 61731 (November 19, 1997)). In our preliminary determination, in reviewing the channels of distribution, we determined that home market sales made by U-SI's affiliate (i.e., UFS/U-SF) from its inventory (either through UFS/U-SF's AP department or non-AP departments) were made at the same LOT because we considered the upgrading and other special operations (6) performed by UFS/U-SF's AP department not to be selling functions, and thus not relevant to our LOT analysis. After further examining the functions the AP department performs in making sales to its customer from inventory, through review of source documentation at verification, we continue to find that the extra operations which the AP department performs (which the respondent claims translate into a LOT distinct from the affiliate's non- AP department sales made from inventory) are manufacturing operations which should not be taken into account in our LOT analysis (see U-SI exhibits 6A, 9A through 9I, and page 12 of the sales verification report). These manufacturing steps are no more selling functions than any of the prior steps in the manufacturing process. We do recognize that in arranging for these additional manufacturing operations, the AP department needs to spend more time with the customer as a result of the need to obtain in the majority of cases additional certifications through third parties for any additional work performed on the SSB purchased sold through the AP department. Nevertheless, for the reason previously explained, we do not find that additional manufacturing and the need to obtain certifications to complete such work a sufficient basis for treating UFS/U-SF's ex-inventory non-AP sales and ex-inventory AP sales as separate LOTs. Accordingly, we find that the selling functions for the two channels of distribution are similar, if not identical. Therefore, for the final determination, we continue to find that these two channels of distribution represent one LOT. With respect to the petitioners' contention that UFS/U-SF's ex-inventory sales are at the same LOT as U-SI's ex-mill sales because the Department has decided to group in the COP the costs associated with the manufacturing operations (i.e., "selling" functions claimed by UFS/U-SF) performed by UFS/U-SF with respect to those ex-inventory sales, we continue to find that ex-mill sales were made at a different LOT than the ex-inventory sales. Specifically, we continue to conclude that there are two LOTs in the home market for the following reasons. First, there are certain selling functions (i.e., inventory maintenance) performed for ex- inventory sales but not for ex-mill sales. Although we do not consider the manufacturing operations UFS/U-SF performs for its ex-inventory sales to be selling functions and have treated the expenses associated with those operations (i.e., the movement and claimed expenses between U-SI and UFS/U- SF) as manufacturing costs rather then selling expenses, the treatment of those costs is separate from how we determine LOTs with respect to the expenses which UFS/U-SF incurred to make the sale to its customer and the selling functions it performed which are associated with those expenses. In this case, the maintenance of product inventory by UFS/U-SF prior to sale to the customer indicates that the selling functions for sales through this channel are significantly different from the selling functions associated with ex-mill sales made by U-SI. For example, selling products from inventory involves a greater degree of intensity in finding the customer and in the overall sales effort than is the case for a company that makes the product to order (i.e., ex-mill). Moreover, based on our verification findings, we confirmed that the selling functions (i.e., sales process/market research, sales calls, interactions with customers, freight, technical advice and warranty servicing) that UFS/U-SF provided for its ex-inventory sales were at a higher level of intensity than the same selling functions U-SI provided for its ex-works sales (see U-SI exhibit 6A of the sales verification report). Therefore, there is a sufficient basis for finding that U-SI's ex-mill and ex-inventory sales in the home market constitute two distinct LOTs and that U-SI's ex-inventory sales are at a more advanced LOT than U-SI's ex-mill sales. With respect to U-SI's sales to its U.S. affiliate, US&A, we stated in the preliminary determination that these sales were made at the same LOT as U-SI's direct ex-mill sales because we were not convinced that U-SI perform different selling functions or selling functions at different levels of intensity for its sales to US&A as it performed for its ex-mill sales. However, after examining at verification data necessary for determining whether or not these two levels were the same, we now determine that there is a sufficient basis for finding that the U.S. LOT is sufficiently different from the ex-mill sales LOT and thus should be treated as different LOTs for the following reasons. Based on documentation examined at verification (i.e., documents showing the sales process from U-SI to its