(65 FR 13943, March 15, 2000) A-201-802 ARP 97-98 Public Document G1O3: SMH MEMORANDUM TO: Robert S. LaRussa Assistant Secretary for Import Administration FROM: Richard W. Moreland Deputy Assistant Secretary For Import Administration SUBJECT: Issues and Decision Memo for the Administrative Review of Gray Portland Cement and Clinker From Mexico -- August 31, 1997 through July 31, 1998 Summary We have analyzed the comments and rebuttals of interested parties in the 97/98 administrative review of the antidumping duty order covering gray portland cement and clinker from Mexico. As a result of our analysis, we have made changes, including corrections of certain inadvertent programming and clerical errors, in the margin calculations. 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 administrative review for which we received comments and rebuttals by parties: 1. Revocation 2. As Invoiced vs. As Produced 3. Ordinary Course of Trade 4. Level of Trade 5. Constructed Export Price Calculation 6. Regional Assessment 7. Bag vs. Bulk 8. Difference-in-Merchandise Calculation 9. Sales-Below-Cost Test 10. Special Cement 11. Assessment-Rate Calculation 12. Adjustments a. Rebates b. Freight c. Advertising d. Early-Payment Discounts e. Credit Expenses f. Other Adjustments 13. Financing Cash Deposits 14. Duty Absorption 15. PROMEXMA Sales 16. Contrucentro’s Employee Sales 17. Further-Manufactured Sales 18. Ministerial Errors a. Model Matching b. CDC’s Employee Sales c. U.S. Direct Selling Expenses Background On September 8, 1999, the Department of Commerce (the Department) published the preliminary results of administrative review of the antidumping duty order on gray portland cement and clinker from Mexico (Gray Portland Cement and Clinker From Mexico; Preliminary Results of Antidumping Administrative Review and Extension of Final Results of Administrative Review, 64 FR 48778 (preliminary results)). The merchandise covered by this order is gray portland cement and clinker. The period of review (POR) is August 1, 1997, through July 31, 1998. We invited parties to comment on our preliminary results of review. Discussion of the Issues 1.Revocation Comment 1: CEMEX, S.A. de C.V. (CEMEX), and Cementos de Chihuahua, S.A. de C.V. (CDC), argue that the Department must terminate this review and revoke the underlying antidumping duty order. CEMEX contends that, at the time of the initiation of the original Less- Than-Fair-Value (LTFV) investigation (October 16, 1989), the Department assumed that the petition was filed "on behalf of" a regional industry without measuring whether a majority of the industry actually supported the request. The Department should have done so, CEMEX argues, because a General Agreement on Tariffs and Trade (GATT) panel decided in July 1992 that the 1980 antidumping code required that an antidumping petition filed "on behalf of"' an industry must be supported by an appropriate majority of the industry and that such support must be ascertained prior to initiating an investigation. According to CEMEX, the panel's decision applies to the instant administrative review for two reasons. First, it contends, the Antidumping Agreement resulting from the Uruguay Round negotiations adopted the requirement of industry support articulated by the GATT panel. CEMEX asserts U.S. law incorporated the new standing requirements contained in the Antidumping Agreement, citing section 732(c)(4)(C) of the Tariff Act of 1930 (the Act). Second, CEMEX asserts, even if the pre-Uruguay Round Agreements Act (URAA) antidumping law applies, the antidumping statute that was in effect in 1989 did not define the term "on behalf of." CEMEX argues that the Department is compelled by the decision in Murray v. Schooner Charming Betsy, 6 U.S. 64, 2 Cranch 64 (1804), to reinterpret U.S. law in accordance with the international obligations of the United States. Based on the above arguments, CEMEX asserts that the Department is therefore required in this review to revisit the issue of initiation in the original LTFV investigation. According to CDC, the plain language of section 771(4)(C) of the Act requires petitions in regional-industry cases to be filed on behalf of the producers that account for "all, or almost all, of the production in the region." Since the antidumping duty order covering cement from Mexico was based on a petition that was unsupported by producers accounting for all or almost all of the region's production, CDC asserts, the Department issued the order in violation of U.S. law. CDC disputes the Department’s assertion in the 96/97 review that, because the issue of petitioner’s standing was not challenged in court, the issue cannot be revisited in the context of an administrative review. CDC argues that the lack of standing to file an antidumping duty petition is a "jurisdictional" defect which parties may raise at any time. CDC contends that the Department has the authority to revoke an order that never had the requisite level of industry support, citing Zenith Electronics Corp. v. United States, 872 F. Supp. 992 (CIT 1994) (Zenith Electronics), Gilmore Steel Corp. v. United States, 585 F. Supp. 670 (CIT 1984) (Gilmore Steel), and Oregon Steel Mills, Inc. v. United States, 862 F.2d 1541 (Fed. Cir. 1988) (Oregon Steel Mills). CDC also contends that the Department dismissed the standing issue raised in the 96/97 review and supported its decision based on cases that pertained to national rather than regional- industry cases. CDC points out that the definition of domestic industry in regional cases imposes a different standard for the support that is required when making the "on behalf of" determination. The Southern Tier Cement Committee (the petitioner) comments that CEMEX and CDC have raised the argument that the Department must terminate the review and revoke the underlying antidumping duty order on the ground that the Department initiated the antidumping investigation improperly in 1989 in the reviews covering the 92/93 through 96/97 periods. The petitioner argues that the North American Free Trade Agreement (NAFTA) binational panels reviewing the final results of the 92/93 and 94/95 reviews both ruled unanimously against CEMEX’s position on a variety of grounds. The petitioner asserts that CEMEX’s and CDC’s claim that the Department must revoke the antidumping duty order and terminate this administrative review has no legal foundation. The petitioner argues that CEMEX’s and CDC’s claim is barred by the statute of limitations which required any appeal of the decision to initiate the antidumping investigation to be filed within 30 days of the publication of the antidumping duty order. The petitioner claims that CEMEX and CDC did not exhaust available administrative remedies when they did not raise the issue before the Department in the original investigation. Furthermore, the petitioner contends that the respondents did not raise the issue in an appeal to the Court of International Trade from the Department’s final determination in the original investigation. The petitioner also cites the Department’s 95/96 review final results (Gray Portland Cement and Clinker From Mexico, 63 FR 12765 (March 16, 1998)) and 96/97 review final results (Gray Portland Cement and Clinker From Mexico, 64 FR 13149 (March 17, 1999)) in which the Department commented that panel reports under the 1947 GATT were not self-executing and had no legal effect under U.S. law and that neither the 1947 GATT nor the 1979 GATT antidumping code obligated the United States to establish industry support in regional-industry cases. The petitioner contends that the Department lacks authority under the statute to rescind its decision to initiate or to re-examine the issue of industry support in a review. Citing Suramerica de Aleaciones Laminda, C.A. v. United States, 966 F.2d 660 (Fed. Cir. 1992) (Suramerica), and the 95/96 final results, the petitioner asserts that courts have affirmed the Department’s presumption of industry support. Department's Position: There are no circumstances in this review that would cause the Department to reevaluate industry support underlying the petition. Nor is there any basis, statutory or otherwise, to revoke the order based on lack of industry support at the time of initiation. Furthermore, as we stated in our last final results of review, we have no obligation at this time to determine whether the petition was filed by or on behalf of the industry. Moreover, the URAA amended the statute by prohibiting the Department from revising the issue of industry support once the Department has initiated an LTFV investigation. See section 732(c)(4)(E) of the Act. Neither the 1947 GATT nor the 1979 GATT Antidumping Code obligated the United States to establish affirmatively prior to the initiation of a regional-industry case that all or almost all of the producers in the region supported the petition. Neither agreement suggested that the standing requirements in regional-industry cases were any more rigorous than the standing requirements in national-industry cases. Furthermore, GATT panel reports, such as the report issued in 1992, had no legal effect or formal status unless and until they were adopted by the GATT Council or, in the case of antidumping measures, the GATT Antidumping Code Committee. This followed from the fact that the 1947 GATT operated, throughout its history, on the basis of consensus for purposes of decision-making in general and the resolution of disputes in particular. It is undisputed in the present case that the Antidumping Code Committee never adopted the GATT panel report. Thus, the recommendations contained in the report were never binding, did not impose any international obligations upon the United States, and did not trigger the rule of statutory construction set forth in Murray v. Schooner Charming Betsy. As we pointed out in previous determinations, the object of the respondents’ comments is not the preliminary results of this review. Rather, the respondents challenge the initiation of the original LTFV investigation, an event which occurred almost ten years ago and over five years before the effective date of the URAA. The time to voice such objections before the Department was during the investigation. Instead, CEMEX and CDC, as well as the other Mexican cement producer that participated in the original investigation (Apasco, S.A. de C.V.), did not raise this argument before the Department. See Final Determination of Sales at Less Than Fair Value; Gray Portland Cement and Clinker From Mexico, 55 FR 29244 (July 18, 1990) (Original LTFV Investigation). Moreover, neither CEMEX nor any other party appealed the agency's final affirmative LTFV determination (including the decision to initiate) to the appropriate court, and the deadline for doing so has long expired. See section 516A of the Act. Therefore, even if the Department, of its own volition, were to reinterpret U.S. law in light of the 1992 GATT panel report, it lacks the legal authority in this review to revoke the order or otherwise rescind the initiation of the underlying investigation based on lack of industry support at the time of initiation. See also Gray Portland Cement and Clinker from Mexico; Final Results of Antidumping Duty Administrative Review, 62 FR 17581 (April 10,1997) (93/94 Review Final Results); Gray Portland Cement and Clinker from Mexico; Final Results of Antidumping Duty Administrative Review, 62 FR 17148 (April 9, 1997) (94/95 Review Final Results); Gray Portland Cement and Clinker from Mexico; Final Results of Antidumping Duty Administrative Review, 63 FR 12764 (March 16, 1998) (95/96 Review Final Results); Gray Portland Cement and Clinker from Mexico; Final Results of Antidumping Duty Administrative Review, 64 FR 13148 (March 17, 1999) (96/97 Review Final Results). Of the cases cited by CEMEX and CDC, none of them supports the argument that the Department has the authority, in an administrative review under section 751(a) of the Act, to reach back almost ten years and reexamine the issue of industry support for the original petition. In Gilmore Steel, the plaintiff contended that the Department lacked the authority to rescind the investigation based upon insufficient industry support for the petition after the 20-day period established in section 732(c) of the Act had elapsed. 585 F. Supp. at 673. In Zenith Electronics, the plaintiff alleged that the petitioner was no longer a domestic "interested party" with standing to request an administrative review. 872 F. Supp. at 994. Nothing in Zenith Electronics or Gilmore Steel supports CDC's argument that a party may challenge industry support for a petition almost ten years after the fact in the context of an administrative review under section 751(a) of the Act. The case cited by CDC, Oregon Steel Mills, involved a challenge to the Department's authority to revoke an antidumping duty order based upon new facts, i.e., the industry's affirmative expression of no further support for the antidumping duty order, not upon a reexamination of the facts as they existed during the original LTFV investigation. The Federal Circuit (CAFC) held that it was lawful for the Department, in the context of a "changed circumstances" review pursuant to section 751(b) of the Act, to revoke an order over the objection of one member of the industry. 862 F.2d at 1544-46. The court did not state that industry support for an order must be affirmatively established throughout the life of an order. Indeed, the court went to lengths to explain that it was not ruling on the claim that "loss of industry support for an existing order creates a 'jurisdictional defect.'" Id. at 1545 n. 4. As courts explained subsequently, the holding in Oregon Steel Mills is limited to the proposition that the Department may, but need not, revoke an order when presented with record evidence which demonstrates a lack of industry support for the continuation of the order. See, e.g., Suramerica de Aleaciones Laminda, C.A. v. United States, 966 F.2d 660, 666 (Fed. Cir. 1992) (Suramerica), and Citrosuco Paulista, S.A. v. United States, 704 F. Supp. 1075, 1085 (CIT 1988) (Citrosuco). Finally, as we mentioned in the final results of the last five administrative reviews, numerous courts upheld our practice under the pre-URAA statute of assuming, in the absence of evidence to the contrary, that a petition filed on behalf of a regional or national industry is supported by that industry. See, e.g., NTN Bearing Corp. v. United States, 757 F. Supp. 1425, 1427-30 (CIT 1991), Citrosuco, 704 F. Supp. at 1085, and Comeau Seafoods v. United States, 724 F. Supp. 1407, 1410-12 (CIT 1989). Indeed, this issue raised by CEMEX and CDC was before the Federal Circuit in the Suramerica case (966 F.2d at 665, 667). In Suramerica the plaintiffs challenged the Department's interpretation of the phrase "on behalf of" which applied to both national-industry and regional-industry cases. In affirming the Department's practice, the court observed that the phrase "on behalf of" was not defined in the statute. Id. at 666- 667. The statute was, in fact, open "to several possible interpretations." In the opinion of the court, the Department's practice with regard to standing and industry support for a petition reflected a reasonable "middle position." 966 F.2d at 667. While there was a gap in the statute, the court stated, "Congress did make [one thing] clear--Commerce has broad discretion in deciding when to pursue an investigation, and when to terminate one." Id. Although not considered legal precedent, our decision not to revoke the underlying order has been upheld by NAFTA panels. See, e.g., NAFTA Panel Review (94/95 Review) Opinion and Order of the Panel (June 18, 1999). Therefore, we reject respondents’ arguments that we lack the authority to assess antidumping duties pursuant to these final results of review and that we must revoke the underlying antidumping duty order. 2. As Invoiced vs. As Produced Comment 2: The petitioner contends that the Department erred by matching merchandise on the basis of the ASTM cement type "as produced" rather than matching, as it had done in the original investigation and in the first five administrative reviews, on an "as invoiced" basis. The petitioner argues that matching merchandise on an "as produced" basis is inconsistent with the Department’s matching practices of commercially significant characteristics. The petitioner asserts, however, that if the Department continues to find home-market sales of cement produced as Type V and Type V LA outside the ordinary course of trade, then it does not object to the Department’s methodology of matching cement on an "as produced" basis because this methodology will have no impact on the dumping margin. CEMEX responds that the Department matched identical merchandise properly on the basis of the ASTM specification to which the cement was produced. CEMEX argues that matching merchandise according to how it was sold does not meet the statutory requirement of section 771(16) of the Act, which requires that the "foreign like product" include only merchandise sold in the comparison market that is physically identical with the merchandise produced for sale to the United States. CEMEX argues that, as the Department recognized in the 95/96 and 96/97 review final results, the statute compels the Department to base normal value on its sales of cement that meet the customers' specifications physically. CEMEX notes that the petitioner raises the same arguments already rejected by the Department in the 95/96 and 96/97 reviews. CEMEX points out that, in this case, no party disputes that product characteristics of cement are determined on the highest ASTM specifications that it meets. Therefore, CEMEX concludes, the Department's identification of home- market cement sales pursuant to the highest ASTM specifications to which the cement is produced continues to be in accordance with law. Department’s Position: For these final results of review, we found that home-market sales of Type V and Type V LA cement are outside the ordinary course of trade. Thus, regardless of whether we adopt an "as produced " or "as sold" methodology, we would be comparing U.S. sales to sales of Type I cement in Mexico. Notwithstanding this fact, we found, as we did in the 96/97 administrative review of this order, that CEMEX continues to produce cement meeting the physical requirements of one type of cement and sells that cement as another cement type. Therefore, we have maintained our practice of matching the foreign like product that is physically identical to the merchandise produced for sales in the U.S. market. As our position has not changed since the 96/97 review, we adopt the discussion in those final results with respect to this issue. See 64 FR at 13154. 3. Ordinary Course of Trade Comment 3: CEMEX argues that its home-market sales of cement produced as Type V LA at its Hermosillo plants were in the ordinary course of trade and should be used in the calculation of normal value. CEMEX challenges the validity of the factors the Department used in its analysis to determine whether sales of cement produced as Type V LA ware made outside the ordinary course of trade. CEMEX argues that the Department applied selected factors in performing its ordinary-course-of-trade analysis and that the Department's analysis was not supported by substantial evidence. CEMEX contends further that sales produced as Type V LA and specifically those invoiced as Type II LA and Type V LA were made pursuant to a bona fide home- market demand for those types of cement, that the merchandise sold was not obsolete or of second quality, that it was sold for its intended purposes, and that there were no special sales arrangements for sales of these cement types. Concerning the sales-volume factor, CEMEX contends that the Department's analysis relies incorrectly on the volume of the sales at issue relative to sales of Type I cement and that the volume of the sales at issue was significant in absolute terms and pursuant to a bona fide demand. CEMEX also argues that judicial precedent and prior administrative practice establish that relatively low sales volume signifies sales outside the ordinary course of trade only when coupled with an absence of a bona fide demand in the home market, which does not exist in this case. With respect to shipping arrangements, CEMEX contends that the Department should focus on the actual terms of delivery for the sales at issue, identical to those of Type I customers, rather than the geographic distance. CEMEX points out that the distance to the customers is a geographic fact rather than a condition or practice of sale as required by section 771(15) of the Act. CEMEX maintains that in no other cases has the Department relied on shipping distances in determining whether sales were outside the ordinary course of trade. Furthermore, CEMEX argues, if the Department continues to consider shipping distance in its analysis, it should do so on an individual- sale basis. CEMEX argues that the Department's reliance on the low profitability of the sales at issue ignores the fact that the profit levels on these sales, though not as high as sales invoiced as Type I, are substantial and significant in absolute terms. CEMEX points out that the Department’s regulations suggest that very high disparities or relatively negative profit rates, as evidenced in the 91/92 administrative review, would demonstrate sales made outside the ordinary course of trade, which is not the situation in this review. CEMEX contends further that the Department's reliance on the small number and type of customers for cement produced as Type V LA, and specifically with respect to those invoiced as Type II LA and Type V LA, is improper because such evidence generally reflects sales outside the ordinary course of trade in cases of sales of export overrun and off-specification sales, rather than when sales are made to a bona fide home market. Moreover, CEMEX argues that the Department should focus its quantity-per-transaction analysis on bulk- to-bulk comparisons rather than bulk-to-bag comparisons, in accordance with the NAFTA panel determination pursuant to the 94/95 administrative review. CEMEX proffers that this type of analysis would demonstrate that the average difference in sales quantity per transaction between cement produced as Type I and cement produced as Type V LA was not significant enough to warrant a determination that its sales of cement produced as Type V LA were outside the ordinary course of trade. With regard to the historical trade-trends factor, CEMEX argues that a period of thirteen years of domestic sales of these products, before and after the imposition of the order, constitutes a "reasonable period of time" within the meaning of section 771(15) of the Act. Concerning Type V cement, CEMEX also argues that the Department's preliminary results are factually incorrect because those results did not appreciate the prior history of this case. Specifically, CEMEX states that the Department did not acknowledge that, when conducting the 91/92 administrative review, the Department verified that Tolteca, a CEMEX subsidiary whose production is subject to this review, has made continuous sales in the home market of Type V cement since 1964. Furthermore, CEMEX points out that it began producing and selling Type V LA cement five years before the issuance of the order, thereby satisfying the conditions and practices which, for a reasonable period of time prior to exportation of subject merchandise, have been normal in the trade under consideration. Thus, CEMEX contends, historical sales patterns in the home market of these cement types meet the statutory definition of sales made within the ordinary course of trade. Finally, CEMEX maintains that the Department’s reliance on facts available to infer that sales of Type V LA cement that was produced as Type V LA exhibit a promotional quality not present for sales of Type I cement is factually incorrect. CEMEX argues that record evidence demonstrates that the promotional qualities of cement sold as Type V LA and Type I cement are identical. CEMEX challenges the overall relevance of promotional quality as a factor in an ordinary- course-of-trade inquiry and argues that there is no judicial or Departmental precedent that has referred to this factor in any other ordinary-course-of-trade analysis. The petitioner states that, for the preliminary results of review, the Department determined correctly that CEMEX’s home-market sales of cement produced as Type V and Type V LA (sold as Type I, Type II LA, or Type V LA cement) were outside the ordinary course of trade.1 The petitioner asserts that the Department should not evaluate just one factor in isolation but should consider all the circumstances particular to the sales in question to determine whether the sales reflect the conditions and practices which, for a reasonable time prior to the exportation of the subject merchandise, have been normal in the trade under consideration. The petitioner argues that the Department’s preliminary ordinary-course-of trade determination is consistent with its determination on the same issue in the final results of several previous administrative reviews. The petitioner maintains that, in the instant administrative review, the record evidence demonstrates that home-market sales of cement produced as Type V and Type V LA continue to be outside the ordinary course of trade. With regard to the changes in production and shipping arrangements, the petitioner argues that the fact that CEMEX made such a radical change in its production and distribution of these products since the issuance of the antidumping duty order for the purpose of bringing down the margin is strong evidence that sales of those products are unusual and, thus, outside the ordinary course of trade. Concerning the absorption of freight costs, the petitioner states that, prior to the issuance of the antidumping duty order, CEMEX passed on the cost of freight for Type II and Type V sales to its customers in higher prices. After the order, the petitioner asserts, CEMEX consolidated all Type II and Type V cement production at Hermosillo which increased the shipping distance to customers in central Mexico significantly, consequently increasing its freight expense. The petitioner contends that, since that time, CEMEX absorbed the high freight costs and continued to do so during the instant review. With regard to the sales-volume factor, the petitioner contends that CEMEX restricted its home-market sales of Type II cement after the antidumping duty order by ceasing to offer Type II for sale as a general-purpose cement and selling it only to those customers demonstrating a need for low-alkali Type II cement. The petitioner asserts that CEMEX offset the much lower prices on Type II LA cement resulting from the increased transportation costs (incurred due to shipping small quantities long distances to customers) with higher prices on Type I cement. With respect to specialty products sold to a "niche" market, the petitioner asserts that this situation reflects CEMEX’s restriction of sales of Type II after the antidumping duty order, which it had sold previously as general-purpose cement. Moreover, the petitioner contends the fact that Type II and Type V cement are specialty products sold to a niche market is demonstrated further by the fact that, in the instant review period, CDC did not produce Type V LA cement and, although it has been producing Type II LA cement for export to the United States, it only began selling a limited quantity of Type II LA in the home market in December 1997 to a related customer. Concerning the percentage sold of cement produced as Type V LA and sold as Type II LA and Type V LA compared with CEMEX’s total home- market sales of cement, the petitioner points out that, since the 90/91 review, the Department has found that these sales accounted for a "minuscule percentage of CEMEX’s total sales of cement". The petitioner contends that the same situation applies for the instant review. The petitioner argues that CEMEX’s home-market sales of Type II LA and Type V LA cement produced as Type V LA also account for only a small percentage of CEMEX’s production of Type V LA produced cement. The petitioner asserts that the extremely limited home-market demand for sales of Type II LA and Type V LA cement compared to its export market for those types of cement is similar to sales of "overrun" merchandise which the Department has found in other cases to be outside the ordinary course of trade. In addition, the petitioner points out the unusual nature of the number, type, and geographic location of customers for CEMEX’s cement produced as Type V LA and sold as Type II LA and Type V LA. The petitioner asserts that information on the current record supports the Department’s conclusion that the number and types of customers for sales of Type II LA cement are very small and very different from those for other home-market sales. The petitioner refutes the argument put forward by CEMEX that the number of customers is relevant only where the merchandise is not sold to a bona fide home market, stating that the Department has no such rule of thumb and evaluates each case according to its own circumstances. The petitioner asserts that there is no reason for the Department to distinguish between bulk and bagged sales in its analysis of the number and type of customers to which CEMEX sold cement produced as Type V LA and sold as Type II LA and Type V LA cement. The petitioner asserts that the only difference between bulk and bagged cement is the manner in which the sales are packed. The petitioner also asserts that there was a difference in the manner in which CEMEX granted discounts and rebates on sales of cement produced as Type V LA and sold as Type II LA and Type V LA cement compared to other home-market sales. With regard to CEMEX’s profit on cement produced as Type V LA and sold as Type II LA and Type V LA sales, the petitioner asserts that the revised data CEMEX submitted indicates a large difference in profit rates and not a "minor" or "insignificant" difference as alleged by CEMEX. In addition to cement produced as Type V LA and sold as Type V LA and Type II LA, the petitioner contends that CEMEX’s home-market sales of cement produced as Type V and Type V LA but sold as Type I are also outside the ordinary course of trade. The petitioner states that many of the factors which were relevant in determining whether CEMEX’s cement produced as Type V LA and sold as Type II LA and Type V LA cement are outside the ordinary course of trade are also relevant in determining whether sales of Type I cement produced as Type V and Type V LA cement are outside the ordinary course of trade. In addition to those factors, the petitioner argues that CEMEX consistently produced Type V and Type V LA cement and sold it as Type I cement at the two Hermosillo plants, even though no legitimate demand in Mexico exists for Type V and Type V LA cement. The petitioner states that, while the circumstances with respect to CEMEX’s Type V LA cement sold as Type I from Hermosillo have not changed materially since the 96/97 review, two other aspects of CEMEX’s home-market sales of cement produced as Type V cement and sold as Type I cement have changed. The petitioner discusses several reasons why these factual changes do not alter the fact that the sales are outside the ordinary course of trade. However, the petitioner comments that CEMEX has reported producing Type V LA cement at the Hidalgo plant and selling it as Type I cement, and that these sales were made under temporary and highly unusual circumstances. Therefore, such sales were outside the ordinary course of trade. Thus, the petitioner maintains, only the two Hermosillo plants produced Type V LA cement that was sold consistently as Type I cement. The petitioner points out further that CEMEX has also reported that, in January 1998, the Yaqui Plant began producing Type V cement that did not meet the ASTM standard for low-alkali cement and which it sold in the home market as Type I cement. The petitioner argues that Type V produced cement sold as Type I is outside the ordinary course of trade for many of the same reasons that Type V LA produced cement sold as Type I cement is outside the ordinary course of trade. The petitioner adds additional factors which it contends support the Department’s analysis on this issue: CEMEX is silent with respect to Type I cement sales which were produced as Type V cement at the Hidalgo plant; lack of demand in Mexico for Type V cement which does not meet the optional low-alkali specification; Type V and Type V LA cement sold as Type I accounts for a very small percentage of CEMEX’s home-market sales; the relatively small number and type of customers for cement produced as Type V and Type V LA cement sold and sold as Type I cement. Department's Position: In the instant administrative review, CEMEX produced Type V LA cement at its Campana, Yaqui (located in the Hermosillo region), and Hidalgo plants in Mexico. It sold this cement as either Type I, Type II LA, or Type V LA cement. It also produced Type V cement which it sold only as Type I cement. We have found that home-market sales of cement produced as Type I and sold as Type I cement are within the ordinary course of trade. Our ordinary-course-of trade analysis is focused on CEMEX’s cement that was produced as Type V LA or Type V and sold as Type V LA, Type II LA, or Type I in Mexico. As we stated in our final ordinary-course-of-trade memorandum, we determined that CEMEX’s sales of cement produced as Type V and sold as Type I were outside the ordinary course of trade. See the Ordinary-Course-of-Trade Memorandum for the Final Results of Review from Davina Hashmi to Laurie Parkhill, dated March 6, 2000, available in the Department’s Central Record Unit, room B-099. We made this determination based on the small sales volume and quantity of this cement type, coupled with the lack of demand in the home market and the fact that CEMEX produced Type V cement with the understanding that it would be sold as another cement type, Type I. In its case brief, CEMEX did not address our finding that cement produced as Type V and sold as Type I was outside the ordinary course of trade. Therefore, the following ordinary-course-of-trade discussion does not pertain to cement produced as Type V. For further discussion of cement produced as Type V, see the final ordinary-course-of-trade memorandum, dated March 6, 2000. CEMEX produced Type V LA cement at its Hidalgo and Hermosillo plants. However, CEMEX produced Type V LA cement at the Hidalgo plant inadvertently as the result of an effort to produce another cement type. Since the production of this cement type was unintentional and unusual and it was sold only as Type I cement, we found sales of this cement to be outside the ordinary course of trade. Therefore, the analysis contained herein focuses on cement produced as Type V LA at the Hermosillo plants, irrespective of how sold (i.e., Type V LA, Type II LA, or Type I). Section 773(a)(1)(B) of the Act states, in part, that normal value is the "price at which the foreign like product is first sold (or, in absence of a sale, offered for sale) for consumption in the exporting country, in the usual commercial quantities and in the ordinary course of trade." The term "ordinary course of trade" is defined as "the conditions and practices which, for a reasonable time prior to the exportation of the subject merchandise, have been normal in the trade under consideration with respect to merchandise of the same class or kind." The Statement of Administrative Action (SAA) which accompanied the passage of the URAA clarifies this portion of the statute further when it states: Commerce may consider other types of sales or transactions to be outside the ordinary course of trade when such sales or transactions have characteristics that are not ordinary as compared to sales or transactions generally made in the same market. SAA, at 834. Thus, the statute and the SAA are clear that a determination of whether sales (other than those addressed specifically in section 771(15) of the Act) are in the ordinary course of trade must be based on an analysis comparing the sales in question with sales of merchandise of the same class or kind generally made in the home market (i.e., the Department must consider whether certain sales of cement in Mexico are ordinary in comparison with other sales of cement in Mexico). The purpose of the ordinary-course-of-trade provision "is to prevent dumping margins from being based on sales which are not representative" of the home market. See NTN Bearing Corp. of America v. United States, Slip Op. 99-71 at 21-22 (CIT, July 24, 1999) (NTN Bearing Corp.). Congress has not specified any criteria that the agency should use in determining the appropriate "conditions and practices". Thus, the Department, in its discretion, chooses how best to analyze the many factors involved in a determination of whether sales are made within the ordinary course of trade. Id. In the instant review, the Department's decision to exclude cement produced as Type V LA from the calculation of normal value centered on the unusual nature and characteristics of these sales compared to CEMEX's sales of other types of gray portland cement, which consisted primarily of Type I cement. The Department's ordinary-course-of- trade analysis is far-reaching. The agency must evaluate not just "one factor taken in isolation but rather all the circumstances particular to the sales in question" (Murata Mfg. Co. v. United States, 820 F. Supp. 603, 607 (CIT 1993) (quoting Certain Welded Carbon Steel Standard Pipes and Tubes from India, Final Results of Antidumping Duty Administrative Review, 56 FR 64753, 64755 (December 12, 1991) (Pipes and Tubes from India); see also NTN Bearing Corp., supra). This broad approach recognizes that each company has its own conditions and practices particular to its trade. In short, the Department examines the totality of the facts in each case to determine if sales are being made for "unusual reasons" or under "unusual circumstances." See Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et. al; Final Results of Antidumping Duty Administrative Review, 64 FR 35590, 35620- 21 (July 1, 1999), and Electrolytic Manganese Dioxide from Japan; Final Results of Antidumping Duty Administrative Review, 58 FR 28551, 28552 (May 14, 1993). We have revisited our ordinary-course-of-trade analysis for these final results of review. For this analysis, we concentrated on the sales-transaction volumes, quantity or tonnage sold, shipping distances, freight costs, and average profit as factors to support our finding that, as we determined for the preliminary results of review, home-market sales of cement produced as Type V LA are outside the ordinary course of trade. See the Ordinary-Course-of-Trade Memorandum. Based on this analysis, we found that, during the period of review, as in previous reviews, CEMEX made very few home-market sales of cement produced as Type V LA compared to its sales of cement produced as Type I cement. We also found that the tonnage CEMEX sold in Mexico of cement produced as Type V LA was very small compared to the tonnage of cement produced as Type I. Thus, it was evident from this analysis alone that the sales of cement produced as Type V LA in comparison to cement produced as Type I constituted a practice that was unusual in the context of CEMEX's sales in Mexico. With regard to shipping distances, the normal practice for CEMEX is to ship cement, a heavy material, over relatively short distances. The majority of CEMEX's sales of cement in Mexico were shipped less than 150 miles from the production site and, during the period of review, shipments of cement produced as Type I cement conformed to this pattern. Shipments of cement produced as Type V LA and sold as Type II LA and Type V LA, however, traveled over vastly greater distances. We found that freight costs for cement produced as Type V LA, and specifically those with regard to sales of Type V LA and Type II LA, were considerably higher than the freight costs for cement produced as Type I and sold as Type I. We also found that CEMEX's profit on cement produced as Type V LA during the period of review was small compared to profit it earned on cement produced as Type I and Type V. These factors indicate that CEMEX is absorbing the high freight costs incurred to transport its cement produced as Type V LA (particularly with regard to cement produced as Type V LA that was sold as Type V LA and Type II LA) over great distances to its customers located more than 1200 miles from the production site. For the years preceding the antidumping order, it was CEMEX's normal business practice to pass along the cost of pre-sale freight to purchasers of its Type II cement. Therefore, we find it an "unusual circumstance" for CEMEX to absorb freight costs on its cement produced as Type V LA compared with cement produced as Type I. For further discussion on the factors with reference to proprietary data, see the final ordinary- course-of-trade memorandum from Davina Hashmi to Laurie Parkhill, dated March 6, 2000. We disagree with CEMEX's contention that our analysis relied incorrectly on the relative sales volume of cement produced as Type V LA compared with cement produced and sold as Type I cement. The relative sales volume has often played an important