U.S. affiliate, corporate business plans prepared by U-SI and US&A, and other documentation related to U.S. sales), it is evident that U-SI has no significant role in performing any selling functions in making sales to US&A for resale in the U.S. market. In fact, U-SI's selling functions are limited to simply invoicing US&A, arranging freight, and occasionally providing warranty services for the SSB it sells to US&A (see US&A and U-SI exhibits 6A of the sales verification report). On the other hand, for its ex-mill home market sales, U-SI performs certain selling functions (i.e., marketing activities, and follow-up customer support services such as technical services) which are not performed by it when making its U.S. sales to US&A (see also US&A exhibit 6A and U-SI exhibit 6A of the sales verification report). Based on this information, we find that the selling functions performed by U-SI on behalf of its ex-mill sales are at a higher level of intensity and greater in number than the selling functions it performs on behalf of its U.S. sales to US&A. Therefore, for the final determination, we are treating the home market ex-mill LOT and U.S. LOT as distinct LOTs. As a result, we now consider both the ex-mill and ex-inventory home market LOTs to be distinct LOTs in the home market and each different from the U.S. LOT. Therefore, because we compared the CEP sales to home market sales at a different LOT, we examined whether a LOT adjustment was appropriate. Since both home market LOTs are not the same as the U.S. LOT, we are unable to determine whether the differences between the home market LOTs affect price comparability, and thus unable to make an LOT adjustment. However, since both LOTs in France are a more advanced stage than the LOT of the CEP sales, we have granted a CEP offset as provided in section 773(a)(7)(B) of the Act. In making our comparisons, since the ex- mill home market sales are at a LOT that is less remote than the U.S. LOT, we have first compared the ex-mill home market sales to U.S. sales. Where such comparisons were not possible, we then compared U.S. sales to the more remote home market LOT (i.e., ex-inventory sales). Comment 7: Home Market Expenses Reported In Lieu of Commissions In some situations, UFS/U-SF, U-SI's home market affiliate, acts as an agent for U-SI by finding the home market customer for U-SI and directing that customer to it so that U-SI can make the sale to that customer on an ex-mill basis. UFS/U-SF reported in a separate field of information the expenses it incurred acting as a commission agent for U-SI and claimed these expenses in lieu of commissions. UFS/U-SF calculated these expenses by allocating a portion of its total indirect selling expense to this activity based on the total sales value associated with this activity. The petitioners argue that the Department should deny the respondent's claim for a commission adjustment on home market sales because there is no support for finding that the commissions U-SI paid its affiliate UFS/U-SF were at arm's-length. The petitioners maintain that the Department's practice is to deny this type of adjustment in situations where it cannot test the arm's-length nature of the affiliated party commission. Moreover, the petitioners argue that it is the Department's practice to allow an adjustment for related party commissions made at arm's-length only if they are tied directly to sales. Therefore, the petitioners argue that in the final determination the Department should not adjust normal value for the so-called commission paid between U-SI and UFS/U-SF. In support of its position, the petitioners cite to the Final Determination of Sales at Less Than Fair Value: Coated Groundwood Paper from Belgium, Finland, France, Germany and the United Kingdom, 56 FR 56359 (November 4, 1991). The respondent maintains that it is not claiming that the actual commissions U-SI paid to UFS/U-SF are at arm's-length or should be deducted from normal value. Rather, the respondent argues that the Department should subtract from normal value the expenses reported by UFS/U-SF which it incurred acting as a commission agent for U-SI. Moreover, the respondent contends that UFS/U-SF only incurs expenses associated with this type of activity if U-SI actually pays UFS/U-SF a commission after UFS/U-SF finds the customer for it and U-SI makes the sale to that customer. As such, the respondent points out that UFS/U-SF would not have incurred these type of expenses if it had not acted as a commission agent on behalf of U-SI and U-SI did not make a sale through this channel of distribution. Therefore, the respondent claims these expenses bear a direct relationship to the particular sale in question (i.e., the home market sales U-SI made on an ex-mill basis), and should be treated as a direct selling expense. In addition, the respondent points out that the Department cannot offset these home market expenses at issue by the amount of the U.S. commission because the Department has already deducted U.S. commissions from the U.S. gross unit price. Moreover, the respondent asserts that the home market expenses at issue can not be offset by the amount of U.S. expenses which were not incurred as a result of U.S. economic activity (i.e., indirect selling and inventory carrying expenses incurred in the home market on behalf of U.S. sales), because these U.S. expenses are not related to activities for which commissions are being paid. Therefore, the respondent requests that the Department consider the actual expenses associated with U-SI's commissioned sales to be direct in nature and deduct such expenses from normal value in the final determination. Department's Position: We agree with respondent that the expenses incurred by UFS/U-SF should be accounted for in the final determination but disagree with the respondent that they should be treated as direct selling expenses based on the fact those expenses are the same as other expenses reported as home market indirect selling expenses and based on the fact that those expenses cannot be linked directly to the sales at issue. Upon our examination of this issue at verification, we found that these expenses were extracted by UFS/U-SF from its total indirect selling expenses recorded in its accounting records. The Department notes that in general UFS/U-SF took into account the same expense items when it allocated its total pool of indirect selling expenses over the sales made on behalf of U-SI and over the sales it made from its inventory based on channel-specific sales ratios used for this purpose (see U-SI exhibit 10D of the sales verification report). Therefore, the Department concludes that the expense items (which UFS/U-SF reported for this expense at issue) are incurred for not only the sales it makes on behalf of U-SI but also for the sales it makes on its own behalf. Therefore, these expenses are common to all sales and accordingly should be treated as such. Therefore, the Department has treated the expenses at issue as UFS/U-SF indirect selling expenses associated with sales U-SI made on an ex-mill basis in the final determination. We have also included these expenses in the pool of home market expenses used to determine the CEP offset. Comment 8: Treatment of Movement and Selling Expenses Between Home Market Affiliates as Manufacturing Costs The respondent argues that the Department erred in its preliminary determination by treating U-SI's selling and movement expenses associated with it sales of SSB to UFS/U-SF as production costs, instead of adjusting normal value to account for these expenses. The respondent argues that the antidumping duty law and the Department's regulations recognize that manufacturers and their affiliated resellers incur separate and independent selling expenses for their respective sales and that normal value must be adjusted accordingly. The respondent further maintains that the adjustment must account for additional costs, charges and expenses incident to bringing the foreign product from the original place of shipment to the place of delivery, which in its case, is UFS/U-FS's warehouse. The respondent contends that the Department's preliminary determination appeared to mischaracterize the product sold by U-SI to UFS/U-FS as semi-finished rather than as finished SSB. The respondent maintains that this treatment is contrary to the facts on the record and what it demonstrated at verification. The respondent reiterates that U-SI produces and sells finished SSB directly to unaffiliated parties in France as well as to its affiliate UFS/U-SF. Furthermore, the respondent argues that the Department has long recognized that affiliated resellers, like UFS/U-SF, which function as steel service centers, represent a separate and distinct step in the selling-distribution chain. Moreover, the respondent maintains that the further processing operations performed by UFS/U-SF (i.e., cutting, subcontracted grinding and certification) are consistent with those activities considered by the Department to be within the province of affiliated resellers and steel services centers. The respondent argues that treatment of these entities as selling entities is supported by the North American Industry Classification System (formerly known as the "Standard Industrial Classification System"). Therefore, the respondent contends that the Department's preliminary determination is unsupported by law, fact or reason, and improperly disallowed U-SI the statutorily prescribed adjustment to normal value and CEP. In support of its argument, the respondent cites to the Amended Final Determination of Sales at Less Than Fair Value and Antidumping Duty Order: Stainless Steel Sheet and Strip in Coils from Germany, 64 FR 40557 (July 27, 1999) ("Sheet and Strip from Germany"); Preliminary Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Stainless Steel Sheet and Strip in Coils from the United Kingdom, 64 FR 85, 89 (January 4, 1999) ("Sheet and Strip from the United Kingdom"); and Preliminary Results of Antidumping Administrative Review: Certain Hot-Rolled Lead and Bismuth Carbon Steel Products from the United Kingdom, 64 FR 68316, 68318 (December 7, 1999) ("Lead and Bismuth from the United Kingdom"). The