role as one of the factors in our overall analysis of whether sales are within the ordinary course of trade. See Furfuryl Alcohol From Thailand; Final Results of Administrative Review, 60 FR 22557, 22559 (May 8, 1995) (subject sales compared to "standard" volumes). In previous cases we have stated that, "while the number of sales or volume sold are not in and of themselves definitive factors in determining whether the sales in question are in the ordinary course of trade, this factor ‘coupled’ with others may result in a finding that ‘sales are not in the ordinary course of trade’" (Pipes and Tubes from India; Final Results of Administrative Review, 56 FR at 64755 (emphasis added); Murata, 820 F. Supp. at 607 (quoting Pipes and Tubes from India); Certain Corrosion-Resistant Carbon Steel Flat Products From Australia; Final Results of Administrative Review, 61 FR 14049, 14051 (March 29, 1996) (relative sales volume coupled with other factors led to conclusion that sales were within the ordinary course of trade)). Thus, we are not questioning whether the sales of cement produced as Type V LA constitute bona fide sales, as CEMEX suggests we should. Rather, we question whether the sales of cement produced as Type V LA were made in the usual course of business. Given the number of sales of cement produced as Type V LA made in Mexico during the instant period of review relative to the number of sales of Type I cement, it is evident that CEMEX's sales of the cement in question was, in fact, unusual and not representative of its normal sales pattern. This sales pattern demonstrates that CEMEX's main business in Mexico is Type I cement, not Type V LA cement. In addition, the CAFC in CEMEX, S.A. v. United States, 133 F.3d 897 (Fed. Cir. 1998) (CEMEX v. U.S.), affirmed our determination that CEMEX's sales in the home market of Types II and V were outside the ordinary course of trade based, in part, on relative sales volume. We have excluded transactions from the calculation of normal value based upon low or negligible sales volume in other cases as well. See Tapered Roller Bearings and Parts Thereof, Finished and Unfinished, From Japan; Final Results of Administrative Review, 52 FR 30700, 30704 (August 17, 1987), and New Minivans from Japan; Final Results of Administrative Review, 57 FR 43, 46 (January 2, 1992). Thus, it has been our long-standing practice to consider the relative sales volume, along with other factors, in our ordinary-course-of-trade analysis. We also disagree with CEMEX's claim that, instead of considering shipping distances and freight costs, we should focus on shipping terms and practices. We have taken into consideration both the shipping distances and shipping terms and practices for these final results of review. With respect to shipping distances, we found that the normal practice in Mexico is to ship cement over relatively short distances. CEMEX itself has conceded that shipping long distances by rail or truck is not ordinary in the cement industry. As we noted earlier, the majority of CEMEX's cement shipments in Mexico of Type I cement cover distances of less than 150 miles whereas its shipments of cement produced as Type V LA traveled over substantially greater distances. CEMEX claims that the "differences in shipping distances is simply a geographic fact" and the result of a "legitimate business decision" and that the Department has not relied on shipping distances in determining whether sales were outside of the ordinary course of trade in prior cases. These claims are inapposite. We are not questioning the reasoning behind its sales decision but rather the effect of the decision to ship over long distances. As we noted in earlier reviews, a company may have sound business reasons for changing its methods of operation but, if sales resulting from this new business practice are not normal for the company, then they cannot be within that company's ordinary course of trade. The CIT and CAFC affirmed this analysis in their examination of the 90/91 administrative review. See CEMEX v. U.S. The shipping-distance factor is an important factor as it relates to the freight costs involved in making the sale and also centers on the question of whether a company absorbs such freight costs. CEMEX has acknowledged that transportation is a major cost factor in the industry. In support of this fact, we found that CEMEX incurs much greater freight costs on sales of Type V LA produced cement. Moreover, we found that CEMEX is absorbing the costs of freight in making these sales as compared with sales of Type I cement. Not only is this an unusual business practice for CEMEX but, for years preceding the antidumping duty order, it was CEMEX's normal business practice to include the cost of pre-sale freight in the sales price to its customers of its Type V LA cement. For CEMEX to absorb freight costs after the issuance of the order is an "unusual circumstance," particularly given the high freight costs for cement. With respect to shipping terms, while it is true, as CEMEX points out, that shipping terms (e.g., CIF or FOB plant) for sales of cement produced as Type V LA are similar to those for cement produced as Type I, we believe this contention proceeds from an incorrect premise. In an ordinary-course-of-trade analysis, the pertinent issue is whether the conditions and practices are "normal" for the company in question. As noted above, shipping heavy merchandise, such as cement, over vast distances, incurring and absorbing high freight costs, is not a normal practice in the cement industry. CEMEX argues that the Department misapplied the factors of historical sales trends and promotional quality to the record evidence. We did not take these factors into account in our revised ordinary-course-of-trade analysis. We found that, of the factors that we considered for these final results of review, the record information provided strong enough support, absent the factors concerning historical sales trends and promotional quality, to find that sales of cement produced as Type V and Type V LA were outside the ordinary course of trade. 4. Level of Trade Comment 4: The petitioner contends that, in determining the levels of trade of both CEMEX and CDC, the Department focused only on selling functions in its level-of-trade analysis and did not consider whether the merchandise was sold at distinct stages of distribution. The petitioner points out that, consistent with the preamble to the Department’s regulations and Department practice and with the analysis performed for the final results of the 94/95 review, selling functions alone are not sufficient to establish differences in levels of trade. The petitioner asserts that, had the Department considered the stages of distribution in its level-of-trade analysis, the Department would have found that no differences exist between CDC’s and CEMEX’s home-market and U.S. stages of distribution and, consequently, found that there was no basis for finding differences in levels of trade. CEMEX and CDC disagree with the petitioner’s assertion that the Department did not take into account stages of distribution in its level-of-trade analysis. CDC points out that the Department’s level- of-trade memorandum specifies that the Department began its analysis by examining the customers to which it sold comparison merchandise, the distribution channels through which such merchandise was sold ("stages of distribution"), and the selling functions associated with each stage of distribution. CDC maintains that this type of examination was performed in accordance with Department practice and was also addressed in the verification report and in the level-of- trade memorandum. Moreover, CEMEX and CDC assert that the Department addressed the relevant stages of distribution in the preliminary results. Department’s Position: Our level-of-trade analysis focuses both on the selling functions performed by the foreign producer/exporter as well as the stages of distribution. We cannot divorce one factor from the other in performing our analysis. The selling functions, themselves, play a prominent role in our level-of-trade analysis. The regulations at 19 CFR 351.412(c)(2) make clear that differences in selling activities are indicative of differences in stages of marketing. Furthermore, as stated in the SAA at 829, "[n]ominal reference to a company as a ‘wholesaler’ [i.e., customer category], for example, will not be sufficient." While the stages of distribution play a role in our analysis, the petitioner’s suggestion that we consider the levels of trade on the basis of customer categories is misplaced. For normal value, our level-of-trade analysis is concerned with the sale made to the customer in the home market. That is, the normal-value level of trade is that of the starting-price sale in the home market. For constructed-export-price (CEP) transactions, we are concerned with the sale made to the U.S. affiliated importer. That is, the U.S. level of trade is the level of the constructed sale from the exporter to the affiliated importer. In performing our level-of-trade analysis for CEP transactions, we are not concerned with the sale made to the unaffiliated customer in the United States. Therefore, even if all of the selling functions are the same for the sale made to the unaffiliated customers in the home market and in the United States, we deduct the selling functions that arose in making the sale to the unaffiliated customer in the United States pursuant to section 772(d) of the Act. While the analysis may begin with the starting price for CEP, after the 772(d) deductions, we find a different level of trade, that of the CEP, not the CEP starting price. Thus, in a CEP situation, the levels of trade in the home market and the United States are not likely to be the same. Having made the deduction for selling functions pursuant to section 772(d) of the Act, the selling functions at issue are those that the producer performed in making the sale to the U.S. affiliate or reseller. In the instant case, having made the deductions from the starting price for CEP, in accordance with section 772(d) of the Act, we find that the normal- value level of trade is different and more remote from the factory than the CEP level of trade. Comment 5: The petitioner argues that the Department's methodology for determining the level of trade for CEP sales based on the level of the CEP from the exporter to the related affiliated importer (after deductions required by section 772(d) of the Act) is contrary to the Act and inconsistent with the methodology the Department used to determine level of trade for export-price and normal-value sales. Citing Borden, Inc. v. United States, 4 F. Supp.2d 1221 (CIT 1998) (Borden), the petitioner contends that the CIT found this methodology to be contrary to the requirements of the plain language of the statute. The petitioner asserts that, in order to make an ``apples-to- apples'' level-of-trade comparison, the statute requires the Department to analyze the level of trade for both home-market and CEP sales equivalently, based on the selling functions performed with respect to the sales to the first unaffiliated customer in both markets. In particular, the petitioner contends that, in determining the levels of trade of both CEMEX and CDC as outlined in its level-of- trade memorandum dated August 30, 1999, the Department only focused on selling functions in its level-of-trade analysis and did not consider whether the merchandise was sold at distinct stages of distribution. The petitioner asserts that, had the Department considered the stages of distribution in its level-of-trade analysis, the Department would have found that no differences exist between CDC’s and CEMEX’s home-market and U.S. stages of distribution and, consequently, found that there was no basis for finding differences in levels of trade. The petitioner contends further that, by using the adjusted CEP rather than the CEP starting price in the level-of- trade analysis, the Department is creating and, thus, guaranteeing differences in levels of trade. The petitioner concludes that the Department's practice results in an unfair, skewed comparison between an adjusted CEP and an unadjusted normal value. CEMEX and CDC respond that the Department interpreted section 772(d) of the Act properly and based the CEP level of trade appropriately on the U.S. price after adjustments. CEMEX and CDC argue that the petitioner's sole reliance upon the CIT decision in Borden is misplaced because, as the Department stated in prior determinations, the decision is not final and the Department is appealing the decision. The respondents also assert that the Department's interpretation of the statute is supported in the SAA and the Department's regulations as well as by Department practice. CEMEX also comments that the Department addressed this specific issue in the preamble to the proposed regulations which indicates that using the starting price to determine level of trade results in a meaningless comparison. In addition, CEMEX and CDC disagree with the petitioner’s assertion that the Department did not take stages of distribution into account in its level-of-trade analysis. CDC points out that the Department’s level-of-trade memorandum specifies that the Department began its analysis by examining the customers to which it sells cement, the distribution channels through which such merchandise was sold ("stages of distribution"), and the selling functions associated with each stage of distribution. CDC maintains that this type of examination was performed in accordance with Department practice and was also addressed in the verification report and in the level-of- trade memorandum. Moreover, CEMEX and CDC assert that the Department addressed the relevant stages of distribution in the preliminary results. Department’s Position: Our level-of-trade analysis in this review is consistent with the statute. In determining whether a level-of- trade adjustment is necessary, section 773(a)(7)(B) of the Act directs us to either (1) make due allowance for any difference between the CEP and normal value, or (2) include a CEP offset in the situation where "normal value is established at a level of trade which constitutes a more advanced stage of distribution than the level of trade of the CEP...." Section 772(b) of the Act defines the CEP as "the price at which the subject merchandise is first sold...in the United States...by or for the account of the producer or exporter of such merchandise or by a seller affiliated with the producer or exporter, to a purchaser not affiliated with the producer or exporter, as adjusted under subsections (c) and (d)" (emphasis added). In the case of CEP, section 351.412(c)(1)(ii) of the Department’s regulations directs us to identify "the starting price, as adjusted under section 772(d) of the Act" (emphasis added). Therefore, the statute and regulations are clear that we are to analyze the adjusted price pertaining to the sale made by the U.S. affiliated importer to the unaffiliated purchaser. Accordingly, our level-of-trade analysis is in full compliance with both the statute and regulations. The CIT held in Borden that our practice of determining levels of trade for CEP transactions after CEP deductions is an impermissible interpretation of section 772(d) of the Act. We believe, however, that our practice is in full compliance with the statute. On June 4, 1999, the CIT entered final judgment in Borden on the level-of-trade issue. Borden, Inc. v. United States (Court No. 96-08-01970, Slip Op. 99-50 (CIT June 4, 1999)). We have filed an appeal of Borden which is pending before the CAFC. Consequently, we have continued to follow our normal practice of adjusting the CEP under section 772(d) of the Act prior to starting a level-of-trade analysis, as articulated by our regulations at 19 CFR 351.412. Comment 6: The petitioner contends that the Department’s level-of- trade analysis in this review is inconsistent with the final results of the 94/95 review where the Department compared an unadjusted CEP to an unadjusted normal value and determined that no differences exist with respect to CEMEX’s and CDC’s levels of trade. The petitioner maintains that the facts underlying CEMEX’s and CDC’s distribution systems are not different from those reported by both companies in the 94/95 review and, therefore, the Department should maintain a consistent approach, citing Certain Fresh-Cut Flowers From Colombia; Final Results of Review, 62 FR 53287, 53296-98 (October 14, 1997). Accordingly, the petitioner argues, the Department should reach the same conclusions in this review as it did for the final results of the 94/95 review and deny a level-of-trade adjustment to normal value for CEMEX and CDC. CEMEX and CDC maintain that, since the submission of the petitioner’s case brief, the Department issued its draft results of the remand in the 94/95 review in which the Department reversed its original position and ultimately granted a CEP offset to both respondents. Accordingly, CEMEX and CDC argue that, since the preliminary results now mirror those determined by the Department in the remand of the 94/95 review, this is a moot issue. Department’s Position: Since the time that the petitioner filed its case brief, we have issued the final remand determination for the 94/95 review. In that remand, we revisited our level-of-trade analysis that we had performed for the administrative review and found that, based on record information, the circumstances warranted a CEP-offset adjustment to normal value. Consequently, our level-of- trade analysis in this review is not inconsistent with our final analysis in the 94/95 review. Comment 7: The petitioner argues that CEMEX and CDC did not fulfill their burden of substantiating that their sales in the home market were at a different and more advanced stage of distribution than their sales in the United States. The petitioner challenges the selling functions reported by CEMEX and CDC in their selling- functions charts and argues that the Department should not consider selling functions associated with movement, packing, and direct selling expenses. The petitioner contends that neither company demonstrated that its reported selling functions correspond to the expenses reported in its sales listing, including CEMEX’s reported itemized home-market indirect selling expenses. The petitioner maintains that the Department’s level-of-trade analysis was flawed further by neglecting to measure properly the intensity of the reported selling functions and by not determining the frequency with which each selling function was performed. The petitioner argues that, in performing its analysis, the Department did not quantify the expenses incurred with regard to each selling function. Moreover, the petitioner asserts, certain selling functions CEMEX and CDC reported were not described fully or supported with sufficient information to be deemed selling functions, including market research, procurement and sourcing, personnel training, and after-sales services/warranties. With respect to CEMEX, the petitioner alleges that an inconsistency exists between the selling functions that CEMEX discussed in the narrative portions of its response and those included in its selling- functions chart. The petitioner points out that CEMEX did not respond to the Department’s requests to reconcile such inconsistencies, particularly with respect to the personnel training/exchange and procurement and sourcing selling functions. The petitioner notes that, contrary to information on the record, the Department discerned incorrectly that CEMEX did not perform advertising in the U.S. market and requests that the Department change the designation of this selling function in the level-of-trade memorandum. In addition, the petitioner asserts that, with respect to CDC, home- market selling functions must be provided to at least the majority of customers and that minor or relatively insignificant selling functions cannot provide the basis for a determination that there are different levels of trade and that a CEP-offset adjustment is warranted. Further, the petitioner provided its own selling-functions chart listing eight selling activities performed by CEMEX and CDC which the petitioner believes are relevant to determining whether the levels of trade in the United States and Mexico are similar. The activities listed in the petitioner’s chart includes the following: (1) inventory maintenance, (2) market research, (3) after-sales service/warranties, (4) technical advice, (5) advertising, (6) solicitation of orders, (7) computer/legal/accounting/business, (8) accounts receivable management. In this chart, the petitioner indicates that the respondents performed