petitioners argue that the Department should reject the respondent's argument and continue to treat these expenses at issue as manufacturing costs. The petitioners maintain that the respondent's premise that the Department must adjust for expenses incurred from the original place of shipment to the place of delivery has misconstrued the meaning of the original place of shipment. The petitioners argue that the production of the merchandise sold by UFS/U-SF takes place in two locations. The petitioners maintain that the initial production of the merchandise takes place at U-SI and that this product is semi-finished product. The petitioner then contends that the semi-finished product U-SI sells to UFS/UFS is then further processed by UFS/U-FS via cutting, grinding or heat treatment operations and only then does it become finished product which UFS/U-SF sells to its unaffiliated home market customers. The petitioners conclude that, in this instance, the original place of shipment occurs after the completion of the manufacturing of the product, or shipment from UFS/U-SF to its customer. Therefore, the petitioners argue that the movement and selling expenses incurred by U-SI in selling its product to UFS/U-SF should be treated as manufacturing costs because they are part of the cost of the semi-finished inventory that is transferred to UFS/U-SF. Therefore, the petitioners request that, for the final determination, the Department affirm its preliminary determination and consider all intercompany transfer expenses as part of U-SI's cost of manufacture ("COM"). Department's Position: We agree with the petitioners and have continued to treat the movement and selling expenses between U-SI and UFS/U-SF as expenses associated with the cost of manufacturing the final product sold by U-SI's home market affiliate to its customer in the final determination. Although the respondent claims that the product sold to UFS/U-SF is finished SSB, UFS/U- SF performs additional further processing operations on the product either through subcontracting (i.e., grinding and/or heat treatment) or by altering the product itself (i.e., cutting), which effectively transforms the product prior to selling it to its customers. In this situation, U-SI has not made the claim that its sales to its affiliate, UFS/U-SF, were made at arm's length. Therefore, we have examined the issue in this case in the context of the downstream sales made by U-SI's affiliate to its customer. Because we view the downstream sale as one continuous sales process, we appropriately are treating the intercompany expenses incurred between U-SI and UFS/U-SF in making the sale to the first unaffiliated home market customer as manufacturing costs rather than selling expenses, irrespective of whether UFS/U-SF's customers request on occasion that some further processing be performed to the SSB prior to their purchase. Although the respondent maintains that further processing operations performed by UFS/U-SF are minor in nature and do not encompass all of the products UFS/U-SF sold to its customers, our analysis of certain information requested at verification indicates that with respect to one of the two channels of distribution UFS/U-SF claimed in the home market (i.e., AP ex-inventory sales), a significant portion of the merchandise sold through this channel of distribution further underwent heat treatment and grinding (see U-SI exhibit 6A of the sales verification report). The Department finds that these operations are significant, not minor, manufacturing operations which substantially altered the product from what U-SI originally sold to UFS/U-SF. In addition to these operations, a significant portion of the product sold through this channel of distribution underwent cutting. This operation also altered the product prior to its sale to the first unaffiliated home market customer. With respect to the other channel of distribution (i.e., non-AP ex-inventory sales), the respondent has indicated that on occasion, UFS/U-SF will perform cutting operations on the product sold through this channel as well. The Department also examined this channel of distribution at verification and determined that cutting, performed even on an occasional basis, still constitutes an operation which transformed the product sold between affiliated parties prior to its resale to the unaffiliated customer. Finally, with respect to the above-referenced cases cited by the respondent, the Department finds that in two of those cases (i.e., Sheet and Strip from the United Kingdom and Lead and Bismuth from the United Kingdom), the issue of how to treat the movement and selling expenses between the respondent and its affiliated reseller was not specifically raised in the same context for Department consideration in those cases. With respect to Sheet and Strip from Germany, where the Department determined that it was inappropriate to adjust the respondent's COP for certain products subjected to further processing by the affiliated home market service center, we find that this decision was made as a result of the Department incorrectly adding processing costs to certain sales for which none were incurred and/or adding multiples of those costs to the same sales where such costs were incurred (see "NSC's Processing Costs" issue contained in the public version of the Ministerial Error Memorandum titled "Allegations of Ministerial Errors - Final Determination in the Investigation of Stainless Steel Sheet and Strip in Coils from Germany," dated July 23, 1999, which has been placed on the record of this proceeding). Comment 9: Home Market Warranty Expenses The petitioners argue that the Department should not use U-SI's and UFS/U- SF's reported home market warranty expenses because the respondent continued to allocate these expenses over all of its sales instead of on a customer-specific basis as requested by the Department in its May 21, 2001, supplemental questionnaire. Furthermore, the petitioners argue that the Department noted in the sales verification report that both companies could have reported their warranty expenses on a customer-specific basis rather than over each companies' total home market sales, respectively. The respondent maintains that U-SI and UFS/U-SF reported their warranty expenses on a product-specific basis, not over all of each company's respective home market sales, based on instructions contained in the Department's February 20, 2001, questionnaire. The respondent contends that reporting this expense on a customer-specific basis fails to account for the fact that it provides warranties to all customers even though only some customers actually had filed warranty claims with the respondent during the POI. In addition, the respondent argues that attempting to relate particular warranty expenses to particular SSB purchases during the POI does not improve the accuracy of this expense calculation. Moreover, the respondent states that this expense is incurred by it after the sale has been made and simply represents claims paid during the POI. Therefore, the respondent asserts that the Department should accept its allocation methodology for reporting this expense in the final determination. Department's Position: We agree in part with the petitioners. In this case, U-SI and UFS/U-SF each separately allocated the total expense it incurred in warranty claims during the POI over all of its sales of the subject merchandise during the POI. In this case, as requested by the Department, the respondent and its affiliates have also provided the data for both the U.S. and home market sales which would allow it to isolate and calculate this expense on a customer-specific basis, rather than on a general product basis. In this case, the respondent was able to provide lists of credit notes issued to each customer for which a warranty expense was incurred and those credit notes reflected the amount of subject merchandise returned to the respondent and the amount credited to the customer for the defective merchandise. We examined this customer-specific data at verification and found it to be reliable (see U-SI exhibits 10B and 10F of the sales verification report). We have the data necessary to identify the amount of this expense which the respondent incurred on behalf of its sales to particular customers rather than on a general product basis. Therefore, we have used this data for purposes of determining the per-unit amounts for these expenses in the final determination. Moreover, since we consider these expenses to be directly related to sales made in both the U.S. and home markets, we are treating this expense as a direct selling expense in both the home and U.S. markets. Comment 10: Treatment of UFS/U-SF's Restructuring Costs as Selling Expenses The petitioners argue that the Department found at verification that a portion of expenses attributable to layoffs, relocation, severance pay and loss on fixed assets (i.e., restructuring expenses) as a result of UFS/U- SF's closing certain of its sales offices were general and administrative ("G&A") expenses and not indirect selling expenses. Therefore, the petitioners request that the Department treat these expenses as G&A expenses in the final determination. The respondent maintains that UFS/U-SF reported the expenses at issue as indirect selling expenses associated with sales office rent and salesmen's salaries incurred to sell the product in the foreign market based on instructions contained in the Department's February 20, 2001, questionnaire. Therefore, the respondent argues that the restructuring expenses at issue are related to the cessation of selling activities due to the closure of some UFS/U-SF's sales offices and reflect expenses associated with sales office rent and salesmen's salaries. Therefore, the respondent maintains that these expenses were properly reported as indirect selling expenses and should be treated as such in the final determination. Department's Position: We agree with the respondent and have continued to treat the restructuring expenses at issue as indirect selling expenses in the final determination. The Department finds that the expenses associated with closing down some of UFS/U-SF's sales offices and