all eight of these activities in Mexico and six activities performed in the United States, indicating that the market research and after-sales warranties were not performed in the United States. On this basis, the petitioner asserts that the levels of trade in the United States and Mexico are similar such that a CEP-offset adjustment is not warranted. CEMEX and CDC contend that they provided in their responses detailed descriptions of their reported selling functions and the degree to which each selling function was performed in both the home market and the United States. The respondents point out that, at verification, the Department examined their claimed selling functions and the degree to which such functions were performed. CEMEX and CDC note that the Department verified similar information in the last three administrative reviews and found in those reviews, as well as in the instant review, that the level-of-trade information was reported accurately. CEMEX asserts that, based on the law and verified information on the record, the Department's preliminary results included a CEP offset properly. CEMEX contends that it performs significantly more selling functions at a higher level of intensity for sales made in the home market than for sales made to its affiliated U.S. distributor, Sunbelt Cement. CEMEX argues that, in its response, it indicated that numerous links in the distribution chain exist with respect to sales made in the home market whereas only one link in the distribution chain exists with respect to sales made to Sunbelt Cement. In addition, CEMEX refutes the petitioner’s argument that its itemized indirect selling expenses do not correspond to its reported selling functions. CEMEX argues that there is no statutory, regulatory, judicial, or administrative precedent that requires consistency between the indirect selling expenses and the reported selling functions or that requires quantifying the expenses in order for the Department to conduct a proper level-of-trade analysis. CEMEX explains that the itemized indirect selling expenses are representative of cost centers in sales departments whereas the selling functions designate activities performed by department personnel. For these reasons, CEMEX and CDC argue, the Department determined appropriately that the respondents performed significantly different selling functions for the CEP sales than for the home- market sales, that the home-market level of trade was more advanced than the level of the CEP, and that a CEP-offset adjustment was warranted. Department’s Position: At the home-market and U.S. verifications of CEMEX and CDC, we verified the information that the respondents reported with respect to the level-of-trade issue and found no discrepancies. While we did not verify certain reported selling functions, such as CEMEX’s reported procurement and sourcing and personnel training selling functions performed in the home market, it is administratively impossible for us to examine every topic (i.e., selling function or expense) at verification. Rather, our objective is to spot-check the accuracy and completeness of the information that is submitted to us in the course of an administrative review rather than to conduct an exhaustive examination of such information. See Certain Internal-Combustion Industrial Forklift Trucks From Japan; Final Results of Antidumping Duty Administrative Review, 62 FR 5592, 5602 (February 6, 1997). Notwithstanding the fact that we verified the level-of-trade topic pertaining to both companies and were satisfied with the information that we examined, we address below specific aspects raised by the petitioner in its case brief. As a result of the home-market verification we conducted of specific CEMEX selling functions, the petitioner argues incorrectly that we should examine individual transactions because the performance of certain selling functions was volume-driven. Nowhere in the statute or the Department regulations is there a suggestion that we are to conduct level-of-trade analyses on a transaction-specific basis. Furthermore, to conduct a level-of- trade analysis based on the level of detail suggested by the petitioner would not only be administratively burdensome but would suggest an overhaul of the manner in which the Department conducts its level-of-trade analyses. With respect to the petitioner’s issue of quantifying selling functions, our purpose of requesting a respondent to "quantify" the selling functions that it submits in its response is to measure the intensity of or the degree to which such selling functions were performed. Ascertaining this information provides us with a more complete picture of the respondent’s selling practices. We do not, as a matter of practice, quantify or sum for purposes of comparison the expenses that are associated with such selling functions. Such a quantitative analysis may be necessary only in factual scenarios in which we need additional information to better understand the reported information for purposes of our analysis. However, in the instant review, both CEMEX and CDC provided sufficient information regarding the selling functions and the intensity to which such functions were performed such that a further analysis was unnecessary. The petitioner challenges the Department’s decision to grant a CEP offset to normal value for CEMEX by asserting that CEMEX’s itemized list of indirect selling expenses included in the appendices of its response and the selling functions reported in its selling-functions chart do not correspond. However, the list to which the petitioner refers (page B-48 of CEMEX’s December 4, 1998, Section B response) itemizes the names of CEMEX’s accounts for its indirect selling expenses in the home market and does not provide the services performed as a result of those expenditures. Because the list of accounts upon which the petitioner relies is not an itemization of CEMEX’s selling functions but, rather, lists the accounts in which CEMEX’s selling functions are recorded, we would, therefore, not expect CEMEX’s list of indirect selling expenses and selling- functions chart to correspond. We disagree with the petitioner’s contention that selling functions associated with movement and packing expenses should not be considered in identifying levels of trade. Our original questionnaire (Section A, question 3.b) requests respondents to list the types of functions or services performed in each market. We identify, as examples ("among others"), the types of activities that we consider to be functions. However, this list is not exhaustive. "Freight and delivery arrangements" or movement is one of the selling functions that we list. Moreover, a respondent may deem a number of activities to be selling functions, such as personnel training or packing, provided such functions or activities arose as a result of arranging for, or carrying out, the sale of merchandise. We agree with the petitioner’s assertion that we should change in the selling-functions chart included in our level-of-trade memorandum the designation of "no" to "yes" for the advertising selling function. However, this does not change our conclusion. Even if we remove CEMEX’s market research and personnel training or CDC’s selling promotions/discount functions from the list of selling functions that were performed in Mexico and included in our level-of- trade memorandum, the end result is the same as that we found in the preliminary results. In addition, if we did find that the selling functions respondents performed for sales made to unaffiliated customers in the home market and the United States were the same, such as those identified by the petitioner in its case brief at page 71, contrary to the petitioner’s assertion, we would not find the U.S. or home-market levels of trade to the be same. The petitioner disregards the aspect of the analysis that is subject to section 772(d) of the Act. As described in our responses to Comments 4 and 5, this section of the statute requires us to deduct from the CEP starting price the "expenses generally incurred by or for the account of the producer or exporter, or the affiliated seller in the United States in selling the subject merchandise" (emphasis added). We remove from the analysis, therefore, those functions that give rise to as a result of the expenses incurred in selling the subject merchandise to the unaffiliated customer in the United States. Using, for instance, the modified selling-functions chart the petitioner provided in its case brief, we would remove the selling functions that it lists as being performed in the United States and, by virtue of these deductions, find different levels of trade. Based upon our analysis of the record, we determined, as in the preliminary results of review, that CEMEX’s and CDC’s home-market sales occurred at a different and more advanced levels of trade than the U.S. CEP levels of trade and, accordingly, made a CEP-offset adjustment to normal value in our margin calculations. 5. Constructed Export Price Calculation Comment 8: The petitioner contends that the Department’s calculated profit ratio using information from CEMEX’s and CDC’s financial statements is understated. The petitioner requests that, in accordance with section 771a(f)(2)(C)(i) of the Act, the Department incorporate into its final margin calculation the verified cost information submitted by CEMEX and CDC for purposes of calculating the CEP-profit ratio. Department’s Position: As we stated in the preliminary analysis memorandum of September 2, 1999, we relied upon CEMEX’s and CDC’s financial statements submitted in the responses because we did not have the cost data available for the preliminary analysis to determine the CEP-profit ratio. For these final results of review, however, we have incorporated CEMEX’s and CDC’s post-verification cost data into our calculations and have used such data for purposes of calculating the CEP-profit ratio. Comment 9: The petitioner disagrees with the Department's decision not to deduct CEMEX’s and CDC’s indirect selling expenses incurred in the home market on sales to its affiliated importer in the United States (indirect expenses) in calculating the CEP. The petitioner believes that this decision, although consistent with the Department's regulations and current practice as well as the final results of earlier reviews, is contrary to the Act, the SAA, the URAA, and judicial precedent. The petitioner asserts that indirect selling expenses are not limited under section 772(d)(1) of the Act and are therefore required to be deducted in accordance with section 772(d)(1)(D) of the Act. The petitioner explains that, when Congress amended the antidumping law in the URAA, it intended to retain the pre-URAA deduction made for direct and indirect expenses, citing House Rep. No. 103-826, at 79 (1994). In accordance with a clear reading of section 772 of the Act, the petitioner maintains that Congress intended to treat indirect expenses in the same way it treated such expenses under the old law. Thus, the petitioner argues that the Department’s regulations and current practice concerning the treatment of the indirect expenses in question are unsupported by the statute and legislative history. The petitioner contends that the Department has discretion to deduct from the CEP all indirect expenses associated with U.S. sales. The petitioner contends that the Department's use of the term "U.S. expenses" is limited incorrectly to expenses incurred on and associated with sales to unaffiliated customers in the United States. The petitioner asserts that the use of the term should be expanded to include all expenses incurred in relation to sales made to the affiliated importer. Further, the petitioner disagrees with the Department's narrow interpretation of the language in section 772(d) of the Act referring to expenses "associated with economic activities occurring in the United States" to be defined as only those expenses related to sales by the affiliated importer to unaffiliated purchasers. The petitioner contends that the ambiguous language is interpreted more properly to include all expenses related to U.S. sales and asserts that indirect expenses are expenses associated with economic activity occurring in the United States. The petitioner cites the final results of the 94/95 review to demonstrate that the Department acted inconsistently with section 772(d) of the Act by limiting the deduction of "any" expenses to those that arise with respect to activities conducted by the affiliated importers in order to make a sale. The petitioner maintains that, by not deducting indirect expenses from the CEP starting price, the Department calculated the CEP at an ex-importer rather than ex-factory price, a price that does not represent an "apples-to-apples" comparison. In addition, the petitioner claims that the Department misinterprets Article 2.4 of the Antidumping Agreement to require only the deduction from the CEP price of costs incurred between importation and resale. The indirect expenses can be categorized, the petitioner asserts, as "any other differences which are also demonstrated to affect price comparability," as required by Article 2.4. Even if the Department’s reading of Article 2.4 is correct, the petitioner maintains that the statute takes precedence over GATT, citing Suramerica de Aleaciones Laminadas, C.A. v. United States, 966 F.2d 660, 667-668 (Fed. Cir. 1992). The petitioner contends that to allow a deduction from the CEP of only those indirect expenses incurred in the United States permits respondents to avoid deduction of any selling expenses by shifting U.S.-related selling activities offshore. The petitioner also maintains that the Department must interpret section 772(d) of the Act according to its plain meaning, citing Mitsubishi Heavy Industries, Ltd. v. United States, 15 F. Supp. 2d 807 (CIT 1998) (Mitsubishi). The CIT in Mitsubishi, the petitioner asserts, held that the plain language of section 772(d) of the Act requires the deduction, without limitation, of all expenses, including indirect selling expenses, generally incurred in selling the subject merchandise in the United States, regardless of where or when paid. CEMEX and CDC respond that the Department is correct in not deducting indirect selling expenses incurred in the home market from the CEP calculations. CEMEX and CDC state that the petitioner raised the same argument unsuccessfully in the last three administrative reviews. CEMEX argues further that the petitioner attempts to rewrite the legislative history of the URAA and argues that the preamble to the Department’s regulations addresses the petitioner’s claim that the Department’s regulations are inconsistent with the statute. CEMEX also points out that the NAFTA panel in the 94/95 review considered the issue raised by the petitioner and affirmed the Department’s treatment of the expenses in question. CDC maintains that it did not provide any activities directly related to U.S. sales activity and that such activities were performed by its U.S. subsidiary, RGPCC, only. CDC refutes the petitioner's claim that Mitsubishi compels the Department to deduct from the CEP expenses incurred in the home market by a foreign producer and distinguishes the facts in Mitsubishi from those in this case. Moreover, CDC believes that the court’s remand in Mitsubishi reinforces the Department's position to limit acceptable deductions from the CEP. Department’s Position: The SAA at 823 states that the CEP shall be reduced by certain expenses "associated with economic activities occurring in the United States,..." (emphasis added). The SAA at 823 also states that the CEP should be "calculated to be, as closely as possible, a price corresponding to an export price between non- affiliated exporters and importers." The interpretation of both statements in the SAA are codified at 19 CFR 351.402(b), which provides that we will make deductions to the CEP for expenses that are associated with commercial activity in the United States that relate to the sale to an unaffiliated purchaser, no matter where incurred. Consistent with section 772(d) of the Act, the SAA, and section 351.402(b) of the regulations, we did not deduct the respondents’ indirect selling expenses incurred in the country of manufacture because we do not deem these expenses to be related to commercial activity in the United States. Rather, we deduct only those expenses representing activities that relate to the sale made to the unaffiliated customer and not those expenses, such as indirect selling expenses incurred in the country of manufacture, that were incurred in selling to the affiliated U.S. importer. See Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et. al; Final Results of Antidumping Duty Administrative Reviews, 64 FR 35590, 35619 (July 1, 1999), and Gray Portland Cement and Clinker From Mexico; Final Results of Antidumping Duty Administrative Review, 64 FR 13148, 13162 (March 17, 1999). Furthermore, the petitioner’s argument with respect to Mitsubishi is misplaced. In that case, we deducted Mitsubishi’s indirect- selling expenses incurred in Japan from U.S. price. The CIT found, however, that "indirect selling expenses must be associated with economic activity occurring in the United States." Consequently, the CIT ruled that the Department "erred by deducting certain expenses that were not so associated," referring specifically to the indirect selling expenses incurred in Japan that the Department deducted from U.S. price. The CIT ordered the Department to remove Mitsubishi’s indirect selling expenses that were included in the pool of expenses deducted from U.S. price. See Mitsubishi, 15 F. Supp. 2d at 819. Accordingly, Mitsubishi provides additional support for our decision not to deduct CEMEX’s and CDC’s indirect selling expenses incurred in Mexico from CEP. Comment 10: The petitioner maintains that the Department neglected to include indirect selling expenses incurred in the home market on sales to the United States in "total U.S. expenses" for purposes of calculating CEP profit under section 772(f) of the Act. The petitioner argues that including indirect selling expenses incurred in the home market in calculating "total actual profit" but excluding them from "total U.S. expenses" in determining the expense ratio renders the calculation of CEP profit inconsistent. The petitioner argues that this contradictory treatment of the same expenses cannot be reconciled with the statute, the SAA, Department practice, and legislative history. The petitioner also cites U.S. Steel Group v. United States, 15 F. Supp. 2d 892 (CIT 1998) (U.S. Steel), wherein the CIT rejected the Department's inconsistent treatment of movement expenses in the calculation of the CEP. The petitioner concludes that, if indirect selling expenses incurred in Mexico are properly attributable to U.S. sales for the purpose of calculating U.S. selling expenses in the computation of "total actual profit," they must be similarly attributable to U.S. sales for purposes of calculating the expense ratio. By including the indirect selling expenses in question in the numerator of the expense ratio, the petitioner maintains that the Department will ensure that the actual profit and the expense ratio are calculated on the same basis. The petitioner also argues that, in calculating the CEP-profit allocation ratio, the Department included movement expenses improperly in the calculation of total expenses. Citing U.S. Steel, the petitioner contends that, for the purpose of consistency, if the numerator in this ratio is calculated exclusive of movement expenses, then the denominator should also be calculated exclusive of movement expenses. The petitioner argues that section 772(f)(2)(C) of the Act requires the calculation of the denominator to be limited to the production and sale of the merchandise. Because movement expenses are not related to the production or sale of merchandise, the petitioner asserts, such expenses should be excluded from the CEP- profit calculation. The petitioner argues that, for the final results, the Department should recalculate the profit-allocation ratio by calculating the total expenses exclusive of movement expenses. CEMEX and CDC argue that the Department calculated "total U.S. expenses" correctly in its "total expenses" calculation for CEP profit. CEMEX contends that the petitioner has not cited a determination supporting its argument that the Department excluded foreign indirect selling in "total U.S. expenses" incorrectly. CEMEX points out that this same issue was raised in the last three administrative reviews and, for the same reasons, the Department rejected the argument in each of those reviews. CDC cites section 772(f) of the Act, stating that it directs the Department to exclude foreign indirect selling expenses and foreign inventory carrying costs associated with U.S. sales from its definition of total U.S. expenses. CDC cites to a number of recent determinations supporting the Department's treatment of such expenses for purposes of calculating the CEP profit in accordance with the Act. With regard to the petitioner's argument that the Department included movement expenses improperly in its calculation of total expenses to use to calculate CEP, CEMEX and CDC respond that the Department's calculation is a reasonable interpretation of section 772(f) of the Act and is consistent with the Department's past practice. CEMEX and CDC point out that the Department's Policy Bulletin 97-1 enumerates its treatment of movement expenses in the calculation of CEP profit. The respondents argue that the CIT's decision in U.S. Steel is not final and that the Department has not indicated its intention to abandon its prior policy and adopt the decision. CDC notes that the Department is appealing the U.S. Steel decision. Department's Position: We did not make an error by excluding U.S. indirect selling expenses incurred in the country of manufacture from the total U.S. expenses portion of the profit-ratio calculation. Section 772(f)(2)(B) of the Act directs us to include in the calculation of "total U.S. expenses" for purposes of calculating the profit ratio only those expenses that would otherwise be deducted under sections 772(d)(1) and (2) of the Act. We did not find U.S. indirect selling expenses incurred in the country of manufacture to be associated with commercial activity in the United States, nor did we find that such expenses pertain to the sale made to the unaffiliated customer. Accordingly, we must exclude the indirect selling expenses incurred in the country of manufacture from the numerator of the calculation of "total U.S. expenses" in the profit- ratio calculation. This approach is consistent with the intent of the statute and with our practice. With regard to U.S. Steel, we disagree with the CIT's holding with respect to this issue. The decision in U.S. Steel is not final and conclusive as the case is on appeal. Notwithstanding that fact, the SAA explains, at 824, that section 772(d)(3) of the Act refers to profit allocable to "selling, distribution, and further manufacturing" activities in connection with the affiliate's U.S. sale. Excluding movement expenses from "total expenses" would discount the proportionality incorrectly that must logically exist between the "total expenses" calculated and the profits attributable to those expenses, when those profits are based on expenses that include movement. Moreover, such an exclusion does not achieve the statutory purpose of removing the profits associated with all aspects of the affiliate's sale in the United States. Accordingly, for purposes of the final results, we have included movement expenses in "total expenses" for the CEP- profit calculation. 6. Regional Assessment Comment 11: CEMEX and CDC argue that the United States has not honored its obligations under Article 4.2 of the WTO Antidumping Agreement and its predecessor, Article 4.2 of the 1979 Tokyo Round Antidumping Code. CEMEX and CDC claim that the Department has not implemented the special antidumping duty assessment requirements for regional-industry cases set forth in Article 4.2 adequately because it has imposed antidumping duties on all imports of subject merchandise, including those consigned for consumption outside the Southern Tier region, as defined by the International Trade Commission (ITC) in the original investigation. CDC also asserts that the Department’s implementation has been inadequate and belated. It contends that the Department should either terminate this review and revoke the antidumping duty order or should assess antidumping duties only on CDC’s entries of merchandise consumed within the Southern Tier region. CEMEX requests that the Department assess duties and establish cash deposits only on its future entries consumed within the Southern Tier region. CEMEX and CDC argue that the United States is obliged to comply with Article 4.2 of the Antidumping Agreement, which states: "When the industry has been interpreted as referring to the producers in a certain area, i.e., a market as defined in paragraph1(ii), anti-dumping duties shall be levied only on the product in question consigned for final consumption to that area. When the constitutional law of the importing country does not permit the levying of anti-dumping duties on such a basis, the importing Member may levy the anti-dumping duties without limitation only if (a) the exporters shall have been given an opportunity to cease exporting at dumped prices to the area concerned or otherwise give assurances pursuant to Article 8 of this Agreement, and adequate assurances in this regard have not been promptly given, and (b) such duties cannot be levied only on products of specific producers which supply the area in question." According to CEMEX and CDC, Article 4.2 compels the Department to refrain from assessing duties on its subject merchandise destined for consumption outside the Southern Tier region. CDC contends that the exception to Article 4.2 (when the constitutional law of the importing country does not permit levying of antidumping duties on a regional basis) does not apply because none of the conditions necessary to justify an exception to Article 4.2 are satisfied in this case. First, both CEMEX and CDC assert that there is no U.S. constitutional prohibition against levying antidumping duties on a regional basis. CEMEX and CDC contend that neither the port- preference clause of the Constitution, which prohibits Congress from regulating commerce or revenue of ports in a discriminatory manner that would confer preferential treatment for the ports of one state over the ports of another state, nor the uniformity clause, which requires the uniform imposition of taxes throughout the United States, render the regional assessment of antidumping duties unconstitutional, citing U.S. Const. Art. I, Sec. 9, cl. 6, and Art. I, Sec. 8, cl. 1. CDC argues further that the United States has never explained its theory that implementing the general assessment rule would result in a constitutional violation. Second, CDC argues that the condition by which the Department would be exempted from assessing antidumping duties regionally provided it affords the exporter an opportunity to enter into a suspension agreement has not been met. CDC argues that the Department did not permit CDC to enter into a suspension agreement at the time of the original investigation because, at the time of the investigation, the Department's policy was one of refusal to enter into suspension agreements. Moreover, CDC maintains, the Department's decision to collapse CEMEX and CDC in the original investigation diminished CDC's opportunity further to enter into a suspension agreement. CDC also argues that the U.S. implementation of the Article 4.2 assessment rules included no provisions by which these rules could apply to orders predating the URAA. Third, CDC argues that the condition that duties cannot be levied only on products of specific producers which supply the area in question has not been met. It maintains that the language of Section 218 of the URAA and the Department's regulations demonstrate that assessment on less than a national basis is possible. CDC contends that the fact that Congress enacted Section 218 with language calling for the regional assessment of duties attests to the absence of a U.S. constitutional prohibition against regional assessment. CDC points out, however, that Section 218 falls short of implementing the regional-industry rule because it does not address producers or exporters which, like CDC, export merchandise both into and outside the region. In addition, CDC contends that, in the 96/97 review, the Department avoided the substantive regional-assessment issue CDC had raised because the Department did not address and could not demonstrate a constitutional bar prohibiting assessment of duties on a regional basis. CDC points out that the Department indicated that no inconsistency exists between the U.S. antidumping law and GATT. CDC maintains that, if this holds true, the Department must yield to Article 4.2 which states clearly that antidumping duties may be assessed only on products imported for consumption in the relevant, or Southern Tier, region. Accordingly, CDC argues, the Department should assign a zero margin to all sales made outside the Southern Tier region. The petitioner states that the Department has assessed antidumping duties on all nationwide entries of the subject merchandise properly. The petitioner contends that the respondents do not base their argument upon either U.S. law or the Department’s practice but rely upon international trade agreements that have no direct legal effect under U.S. law. The petitioner cites the final results of the last administrative review, where the Department explained that "its determinations are governed by the U.S. antidumping statute, specifically, Title VII of the Tariff Act of 1930, as amended by the URAA in 1995." Moreover, the petitioner adds, Congress has declared that the collection of antidumping duties on a region-specific basis is unconstitutional. The petitioner discusses the statutory provisions that provide for the assessment of antidumping duties in regional-industry cases in a manner that is in accord with both the constitutional constraints and U.S. international obligations. In addition, the petitioner contends that, even if, for the sake of argument, the international agreements cited by CEMEX and CDC were applicable, they would not prevent the Department from assessing antidumping duties on all entries of the subject merchandise. The petitioner notes that neither CEMEX nor CDC appealed the Department’s final affirmative determination to the appropriate court and within the statutory time limit for appeals with respect to definition of "industry" in a regional case or the Department’s alleged failure to offer an opportunity for a suspension agreement during the original investigation. Furthermore, the petitioner contends, the international agreements the respondents cite refer only to the assessment of duties and not to the collection of cash deposits. The petitioner asserts that the arguments brought forward by CEMEX and CDC are based on a fundamental misconception regarding the role of international law, specifically the role that international agreements play in the legal framework of the United States. The petitioner argues that it "is the implementing legislation, rather than the agreement itself that is given effect as law in the United States." The petitioner also argues that CDC’s reliance on Article 4.2 of the Tokyo AD Code is misplaced because the Code was superceded by the WTO AD Agreement. The petitioner asserts that the Department must act within its authority under sections 736(d)(1)-(2) and 734(m)(1)-(2) of the Act, which were amended by the URAA to conform to the regional-industry provisions of the antidumping agreement. The petitioner contends that these provisions are not applicable to respondents and thus confer no authority upon the Department to refrain from assessing antidumping duties outside the Southern Tier. The petitioner argues further that these sections of the Act only apply in investigations and not reviews. Also, the petitioner notes that both CEMEX and CDC do not qualify for the regional assessment of duties under section 736(d)(2) of the Act because both respondents exported subject merchandise into the Southern Tier during the original period of investigation (POI). The petitioner contends that the Department has no obligation under sections 734(m)(1)-(2) of the Act to offer respondents a suspension agreement because the Department may only accept a suspension agreement during the pendency of an investigation or within 60 days after the publication of the antidumping duty order. The petitioner notes that no respondent appealed the Department’s final determination in 1990 based on an alleged lack of an opportunity for a suspension agreement. For these reasons, the petitioner concludes that the Department has complied fully with U.S. law. Department’s Position: As we have maintained in previous reviews, our determinations in an antidumping proceeding are governed by the U.S. antidumping statute - specifically, Title VII of the Tariff Act of 1930, as amended by the URAA in 1995. As numerous courts have recognized, in the event of a conflict between a GATT obligation and a statute, the statute must prevail. See Federal Mogul Corp. v. United States, 63 F.3d 1572, 1581 (Fed. Cir. 1995), citing Suramerica. Congress codified this principle in the URAA at section 102: "[n]o provision of any of the Uruguay Round Agreements, nor the application of any such provision to any person or circumstance, that is inconsistent with any law of the United States shall have effect." See also SAA at 659 ("The WTO will have no power to change U.S. law. If there is a conflict between U.S. law and any of the Uruguay Round agreements . . . U.S. law will take precedence."). Thus, even if respondents were correct in asserting that the statutory provisions relating to regional assessment of duties conflicted with the obligations contained in Article 4.2 of the Antidumping Agreement, the Department must act in conformity with the antidumping statute. Contrary to CDC’s argument that we avoided the issue in the last review, we emphasized in that review, as we do in this review, that we must focus on the law governing determinations in antidumping proceedings. Section 736(d)(1) of the Act provides that, in an investigation in which the ITC makes a regional-industry determination, the Department "shall, to the maximum extent possible, direct that duties be assessed only on the subject merchandise of the specific exporters or producers that exported the subject merchandise for sale in the region during the period of investigation." Because the original antidumping investigation on cement from Mexico occurred in 1989-90 and the URAA applies only to investigations initiated on the basis of petitions filed after January 1, 1995, this provision does not apply to CEMEX’s and CDC’s exports. However, even if section 736(d)(1) of the Act did apply to this review, since CEMEX and CDC exported subject merchandise into the region during the original POI, the Department directed properly that antidumping duties be assessed on all entries of merchandise produced by CEMEX and CDC. Moreover, section 736(d)(2) of the Act provides that, "after publication of the antidumping order, if the administering authority finds that a new exporter or producer is exporting the subject merchandise for sale in the region concerned, the administering authority shall direct that duties be assessed on the subject merchandise of the new exporter or producer consistent with the provisions of section 751(a)(2)(B)." Because neither CEMEX nor CDC is a new exporter or producer as described in this provision, section 751(a)(2)(B) of the Act is inapplicable to the assessment of antidumping duties on subject merchandise exported to the United States by CEMEX or CDC. Furthermore, pursuant to section 734(m) of the Act, in an investigation in which the ITC makes a regional-industry determination, the Department "shall offer exporters of the subject merchandise who account for substantially all exports of that merchandise for sale in the region concerned the opportunity to enter into [a suspension] agreement." Any such agreement is "subject to all the requirements imposed under this section for other [suspension] agreements, except that if the Commission makes a regional industry determination . . in its final determination . . . but not in the preliminary affirmative determination . . . any agreement . . . may be accepted within 60 days after the antidumping order is published under section 736." Under section 734(b) of the Act, we may only accept a suspension agreement during the pendency of an investigation. Because the Department cannot enter into a suspension agreement once the 60-day post-order period has passed (and, indeed, seven administrative reviews have passed), the Department’s decision not to offer respondents an opportunity to enter into a suspension agreement in this review does not violate section 734(m) of the Act. Moreover, although CEMEX argues that the payment of cash deposits should not be required on CEMEX’s entries outside the Southern Tier, the Act contains no provision and describes no circumstances under which we may waive an importer’s requirement to pay cash deposits on subject merchandise sold for consumption in the United States except when conducting new-shipper reviews under section 751(a)(2)(3) of the Act. Accordingly, upon publication of these final results, we will instruct the Customs Service to require the payment of cash deposits and to assess antidumping duties on entries of CEMEX’s and CDC’s subject merchandise that have entered or will enter for consumption both inside and outside the Southern Tier. As demonstrated above, the Department’s decision to assess duties on all subject merchandise exported to the United States by CEMEX and CDC is consistent with the antidumping statute. Indeed, neither CEMEX nor CDC argue that the Department’s actions do not conform to these statutory provisions. For purposes of this administrative review, therefore, the Department need not consider respondents’ arguments further concerning the United States’ implementation of its obligations under the Antidumping Agreement. Nonetheless, we disagree with the respondents’ contention that the antidumping statute does not implement the United States’ obligations under the Antidumping Agreement fully. Because qualifying exporters are given an opportunity for exemption from the assessment of antidumping duties, the statutory scheme described above is consistent with Article 4.2 of the Antidumping Agreement. Thus, the United States has implemented its obligations fully with respect to the assessment of antidumping duties in regional-industry cases. We also disagree with CDC’s contention that we must terminate the review and revoke the underlying antidumping duty order because U.S. implementation of its international obligations is allegedly untimely and inadequate. First, as we have discussed in the "Revocation" section of this document, we have no authority to revoke the order on the basis of industry support at the time the petition was filed. Moreover, neither CEMEX nor CDC appealed the Department’s final determination in the LTFV investigation based upon the Department’s alleged refusal to offer a suspension agreement to the exporters. Thus, the antidumping duty order, based upon the final determinations of the Department and the ITC, is final and binding. 