relocating its staff to other sales offices a selling expense, not a G&A expense, because these expenses were incurred as a result of UFS/U-SF transferring and/or relocating its selling operations to other offices still in operation. In addition, through an examination of these expenses at verification, we found no evidence that these expenses were G&A expenses (see U-SI exhibit 10D of the sales verification report). Comment 11: U.S. Credit Expense for Consignment Sales The petitioners argue that the respondent in reporting its U.S. credit expenses for US&A's consignment sales failed to consider the full credit period from the date of shipment of the merchandise to the customer to the date of payment. The petitioners argue that the respondent in its section C supplemental questionnaire response reported the U.S. credit expenses to reflect the time between the date of payment and the date of the customer's withdrawal of merchandise from inventory. The petitioners contend that this approach is wrong because it fails to consider the full credit period when U-SI was incurring a cost associated with this merchandise. The petitioners request that, for the final determination, the Department correct the respondent's U.S. credit expenses for US&A's consignment sales so that they reflect the full credit period. The respondent maintains that it calculated its U.S. credit expense in accordance with the Department's questionnaire instructions and that the Department found no discrepancies with respect to its date of sale methodology at verification. Further, the respondent argues that the expenses incurred by US&A for its consignment inventory up to the 15th day of the month are fully accounted for in its U.S. inventory carrying cost calculation. Department's Position: We disagree with the petitioners and have continued to accept U-SI's reported U.S. credit expenses for consignment sales in the final determination. As noted in the Department's questionnaire, it is the Department's practice to calculate U.S. credit expense using the number of days between date of shipment to the customer and date of payment. U-SI has properly reported its credit expenses using this time period as part of its calculation methodology. Moreover, as pointed out by U-SI, any costs incurred with respect to consignment sales being held in inventory prior to shipment to the customer are properly included in the reported U.S. inventory carrying costs. The Department found no inconsistencies with the reporting of either of these expense fields at verification (see US&A exhibits 8A through 8K, 9B, and 9K of sales verification report). Therefore, there is no need to revise the U.S. credit expense calculation for consignment sales. Comment 12: Whether to Include Freight Revenue in Calculation Formulas Used to Report Certain U.S. Discounts and Expenses As a result of a correction to its data noted by the respondent at verification, certain U.S. discount and expense fields of information (i.e., early payment discounts, warranty, technical services, and indirect selling expenses) were revised by U-SI in its November 27, 2001, U.S. sales database so that the allocation factor derived for each adjustment noted above was multiplied by the sum of gross unit price, billing adjustments, and other revenue (i.e., net U.S. price) in order to arrive at the per-unit amounts reported for each field of information. The petitioners argue that the Department should also include freight revenue in the calculation of net U.S. price for purposes of deriving the amounts reported for the fields of information noted above because like other revenue, freight revenue is also a part of the net U.S. sales figure contained in U-SI's U.S. affiliate's income statement. Moreover, the petitioners contend that freight revenue should be included in net U.S. price when calculating the fields of information noted above because freight revenue also applies to the U.S. affiliate's sales to its customers and is an integral part of the cost of the subject merchandise. The respondent argues that applying the allocation formulas to a net U.S. price (which includes freight revenue) in order to derive the per-unit amounts for the fields of information noted above would generate expenses in excess of the verified amounts actually incurred. In addition, the respondent states that its U.S. affiliate does not record freight revenue as sales revenue and that freight revenue should not therefore be included in net U.S. price when calculating the per-unit amounts for the fields of information noted above. Department's Position: We agree with the respondent. Based on our verification findings, the Department noted that freight revenue, unlike billing adjustments and other revenue, is not normally a line item on the sales invoice (see US&A exhibits 7A, 8F, and 8G of the sales verification report). In addition, freight revenue is not generally a part of the total sales revenue amount reflected in U-SI's U.S. affiliate's income statement (see also US&A exhibits 7A , 8F, and 8G of the sales verification report). Therefore, freight revenue was not included in the denominator of the allocation ratio. The Department recognizes that on occasion U-SI's U.S. affiliate includes an amount of freight revenue in the total invoice amount charged to the customer, but those instances involved small quantities of shipments and were therefore the exception rather than the rule based on an examination of this issue at verification (see verification exhibit 11A). For this reason, the Department has not included freight revenue in the calculations used to derive the per-unit amounts for the fields of information noted above. Comment 13: Treatment of French Tax Provision The respondent claims that the Department incorrectly included the French tax provision "Provision Pour Hausse de Prix" ("PHP") in the G&A ratio calculation for the preliminary determination. The respondent states that the PHP is a tax provision that reflects the capitalization of certain raw material costs in the current year and a reversal of 1994's capitalized raw material costs. The respondent argues that because these items are recorded for income tax purposes only and do not reflect true costs, these amounts should not be included in the G&A calculation. The respondent also asserts that because consumable packing costs were included in U-SI's reported COM, packing consumables should remain in the cost of goods sold used in the denominator of the G&A calculation. The petitioners argue that the PHP provision should be included in the G&A calculation because it is related to changes in the material cost valuation. The petitioners point out that the statute states at section 773(f)(1)(A) of the Act: Costs shall normally be calculated based on the records of the exporter or producer of the merchandise, if such records are kept in accordance with the generally accepted accounting principles of the exporting country (or the producing country, where appropriate) and reasonably reflect the costs associated with the production and sale of the merchandise. The petitioners claim that the PHP provision is acceptable under French generally acceptable accounting provisions and should therefore be included in the COM. Department's Position: We agree with petitioners and the respondent in part. The PHP provision made by U-SI to adjust for price increases in raw materials should be included in the reported costs. Consistent with section 773(f)(1)(A) of the Act, it is the Department's practice to rely upon a company's normal books and records when they are prepared in accordance with the country's generally accepted accounting principles ("GAAP") and they reasonably reflect the cost of producing and selling the subject merchandise. We noted that in its normal books and records, U-SI included the net price adjustment provision for its raw materials. However, we found that these adjustments are not reflected in U-SI's reported costs. Additionally, because raw material inventories are inputs into the cost of manufacturing the merchandise under consideration, any write-downs of these amounts should be included in determining the reported costs. See Notice of Final Determination of Sales Less Than Fair Value: Stainless Steel Wire Rod from Italy, 63 FR 40422, 40430 (July 29, 1998). Because the PHP provision relates to the general operations of the company as a whole and cannot be segregated between subject and non-subject merchandise we have included the net provision in the calculation of the G&A expense ratio. We agree with the respondent that because consumable packing costs were included in U-SI's reported COM, the packing consumables should remain in the cost of goods sold used in the denominator of the G&A calculation. Recommendation Based on our analysis of the comments received, we recommend adopting all of the above positions. If these recommendations are accepted, we will publish the final determination and the final weighted-average dumping margin for the investigated firm in the Federal Register. Agree____ Disagree____ ______________________ Faryar Shirzad Assistant Secretary for Import Administration ______________________ (Date) _________________________________________________________________________ footnotes: 1. Antidumping Duty Investigations of Stainless Steel Bar from France and Italy. 2. See Techsnabexport, Ltd. v. United States, 802 F. Supp. 469, 472 (CIT 1992) (Finding that antidumping duty "investigations are to focus on merchandise from particular countries."). 3. See 19 CFR 351.202(b)(6); section 771 of the Act. 4. See European Communities - Anti-dumping measures on imports of cotton- type bed linen from India, DS141/AB/R (2001) ("Bed Linen from India"). 5. We have continued to organize the common SSB selling functions into four major categories: (1) sales process and marketing support; (2) freight and delivery; (3) inventory and warehousing; and (4) quality assurance/warranty services. 6. These operations included the following: (1) testing and certifications; (2) upgrading (i.e., heat treatment, machining, drilling, grinding); and (3) "other special" operations (i.e., special conditioning, cutting, marking, and chamfering SSB).