7. Bag vs. Bulk Comment 12: CEMEX and CDC argue that the Department did not make comparisons of the subject merchandise to the foreign like product on a bulk-to-bulk and bag-to-bag basis, which resulted in distortive and unfair comparisons of their sales of cement. CDC explains that, because it sells primarily bagged cement in the home market which is sold at a higher price than the bulk cement that it sells primarily in the United States, the antidumping margin is inherently higher. CEMEX and CDC contend that the mixed comparisons of bagged and bulk cement are contrary to the statute as interpreted by the NAFTA panel in the appeal of the 94/95 administrative review. CDC points out that, while the Department determined in the 94/95 administrative review that bagged and bulk cement were similar in accordance with section 771(16) of the Act, the NAFTA panel determined that bagged and bulk cement were different from one another. CDC contends that the NAFTA panel required the Department to make bag-to-bag and bulk-to-bulk comparisons because record evidence demonstrated that bagged and bulk cement are sold to distinct buyers for different uses commanding different prices and such differences are not due only to packaging. CDC asserts that, in its supplemental questionnaire response for this review, it explained that bagged cement is not the same merchandise as bulk cement owing to the purposes for which it is used, the customers to which it is sold, and, thus, the different prices for which it is sold. CEMEX contends that bagged cement does not satisfy all of the requirements under section 771(16)(B) of the Act to be deemed comparable merchandise. CEMEX contends that, while bagged cement was produced in the same country by the same person as the subject merchandise, sales of bagged cement do not satisfy the other two subparts of the statute. CEMEX argues that, with respect to section 771(16)(B)(ii) of the Act, the component material of bagged and bulk cement and the purposes for which bagged and bulk cement are used are not the same. In this regard, CEMEX argues that the bag itself is a component material that makes bagged cement, as it is sold, different from bulk cement. CEMEX also refutes the Department’s claim that bagged and bulk cement are similar because both cement types are used in the production of concrete. CEMEX asserts that the fact that both cement types are used in the production of concrete is very general and does not take into account the commercial realities of these cement types. CEMEX argues that the inherent meaning of the phrase "purpose for which it is used" in subpart (ii) of the statute is market-driven and, thus, the Department must consider factors such as the physical appearance, interchangeability, and channels of distribution of the merchandise. CEMEX points out that, in considering these market-driven factors, sales of bagged cement in the home-market are made almost exclusively to distributors for resale to small contractors or individuals for do- it-yourself projects. CEMEX also explains that, unlike bulk cement, customers are willing to pay a premium for this bagged cement because of the importance for ease of handling and usability of the cement. In contrast, CEMEX states that sales of bulk cement in both the home market and in the United States are made to end-user customers for use in large projects. CEMEX explains that, in order for customers to handle the bulk cement, specialized facilities and equipment are needed and this type of handling permits the customer to save money by purchasing bulk cement at lower prices. CEMEX and CDC argue further that, with respect to subpart (iii) of the statute, bagged and bulk cement are not equal in commercial value and they take issue with the Department’s position that, because there is no cost difference between these two cement types, except for packaging, these cement types are approximately equal in commercial value. CEMEX contends that the antidumping statute does not use the terms "cost of production" and "commercial value" interchangeably. Moreover, CEMEX and CDC argue that the term "commercial value" refers to the price in the marketplace, not the cost of production. CEMEX asserts that record evidence indicates that, even after adjusting for the cost of the bags, there remains a significant difference in the prices of Type I bagged and bulk cement which demonstrates that bagged and bulk cement are not approximately equal in commercial value. CDC explains that the price for each cement type in the home and U.S. markets is different owing to brand- name recognition of its bagged cement, differences in the markets in which the bagged and bulk cement are sold, purchasing power of bulk customers compared with customers of bagged cement, difference in handling costs for each cement type, and the differences in packing costs. CDC contends that, if the Department compares the prices of bagged and bulk cement on the basis that the only difference between these cement types is product form (i.e., packaging), then its premise for making such comparisons is contrary to recent determinations the Department made in other cases. CDC points to Stainless Steel Round Wire from Taiwan, 64 FR 17336, 17339 (April 9, 1999) (SSRW from Taiwan), and Stainless Steel Round Wire from Korea, 64 FR 17342, 17344 (April 9, 1999), to demonstrate that the Department does not compare products sold in different product forms if two criteria are met: (1) the petitioner must provide evidence that, in the wire industry, product form was a pricing consideration; (2) the petitioner and the Department must identify a distinct correlation between product form and pricing contained in the respondent’s reported sales data. CDC asserts that record evidence in the instant review demonstrates that (1) in the cement industry, pricing for bagged and bulk cement was different, and (2) that sales data submitted for this review demonstrate that the prices of bagged and bulk cement purchased by the same customers were different. Therefore, CDC contends, the Department should not make unfair and distortive comparisons of home-market sales of bagged cement to U.S. sales of bulk cement because the product form is different between these two cement types. In addition, CEMEX disagrees that the record supports the Department’s continued practice of finding Type I bagged cement in the home market comparable to bagged and bulk cement sold in the United States. CEMEX argues that the Department has not had a continual practice of comparing bagged to bulk cement. CEMEX comments that, in the LTFV investigation and in the first four reviews, the Department made bulk-to-bulk and bag-to-bag comparisons in those reviews in which CEMEX sold bagged cement to the United States. CEMEX states that, while the Department compared subject merchandise to the foreign like product which consisted of both bulk and bagged cement in the 94/95 review, the NAFTA panel overturned the comparison methodology employed in that review and limited the model matching to a bulk-to-bulk comparison. CEMEX asserts that such comparisons were affirmed by the CIT with regard to the 91/92 administrative review. CDC argues that the Department’s decision in this review to compare bagged cement to bulk cement is inconsistent with the comparisons the Department made in other cement proceedings, including the LTFV investigation involving gray portland cement and clinker from Venezuela. CEMEX and CDC argue, therefore, that the Department has established a consistent practice of comparing cement on bag-to-bag and bulk-to-bulk bases and should, for the final results of review, follow the recent NAFTA panel decision concerning the 94/95 administrative review. The petitioner states that the Department found that bulk and bagged cement are physically identical merchandise and that the only difference between the two forms of presentation is the manner in which the cement is packed. The petitioner argues that there is no basis for the use of form of presentation as a matching criterion. The petitioner asserts that, although CEMEX is urging the Department to follow the recent NAFTA panel decision concerning the 94/95 review, the panel’s decision is not binding, is not precedent, and is not persuasive. The petitioner argues that, for the current review, CEMEX has not provided the Department with any evidence to warrant departing from its established practice of matching U.S. transactions with home-market transactions without regard to form of presentation. Furthermore, the petitioner comments that CDC’s argument concerning bag-to-bag and bulk-to-bulk comparisons is not supported by any evidence to warrant departing from the Department’s established practice. The petitioner discusses several reasons: there is no statutory basis for excluding a portion of the home-market sales of Type I cement from the foreign like product; CEMEX’s arguments against matching bulk and bagged cement are both legally and factually erroneous; CDC’s arguments against matching bulk and bagged cement are both legally and factually erroneous. Therefore, the petitioner urges the Department to maintain its position in the preliminary results with respect to this issue. Department’s Position: As we stated in the preliminary results, we find that matching the U.S. merchandise which is sold only in bulk to the foreign like product sold in both bulk and bags is appropriate. The cement types the respondents sold in the United States during the instant review period were Type V LA and Type II. CEMEX and CDC sold Type I, Type II LA, Type V LA, Type II, and pozzolanic cement in Mexico. For the reasons discussed in the preliminary results, we did not use home-market sales of Type V LA, Type II LA, Type II, and pozzolanic cement as comparators to sales of subject merchandise. Having rejected sales of Type V LA and Type II as comparators to the sales of subject merchandise, we no longer had a basis to match the subject merchandise to identical sales of the foreign like product. In accordance with section 771(16)(B) of the Act, we matched the sales of subject merchandise to the entire universe of Type I sales, including both bulk and bagged cement. Section 771(16) of the Act is silent with regard to the methodology we should develop for purposes of matching the subject merchandise to the foreign like product. Moreover, Congress, the courts, and even Chapter 19 panels have granted us broad discretion in developing our model-match methodology. See Kerr-McGee Chemical Corp. v. United States, 741 F. Supp. 947, 951-52 (CIT 1990), and Certain Cut-to- Length Carbon Steel Plate from Canada, USA-93-194-04, Opinion at 12 et seq. (October 31, 1994) ("Commerce has discretion in the establishment of a product characteristic hierarchy as an aid in its selection of product matches."). In Koyo Seiko Co., Ltd. v. United States, 66 F.3d 1204, 1209 (Fed. Cir. 1994), the Federal Circuit acknowledged that "Congress has implicitly delegated authority to Commerce to determine and apply a model-match methodology necessary to yield ‘such or similar’ merchandise under the statute." For the reasons discussed below, we believe that comparing sales of bulk cement with sales of bagged and bulk cement is reasonable, consistent with the statute, and supported by administrative practice. The Department has a longstanding practice of developing a model- match methodology in the early stages of each proceeding. In this respect, the Department, in consultation with the parties, selects commercially relevant product characteristics that dictate the selection of a foreign like product. Thus, one may presume that products which are selected based on the model-match criteria meet the statutory elements of section 771(16)(B) of the Act which includes physical characteristics, purposes, and commercial value. For example, in Pasta From Italy, the Department determined that "wheat quality" was an appropriate physical characteristic to be included in the Department’s model-match methodology because these quality differences were "reflected in . . . costs and pasta prices" and were, therefore, "commercially significant." Certain Pasta From Italy, Final Determination of Sales at LTFV, 61 FR 30326, 30346 (June 14, 1996). Since the LTFV investigation in this case, the Department has selected the foreign like product based upon ASTM specifications because all parties have acknowledged the commercial significance of these specifications. In the LTFV investigation, the Department indicated that "both petitioner and CEMEX have noted that customers and producers in both markets rely on ASTM standards to differentiate between products." See Gray Portland Cement and Clinker from Mexico, 55 FR 29244, 29247 (July 18, 1990). Record evidence in this review reflects the fact that the physical differences between Types I, II LA, and V LA affect the purposes for which these cement types can be used. For example, one CEMEX official stated that "the ultimate customer tells the company that it wants Type V LA cement or Type II LA cement used for specific applications." See pages 3-4, Attachment A, of CEMEX’s submission dated January 19, 1999. These differences in the physical characteristics of cement produced to different specifications as well as the different applications for which these cement types may be used also affect the commercial value of cement produced as Type I, Type II LA, or Type V LA. Furthermore, adjusting for these differences is consistent with our methodology for recognizing that cost differences (i.e., additives such as kaolin) affect the physical make-up of the product and, thus, the ASTM specifications. Consequently, these cost differences ultimately result in price differences between the different cement types. Indeed, CEMEX notes that there are "...traditional price premiums for these cements [Types V LA and II LA] compared to Type I." Id. at 4. Therefore, based upon this record evidence, it is apparent that cement types produced to different ASTM specifications are physically different, used for different purposes, and have different commercial values. We disagree with CEMEX’s assertion that the bag itself is a component material which differentiates bagged cement from bulk cement. Section 771(16)(B) of the Act requires, at a minimum, that the foreign like product be similar "in component material or materials". The statute does not contemplate presentation type. The presentation or packaging of the merchandise merely dictates the form in which the merchandise is sold (e.g., wrapped or sealed). It does not affect the constitution or component material of the product in question. Thus, the packaging of merchandise does not alter the material components of the merchandise. Interpreting the statute to allow for packaging (i.e., bags, plastic, etc.) to alter the composition of the merchandise may provide the respondent an opportunity to manipulate the matching of products and, consequently, the outcome of our review. While there have been instances in which the Department compared bag-to-bag and bulk-to-bulk cement, there is also case precedent which supports our methodology to consider both bulk and bagged cement as the foreign like product. See Japanese Cement, 60 FR 43763 (August 23, 1995). Similarly, we do not find that the cases CDC cites are persuasive. In SSRW from Taiwan, for example, the Department did not compare merchandise based upon packing form and, therefore, CDC’s cite to SSRW from Japan does not reflect a Department practice which would compel the Department in this case to exclude home-market sales of bagged cement from the price comparisons. In the instant case, similar to the facts in SSRW from Taiwan, there is no indication that the packing form of cement affects the commercial value of the merchandise. As described above, the commercial value, as reflected in cost variances and prices, is dictated by the ASTM specification of the cement and not necessarily the manner or form in which the cement is delivered to the customer. In addition, case precedent supports our methodology to include both bulk and bagged cement as the foreign like product. See Gray Portland Cement and Clinker from Japan, Final Results of Determination, during Administrative Review, 60 FR 43761, 43763 (August 23, 1995). We also disagree with CEMEX’s argument that bulk and bagged cement are different in the "purposes for which it is used." Section 771(16)(B)(ii) of the Act states that the merchandise must be "like" that merchandise in the purposes for which it is used. In this context, the statute requires that the merchandise have like or similar purposes for which the merchandise sold in the United States is used. Bulk and bagged cement are used to produce concrete in both markets. Therefore, cement sold in both forms will be used for similar or "like" purposes. While CEMEX refutes our finding that bulk and bagged cement are used in the production of concrete and argues that this statement in the preliminary results of review is too general, its argument is not compelling because this subsection of the statute provides us some discretion in identifying merchandise on the basis of component materials and purposes for which the merchandise is used. In its case brief, CEMEX and CDC take issue with our statement in the preliminary results where we indicated that, in addition to other elements that satisfy section 771(16)(B) of the Act, we found no cost difference between bulk and bagged cement and, therefore, concluded that bulk and bagged Type I cement are approximately equal in commercial value. We find that CEMEX's argument that equates commercial value with price is not compelling. The statute does not define "commercial value," nor does it require that commercial value be based upon price. The price or price variance of merchandise is not necessarily indicative, nor the sole determinant, of commercial value. Indeed, we have used cost information previously in determining what is approximately equal in commercial value. See Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and Components Thereof, From Japan; Final Results of Antidumping Duty Administrative Reviews, 63 FR 2558, 2574 (January 15, 1998). Furthermore, neither the price variance of bulk and bagged cement nor the presentation of cement to which CEMEX refers is related to these physical differences and purposes for which the cement is used that we consider relevant to the selection of similar merchandise pursuant to section 771(16)(B) of the Act. Consequently, unlike cement that is produced to different ASTM specifications, we do not find that the commercial value of one cement type (i.e., Type I) varies based upon the manner in which cement is packaged or deliverd. We conclude that the selection of all Type I cement as similar merchandise is consistent with section 771(16)(B) of the Act for the following reasons: Type I cement sold in both bulk and bags is produced in the same country, Mexico, and by the same person, CEMEX and CDC (collectively, "CEMEX"), as that of the subject merchandise; Type I cement, whether sold in bulk or in bags, is "like" the subject merchandise in "component materials" and in the purposes for which it is used; Type I cement, whether sold in bulk or in bags, is approximately equal in commercial value to Type V LA and Type II cement sold in the United States. Further, case precedent supports our methodology to compare bulk and/or bagged cement sold in the United States to both bulk and/or bagged cement sold in the home market. See Calcium Aluminate Cement Clinker and Flux from France, 59 FR 14136, 14143-44 (March 25, 1994), Gray Portland Cement and Clinker from Venezuela, 56 FR 56391 (November 21, 1991), and Frozen Concentrated Orange Juice from Brazil, 57 FR 3995 (February 3, 1992). Finally, we believe CEMEX’s and CDC’s reliance on the NAFTA panel’s decision regarding the 94/95 administrative review which reversed our decision on the bulk/bag issue is misplaced. As the NAFTA makes clear, panel decisions do not have precedential effect since each decision is "limited to the particular matter between the particular parties before the Panel." See NAFTA Article 1904(9). Also, in this case, as of the time of issuance of these final results, the time to appeal the decision for Extraordinary Challenge Committee review has not ended. 8. Difference-in-Merchandise Calculation Comment 13: The petitioner argues that the Department should base CEMEX’s difference-in-merchandise ("DIFMER") adjustment on adverse facts available. The petitioner claims that, despite the Department’s repeated requests for information, CEMEX has not provided information to justify a DIFMER adjustment by isolating and quantifying the variable cost differences. As additional support for the use of adverse facts available, the petitioner states that CEMEX did not reveal until the cost verification that it had not provided accurate cost data for the raw material kaolin, which is the basis for CEMEX’s claimed DIFMER calculation. Further, the petitioner contends that CEMEX’s claim that DIFMER should be based on the methodology of using the processing cost of one plant and the raw-materials costs of three plants is inappropriate. The petitioner claims that the only purpose for weight-averaging the cost of the three plants is to lower the DIFMER adjustment by reducing the effect of the high kaolin cost which is only incurred at one plant. In addition, the petitioner argues that the DIFMER adjustment should not be based solely on the Hidalgo plant’s kaolin cost. According to the petitioner, the record establishes that there are cost differences associated with differences in the physical characteristics for Type I and Type V LA cement other than kaolin costs. Specifically, the petitioner comments that the cost verification report indicates that the raw material mix differs with the production of each type of cement. In addition, the petitioner asserts that an affidavit on the record supports its position that grinding times for each type of cement differs, such that more electricity is consumed when producing Type V LA than when producing Type I. Finally, the petitioner asserts that, if the Department bases the DIFMER adjustment on the difference in the variable costs of producing Type I and Type V LA at the Hidalgo plant, the adjustment should reflect the actual cost of kaolin used in producing Type V LA cement. CEMEX responds that, in calculating the DIFMER adjustment, the Department should average the material cost differences (i.e., kaolin costs) at the Hidalgo plant with the material cost differences at the Yaqui and Campana plants, the other CEMEX plants that produce Type V LA cement. It maintains that such an approach reflects the cost differences between producing the cement sold in the U.S. market and the home market. Alternatively, CEMEX argues that, if the Department does not average the differences in cost at the three plants, the Department should limit the DIFMER calculation to the cost differences between Type I and Type V LA cement produced at its Hidalgo plant, the only plant that produced both Type I and Type V LA cement during the instant review period. CEMEX comments that it was able to segregate the cost differences attributable to the physical characteristics between the cement produced as Type I and Type V LA at its Hidalgo plant from those attributable to plant efficiencies. CEMEX asserts that both types of cement were processed in the same production kiln and final grinding equipment. CEMEX claims that it maintained detailed production statistics for the Hidalgo plant which identify all cost differences between Type I and Type V LA cement. CEMEX points out that the Department verified the accuracy of the Hidalgo plant’s production records and variable cost differences without exception. CEMEX asserts that, if the Department bases DIFMER on the costs differences attributable to Type I and Type V LA cement produced at the Hidalgo plant, then the Department should use the verified cost information to calculate a DIFMER adjustment. CEMEX argues that in this review it produced Type I and Type V LA cement in commercial quantities at its Hidalgo plant and, as such, was able to segregate the cost differences attributable to the physical characteristics between the cement types from those attributable to plant efficiencies. CEMEX asserts that both types of cement were processed in the same production kiln and final grinding equipment. CEMEX claims that it maintained detailed production statistics for the Hidalgo plant which identify all cost differences between Type I and Type V LA cement. CEMEX points out that the Department verified the accuracy of the Hidalgo plant’s production records and variable cost differences without exception. CEMEX asserts that the Department should use the verified cost information to calculate a DIFMER adjustment and not rely upon facts available. CEMEX asserts that it has demonstrated during this review that, at the Hidalgo plant where it produced Type I and Type V LA in commercial quantities, the only cost difference between Type I and Type V LA cement results from a difference in materials. While the clay quarried at Hidalgo produces a natural Type I cement, in order to produce Type V LA cement, CEMEX explains, it uses a special clay additive called kaolin to correct the clay mix which is not necessary for the production of Type V LA cement at the Yaqui and Campana plants. Further, CEMEX maintains that the production records at the Hidalgo plant demonstrate that there is no difference in the variable processing cost of Type I and Type V LA and that the Department verified the accuracy of the information. In CEMEX’s view, it demonstrated at the Hidalgo plant verification that measuring the cost for the two primary cement processes of heating the raw- material mix in the kiln and final grinding of the clinker results in an inconsistent pattern of small differences which does not justify an adjustment. CEMEX also contends that a DIFMER adjustment for final grinding is unwarranted because the record shows that the Hidalgo plant produces Type I and Type V LA cement to the same blaine fineness. With respect to the petitioner’s argument that there are additional material cost differences other than kaolin, which arise in the production of different types of cement due to the percentage of material used, CEMEX contends that all the material inputs, with the exception of kaolin and iron oxide, are extracted from the Hidalgo quarry. As such, CEMEX asserts, all the extracted raw materials have the same per-unit cost per ton and the petitioner’s argument that a DIFMER adjustment persists because of the percentage differences in raw-material inputs is unsupported. In addition, CEMEX argues that the petitioner’s assertion that adverse facts available should be used due to CEMEX’s reporting of an incorrect cost for kaolin and revealed on an untimely basis is misplaced. CEMEX comments that the minor correction for the cost of kaolin, which it presented on the first day of verification, occurred because the transportation cost from the unaffiliated supplier was excluded inadvertently from the transfer price that the Monterrey plant charged the Hidalgo plant. Thus, CEMEX claims, it was in full compliance with the Department’s procedures for submitting minor errors found prior to verification. Department’s Position: Section 773(a)(6)(C)(ii) of the Act requires us to account for and adjust for any differences attributable to physical differences between merchandise exported to the United States and the merchandise sold in the home market where similar products are compared. Section 351.411(b) of the regulations directs us to consider differences in variable costs associated with the physical differences in the merchandise. In this case, during the POR, CEMEX sold Type I cement in the home market and Type V LA cement to the United States. We determined that there are differences in the physical characteristics of cement produced as Type I and Type V LA and, therefore, find that a DIFMER adjustment is appropriate. Accordingly, we must determine the variable cost-of- manufacturing (VCOM) differences attributable to the differences in the physical characteristics of Type I and Type V LA cement. As stated in the SAA at 828, "Commerce will not make an adjustment under this section for cost differences attributable to: (1) the fact that the exporter is charged different prices for its inputs depending on the destination of the finished product; or (2) the fact that the domestic and exported products are produced in different facilities with differing production efficiencies." Further, Import Administration’s Policy Bulletin Number 92.2 (July 29, 1992) states that "it is important in any consideration of a DIFMER to isolate the costs attributable to the difference, not just assume that all cost of production differences are caused by the physical differences. When it is impossible to isolate the cost differences, we should at least determine that conditions unrelated to the physical difference are not the source of the cost differences, such as when different facilities are used...." In this case, CEMEX has numerous facilities which produced cement during the POR. With the exception of one plant, its Hidalgo facility, these facilities produced either Type I or Type V LA cement, not both. As a result, we do not consider it appropriate to simply weight- average the VCOMs for each type of cement across plants because of the varying plant efficiencies. That is, by simply averaging the VCOMs for cement produced at the Yaqui, Campana, and Hidalgo plants, we would be computing a DIFMER which is partially the result of differing plant efficiencies. In order to isolate variable cost differences due strictly to the differences in the physical characteristics between the two cement types being compared, we focused on production costs incurred at the Hidalgo plant because it produced commercial quantities of both Type I and Type V LA cement during the POR. At verification, we found that the only measurable cost difference attributable to the physical differences between Type I and Type V LA cement was the cost of the raw material input, kaolin. Although we found small, inconsistent processing cost differences in the production of both types of cement at the Hidalgo plant, those differences were not due to the differences in the physical characteristics of the two cement types. Rather, the differences were a result of timing differences in the production process (e.g., because the outside temperature has an impact on energy consumption or energy costs may differ depending on which month of the year production occurred). Given the fact that, in the instant review, we were able to isolate the cost differences attributable to the physical differences between Type I and Type V LA cement at the Hidalgo plant, we have calculated the DIFMER adjustment based on the corrected cost of kaolin incurred by the Hidalgo plant which was included in the production of Type V LA cement. We did not find it necessary to apply facts available for the purpose of calculating CEMEX’s DIFMER because CEMEX produced Type I and Type V LA cement at a single plant, Hidalgo. This information supports the calculation of a DIFMER adjustment to normal value for CEMEX in the instant POR. Comment 14: The petitioner refers to the Department’s exclusion of certain months from CDC's DIFMER calculation because of "aberrational factors" affecting the variable cost. As such, the petitioner maintains that, according to a worksheet in the cost verification exhibits, in May, a month which the Department included in the preliminary DIFMER calculation, CDC used Type V raw mix to produce Type I clinker. The petitioner contends that the use of Type V raw mix in the production of Type I cement is aberrational and that the inclusion of the May cost data distorts the DIFMER calculation. CDC responds that the exhibit to which the petitioner refers to demonstrates that the total costs for Type V raw mix was transferred to the Type V clinker processing stage and resulted in an increase in inventory for Type V clinker. Thus, CDC contends, it did not use any of the Type V raw mix to produce Type I cement and, accordingly, the Department should include the May cost data in the DIFMER calculation. Department’s Position: The cost verification exhibit indicates that the total costs for the Type V raw mix were consumed in the production of Type V clinker and not Type I clinker as the petitioner alleges. Thus, for these final results we have continued to include the May cost data in the DIFMER calculation. For further discussion concerning our reasoning behind subtracting certain months in CDC’s DIFMER calculation, see the difference-in-merchandise memorandum, dated August 31, 1999, p. 2. 9. Sales-Below-Cost Test Comment 15: CEMEX and CDC assert that the Department excluded certain below-cost sales from the calculation of normal value erroneously. CEMEX and CDC argue that, under the Department’s administrative precedent, below-cost sales will be disregarded only if they exceed 20 percent of the total home-market sales volume. CEMEX and CDC assert that, in its amended margin-calculation, the Department found less than 20 percent of CEMEX’s and CDC’s home- market sales below cost. CEMEX and CDC argue that, in accordance with the statute and Departmental precedent, no home-market sales should have been disregarded. CEMEX and CDC suggest that the Department delete the reference to manufacturer in the sales-below- cost test to correct this error. Department’s Position: We agree with the respondents that we should delete the reference to manufacturer in our sales-below-cost test. While we have not incorporated the respondents’ suggested language, we have changed our calculations to ensure that we do not take the manufacturer into account in our sales-below-cost test. This is consistent with our approach to collapse the two entities and treat them as one for margin-calculation purposes. Having disregarded the manufacturer in the cost test, we find that there were no sales made at prices below the cost of production. 10. Specialty Cement Comment 16: The petitioner contends that, during the cost verification, CEMEX revealed for the first time that it had not reported its production of an entire type of specialty cement. The petitioner asserts that, not only did CEMEX not report the production of this product prior to verification, but even since then it has not provided any information with respect to the product such as its physical and chemical specifications. The petitioner argues that it is not uncommon for CEMEX to produce a cement meeting a particular ASTM specification and label and sell it as a different cement type. Furthermore, the petitioner asserts that CEMEX has not clarified whether CEMEX included sales of the special cement in the home-market sales database it submitted for the instant review, and, if so, how it reported the sales (i.e., as Type I, Type II LA, or Type V LA) and how they can be identified in the home-market database. Without this basic information, the petitioner asserts that the Department cannot evaluate how it should deal with sales of this product for purposes of calculating the dumping margin and, therefore, the Department should rely on adverse facts available. In addition, the petitioner argues that the use of adverse facts available is especially appropriate given CEMEX’s prior record of persistently misrepresenting the types of cement it produces at particular plants. The petitioner suggests that the Department should, at the very least, apply partial adverse facts available and use the highest normal value for any home-market sale made from one of CEMEX’s plants. CEMEX rebuts the petitioner’s assertion that the existence of the sales of special cement was disclosed only as a result of the Department’s cost verification. CEMEX argues that this information was included in the home-market sales verification and the cost reconciliation submission submitted as part of CEMEX’s supplemental cost response. Moreover, CEMEX argues that the total tonnage at issue is very small when compared with the total production of cement during the POR. CEMEX comments that the Department’s verifiers also considered the quantity at issue to be insignificant and did not pursue the inquiry further. Department's Position: During our verification of CEMEX's sales responses, we examined the completeness and accuracy of CEMEX's reported quantity and value of merchandise sold and found no discrepancies. Thus, we were satisfied that CEMEX reported the universe of sales made during the review period accurately. Even if the existence of sales of CEMEX's special cement was raised prior to the sales verification, our regulations, at 19 CFR 351.307(b) and (d), provide us with significant flexibility in conducting verifications by permitting the examination of a sample of expenses, adjustments, and other topics that we consider relevant to factual information submitted. It is administratively impossible and impractical to examine every topic at verification. Furthermore, verification is not intended to be an exhaustive examination of every topic. See Certain Internal-Combustion Industrial Forklift Trucks From Japan; Final Results of Antidumping Duty Administrative Review, 62 FR 5592, 5602 (February 6, 1997). We did not pursue this issue further in the instant administrative review owing to the insignificant quantity sold of this specialty cement. For further information concerning the tonnage sold of this cement, see the final analysis memorandum from case analyst to file, dated March 6, 2000. However, we do not believe that this topic should be disregarded in its entirety and will pursue inquiry concerning CEMEX's special cement in the 98/99 administrative review which we initiated on October, 1999 (64 FR 53318). 11. Assessment-Rate Calculation Comment 17: CDC argues that the Department should calculate separate assessment rates for each importer. CDC explains that, because (1) it maintains separate facilities and employees from CEMEX, (2) its importers do not share transactions, and (3) its importers import different cement types, there is no reason to warrant the Department’s departure from past practice and calculate a single importer-specific assessment rate. Department's Position: Our decision to calculate a single assessment value stems from our decision to collapse both CEMEX and CDC and treat them as a single entity for the purpose of this administrative review. Therefore, consistent with our decision to treat CEMEX and CDC as a single entity we have calculated a single weighted-average margin and, accordingly, one importer-specific assessment value. The assessment figure represents the actual duties due and is the amount that we instruct the U.S. Customs Service to use in collecting duties from importers, which is how we implement our final results of review. While CDC maintains that it has separate facilities and employees from CEMEX and that it does not share transactions of imported cement, these factors do not have any bearing on our decision to treat these companies as a single entity, let alone our decision to calculate a single importer-specific assessment rate. As mentioned in the collapsing memo, dated August 6, 1999, we find that CEMEX and CDC satisfy the statutory definition of affiliated persons and meet other criteria necessary to treat them as a single entity. Therefore, our calculation of a single weighted-average margin and, thus, a single importer-specific assessment rate for CEMEX and CDC is consistent with our decision to collapse both companies for this administrative review. Calculating two assessment rates, one for CEMEX and one for CDC, is inconsistent with our decision to treat them as a single entity. In addition, while CEMEX and CDC sold different cement types in the United States, we have matched sales of cement sold in the United States to sales of cement sold in Mexico, regardless of the cement type and manufacturer. Again, this is consistent with our decision to collapse these two companies. Finally, our decision to calculate a single importer-specific assessment value in the instant administrative review was not influenced by prior reviews of this order. Rather, our decision was based on and limited to the facts on the record in this administrative review. Indeed, because the statute is silent with respect to the methodology by which we will calculate assessment rates (see Federal Mogul Corp. v. United States, 918 F. Supp. 386, 404 (CIT 1996)), the application of a different methodology in this review does not render our methodology inconsistent with the statute. 12.Adjustments a. Rebates Comment 18: The petitioner argues that the Department should deny CEMEX’s claimed adjustment to normal value for rebates and discounts. Citing SKF USA Inc. v. United States, 180 F.3d 1370, 1377 (Fed. Circ 1999) (SKF USA), and Torrington Co. v. United States, 818 F. Supp. 1563 (CIT 1993) (Torrington), the petitioner contends that CEMEX has not met its burden of demonstrating that its discounts and rebates do not cover non-scope merchandise. The petitioner argues that CEMEX must to demonstrate how it identified those adjustments that pertained to multiple invoices and segregated adjustments paid on non-scope merchandise. It contends that, with respect to certain invoices examined at verification, CEMEX did not provide documentation identifying other invoices to which each rebate pertained. The petitioner argues that many of CEMEX’s large allocated rebates were questionable as they exceeded the gross value of the invoice. The petitioner claims that a number of CEMEX’s large rebates were either allocated incorrectly or were reported on a transaction- specific basis when, in fact, such rebates pertained to numerous invoices. The petitioner maintains that CEMEX’s distortive and inaccurate reporting of its rebates is contrary to the Department’s regulations at 19 CFR 351.401(g)(1). For these reasons, the petitioner asserts that the Department should disallow all of CEMEX’s rebates. CEMEX refutes the petitioner’s claim that the Department should not deduct its discounts and rebates from normal value. CEMEX states that it provided detailed descriptive data of its discount and rebate programs in its response and, with respect to its non-allocated discounts and rebates, the record establishes that it calculated these adjustments on a transaction-specific basis and limited them to scope merchandise. CEMEX points out that the invoices attributable to sales on which these discounts and rebates were paid are limited to a single cement type and presentation. With respect to its allocated rebates, CEMEX argues that the petitioner’s reliance upon SKF USA and Torrington is misplaced because those rulings pertain to the Department’s treatment of circumstance-of-sale adjustments. In contrast, CEMEX explains, discounts and rebates are not synonymous with selling expenses and are held to a different standard in accordance with 19 CFR 351.102. CEMEX maintains that it allocated the rebates based on customer, cement type, presentation (bulk or bag), and calendar year and, thus, the rebates were limited to scope merchandise. CEMEX points out that the Department verified CEMEX’s allocation methodology which the Department has accepted and used for the last three administrative reviews. CEMEX argues that the petitioner has not provided any evidence lending support to its argument that adjustments that are abnormally large should be disallowed. CEMEX also refutes the petitioner’s assertion that its allocated rebates are distortive. While the allocated adjustments may yield results different from the transaction-specific adjustments, CEMEX maintains there is no requirement that the results of these types of adjustments duplicate one another. For the final results, CEMEX argues that the Department should continue to grant its adjustments to normal value for all of its reported discounts and rebates. Department’s Position: We have allowed CEMEX's claimed rebate adjustments because the data was submitted in accordance with our accepted methodology for calculating rebates. Section 351.401(g) of the regulations states that allocations for price adjustments, such as rebates, may be allowed provided the price adjustments are calculated on as specific a basis as is feasible and provided the respondent can demonstrate that the allocation method does not cause inaccuracies or distortions. At verification we found that CEMEX allocated its reported rebates by presentation, cement type, customer, company, and year. See CEMEX’s home-market verification report from analysts to file dated July 23, 1999. CEMEX reported these rebates on as specific a basis as possible, given the manner in which it maintained such rebate data in its computerized commercial system database. Generally, "we have accepted claims for discounts, rebates, and other billing adjustments as direct adjustments to price if we determined that the respondent, in reporting these adjustments, acted to the best of its ability and that its reporting methodology was not unreasonably distortive.'' See Antifriction Bearings (Other than Tapered Roller Bearings) and Parts Thereof from France, et al.; Final Results of Antidumping Duty Administrative Reviews, 62 FR 2081 (January 15, 1997). Furthermore, we allow adjustments to normal value for rebates if we are satisfied that such rebates reflect the respondent’s normal business practice and are not an attempt by the respondent to eliminate dumping margins once we initiate an antidumping investigation or review. See Certain Corrosion-Resistant Carbon Steel Flat Products From Japan, 63 FR 47465, 47468 (September 8, 1998). In this respect, based on CEMEX’s response and our verification of this topic, we are satisfied that, in accordance with 19 CFR 351.401(g), CEMEX reported its rebates on as specific a basis as was feasible given the manner in which it maintains its records in the ordinary course of business and that its allocation methodology is not unreasonably distortive. Finally, the cases the petitioner cites, including SKF USA, relate to the Department’s practice prior to changes made by the URAA and adoption of the Department’s new regulations. Consistent with our 96/97 Review Final Results, we have allowed adjustments for CEMEX’s reported rebates. b. Freight Comment 19: The petitioner contends that the Department should deny CEMEX’s reported home-market freight adjustment. In the alternative, the petitioner contends that the Department should use partial facts available and limit CEMEX’s freight deduction on shipments from warehouse to customer to the smallest value reported for each company. The petitioner asserts that CEMEX did not demonstrate adequately that it is entitled to the adjustment on home- market sales. The petitioner argues that CEMEX did not report its home-market freight expenses on a transaction-specific, customer- specific, or point-of-sale basis and has not demonstrated that it was not feasible to report these expenses on such a basis. The petitioner also maintains that information the Department examined at verification indicates that CEMEX could have provided freight expenses on either a customer-specific or point-of-sale basis. The petitioner argues that allocated movement expenses are allowable under 19 CFR 351.401(g) only if the respondent can demonstrate that it cannot report the expenses on a more specific basis and amounts reported are not distortive of the actual amounts incurred. The petitioner points out that CEMEX determines its freight expenses by the weight of the shipment and the distance the cement was shipped. The petitioner explains that, because freight costs were reported on a per-ton basis, the costs do not take into account the effect that shipping distance has on freight expenses. The petitioner asserts that freight expenses incurred on merchandise transported directly from the plant to the customer are much less than those incurred on merchandise transported from the plant to the terminal to the customer. Combining these expenses, the petitioner argues, results in distortive freight amounts. In addition, the petitioner argues that CEMEX’s calculation methodology skewed the arm’s-length test. The petitioner asserts that, by reporting an average freight cost, CEMEX can mask freight expenses incurred on sales made to affiliated and unaffiliated customers and, thus, cause sales to pass the arm’s-length test that would not be able to pass otherwise. The petitioner argues further that CEMEX’s calculation of average freight costs skewed the sales-below-cost test because the average freight expenses understated actual freight expenses and, thus, decreased the possibility of certain sales failing the cost test. CEMEX responds that the Department has accepted its reporting methodology of the freight expenses in question for the last three reviews. CEMEX argues that, in accordance with 19 CFR 351.401(g)(3), it reported this expense in as feasible a manner as its records permitted. CEMEX maintains that, owing to the high volume of sales made during the POR, it would be burdensome to report transaction- specific freight information. While the allocated freight expenses may yield results different from the transaction-specific adjustment, CEMEX maintains there is no requirement that the results of these types of adjustments duplicate one another, citing 19 CFR 351.401(g)(1). CEMEX points out that it is the Department’s preference to use a weight-based rather than a value-based allocation when expenses are incurred on the basis of weight. Moreover, CEMEX asserts that its freight calculation excluded non-scope merchandise. CEMEX contends that it reduced the possibility of distortion by calculating the freight expenses on the basis of cement type and presentation (bulk or bagged cement) and by calculating monthly factors, given that normal value is also calculated by month. CEMEX refutes the petitioner’s argument that allocating the freight expenses was distortive because of the difference in shipping distances between the point of sale and the customer. CEMEX observes that the petitioner has raised this same argument in the last three administrative reviews and the Department has rejected the petitioner’s argument repeatedly, noting that differences in shipping distances are not necessarily distortive. Moreover, CEMEX points out that the Department verified the reporting methodology of the freight expenses and found no discrepancies. Concerning the arm’s-length test, CEMEX indicates that it reported freight expenses incurred by both affiliated and unaffiliated shippers and an analysis of such data indicates that the methodology does not affect the results. Thus, CEMEX contends, the freight expenses incurred by affiliated parties have no bearing on the margin. With respect to the sales-below-cost test, CEMEX maintains that the petitioner’s theory is incorrect. CEMEX explains that, in conducting its own cost test, it increased the freight costs conservatively by 500 percent which results in only 13 percent of home-market sales failing the cost test, a percentage still far below the Department’s 20-percent threshold. For these reasons, CEMEX maintains that its reported freight expenses are not distortive and the Department should continue to adjust normal value for freight expenses CEMEX reported. Department's Position: Based on submissions from the respondent and on our findings at verification, we determined that CEMEX's reported freight costs for Type I cement are reported on as specific a basis as is feasible, given CEMEX's accounting system. Furthermore, CEMEX’s calculation methodology provides a reasonable estimate of actual transaction-specific freight expenses. We disagree with the petitioner's assertions that CEMEX’s methodology for reporting freight costs skews the arm’s-length test. As the petitioner argues, examination of specific sales may result in differences between allocated and actual freight expenses. However, we determined during verification that combining freight expenses of both affiliated and unaffiliated customers should have no more than a marginal effect on the arm’s-length test. We also disagree with the petitioner’s assertion that CEMEX’s freight expenses affect the results of the sales-below-cost test adversely. With regard to the worksheet which the petitioner cites in its argument that CEMEX’s calculation methodology skewed the sales-below-cost test, the worksheet is merely an example of freight cost incurred and excludes other data submitted by CEMEX, including the data we examined during verification. The use of this worksheet does not provide a compelling reason to support the petitioner’s assertion. Thus, we have accepted CEMEX's claimed adjustment and have deducted its home- market freight expense for Type I cement from normal value for these final results. Comment 20: The petitioner contends that the Department should deny an adjustment to normal value for CDC’s reported freight expenses with respect to sales made by its affiliate, Construcentro. The petitioner argues that, as stated in its questionnaire response, for sales made by Construcentro, CDC reported an average freight expense that included non-scope merchandise. The petitioner argues that a recent judicial ruling requires the Department to disallow an expense that includes non-scope merchandise, citing SKF USA Inc. v. United States, 180 F.3d 1370, 1377 (Fed. Cir. 1999) (SKF USA). CDC argues that the Department deducted Construcentro’s home-market inland-freight expense from normal value properly. CDC argues that its allocation is the most specific possible, given its accounting system. CDC claims further that, because the majority of its total shipments were of subject cement, the freight expenses associated with its shipments is not inherently distortive. CDC observes that the Department made an adjustment for this expense in the last three administrative reviews where CDC used the same methodology. CDC asserts that the petitioner’s reliance upon SKF USA is misplaced because that decision concerned billing adjustments and cash discounts that are guided by rules concerning circumstance-of-sale adjustments which are not applicable to the Department’s treatment of movement expenses. Department's Position: As in prior reviews, we find that CDC reported its freight expenses to the best of its ability given its accounting system. The Department reviewed CDC’s freight expenses and calculation methodology as reported in CDC’s June 1, 1999, response and verified these expenses and its calculation methodology without discrepancies. See Home Market Sales Verification of Cementos de Chihuahua in the 1997/1998 Antidumping Review of Gray Portland Cement and Clinker from Mexico (August 6, 1999). We are satisfied that, consistent with 19 CFR 351.401(g), CDC’s allocation methodology does not cause inaccuracies or distortions. With respect to the petitioner’s reliance on SKF USA, this case relates to the Department’s practice prior to changes made to URAA and adoption of the Department’s new regulations. Since we found that CDC’s allocation methodology was reasonable, given the manner in which the expense was incurred, the accounting records that it maintains in its normal course of business, and the feasibility in which it can report this expense, we have made an adjustment to normal value for these freight expenses. c. Advertising Comment 21: The petitioner contends that the Department treated CEMEX’s home-market advertising expense incorrectly as a direct, rather than indirect, selling expense. The petitioner argues that the Department’s treatment of this expense as direct in nature is inconsistent with the Department practice and with the manner in which the Department has classified CEMEX’s home-market advertising expense in prior reviews of this order. The petitioner also argues that CEMEX has not met its burden of establishing that its advertising expense should be classified as a direct selling expense. Specifically, the petitioner points out that, in order for this expense to be classified as a direct selling expense, it should (a) bear a direct relationship to the particular sale in question, (b) be assumed on behalf of the customer’s customer, and (c) be associated with sales of scope merchandise sold during the POR. The petitioner asserts that the expense in question does not meet any of these criteria necessary to treat CEMEX’s home-market advertising as a direct selling expense. The petitioner contends that record evidence, including the verification report, demonstrates that this expense is directed toward the general public, promotes CEMEX’s corporate image and the company’s entire product line, and, thus, does not meet any of the criteria described above. The petitioner argues that, absent record evidence that establishes that this expense should be treated as a direct selling expense, the Department must treat the expense in question as an indirect selling expense for the final results of review. CEMEX contends that its advertising expense is a direct selling expense and the Department’s treatment of this expense was appropriate. CEMEX argues that the advertising expense in question was directed at its customers’ customer, end-users of the cement. CEMEX points out that the Department verified this expense and noted specifically in the verification report that this expense was incurred on a product-specific basis for advertising aimed towards end-users. CEMEX posits that the treatment of this expense as indirect in previous reviews does not necessarily deem the expense to be indirect in the instant review as every review stands on its own. Department’s Position: Section 351.410(d) of the regulations states that direct selling expenses are those that incurred or "assumed by the seller on behalf of the buyer...." We stated in our original questionnaire, at B-21, dated September 29, 1998, that the cost of advertising is the cost that the respondent incurred to advertise to its customer's customer. In addition, the advertising must be directed toward the specific product under investigation. See Roller Chain, Other Than Bicycle, From Japan, Final Results of Antidumping Duty Administrative Review, 58 FR 52264 (October 7, 1993). At verification, we found that CEMEX's advertising was product-specific and directed towards its customer's customers, end- users. See CEMEX's home-market verification report, at 14, dated July 23, 1999. We have considered such advertising expenses as direct expenses previously. See Notice of Final Results of Antidumping Duty Adminsitrative Review, Certain Pasta From Italy, 65 FR 7349, 7359 (Feb. 14, 2000) (the Department treated home-market advertising as a direct expense as it appeared to be targeted to respondent’s customer’s customer). Accordingly, we have treated CEMEX's reported home-market advertising expenses as a direct selling expense for these final results of review. d. Early-Payment Discounts Comment 22: The petitioner points out that, although CEMEX’s and CDC’s questionnaire responses indicate that the reported early- payment discounts were calculated by applying a discount percentage against the invoice value of their sales, numerous transactions were reported in excess of these amounts. The petitioner contends that the Department should deny granting an adjustment to normal value for those early-payment discounts that exceed the maximum percentage granted to customers as reported in the narrative portions of the companies’ responses. CDC agrees that its early-payment discounts should not exceed the maximum percentage reported in its responses and states that only a small number of its reported transactions exceeded this amount. However, CDC suggests that some of the values of early-payment discounts exceeding the maximum percentage may be due to rounding. CDC argues that, because the Department verified this discount claim, there is no reason to deny an adjustment for those early- payment discounts that exceed the maximum discount percentage indicated in its response. Rather, CDC requests that the Department limit the discount to three percent. Department’s Position: We have not denied an adjustment to normal value for those early-payment discounts that exceeded the maximum percentage as reported in CEMEX’s and CDC’s responses. While the amounts reported were overstated in some instances, we have corrected this where appropriate. Accordingly, we have limited those early- payment discounts that exceed the maximum percentage identified in CEMEX’s and CDC’s responses to their respective maximum discount percentages for these final results of review. e. Credit Expenses Comment 23: The petitioner argues that the Department should not use publicly available Mexican Interbank Average Interest (TIIP) to calculate CEMEX’s short-term interest rate in the credit-expense calculation. The petitioner contends that a verification exhibit demonstrates that CEMEX had short-term peso-denominated debt during part of the POR and, accordingly, the Department should use the weighted-average interest rate applicable to peso-denominated loans for the final results of review. CEMEX argues that it did not have any peso-denominated short-term loans outstanding during the POR. CEMEX points out that the interest rate included in the verification exhibit to which the petitioner refers represents the interest rate applicable to the current maturity of long-term debt. CEMEX states that, in the 94/95 and 95/96 reviews, the Department rejected the use of a short-term interest rate applicable to the current portion of long-term debt in the calculation of credit expense. CEMEX asserts that the TIIP rate is a more accurate reflection of the short-term borrowing rate in Mexico and meets the criteria set forth in the Import Administration Policy Bulletin 98.2 (February 23, 1998) in situations where the respondent has no short-term borrowings during the POR. Department’s Position: Although the verification exhibit to which the petitioner refers has positive amounts listed under a column heading labeled "Short Term Debt", this does not demonstrate that CEMEX had short-term borrowings outstanding during the POR. Based on CEMEX's records that we examined at verification, we found that CEMEX did not have short-term debt outstanding during the review period and that the loan amounts included in the verification exhibit represent the current portion of long-term debt. See memorandum from David Dirstine to The File concerning addendum to CEMEX's home-market verification of credit expenses, dated February 29, 2000, and CEMEX's home-market verification report, dated July 23, 1999. In accordance with our long-standing policy, when a respondent has no short-term borrowings during the review period, "we will use publicly available information to establish a short-term interest rate..." and, thus, can "accept 'external' information about the cost of borrowing in the relevant currency." The available information used to establish a surrogate interest rate should be reasonable, readily obtainable, and representative of "usual commercial behavior." See Import Administration Policy Bulletin Number 98.2 (February 23, 1998). We found that CEMEX's use of the TIIP rate in the calculation of its home-market credit expenses meets these three criteria. We also found that, in the absence of short-term debt during the instant review period, the TIIP is an appropriate surrogate interest rate to use in CEMEX's home-market credit expense. f. Other Adjustments Comment 24: The petitioner contends that CDC is not entitled to a post-sale price adjustment to normal value for the amounts it reported in its Other Adjustments (OTHADJGH) field. The petitioner asserts that the field includes expenses that do not relate specifically to merchandise under review. The petitioner also asserts that, with respect to one of the customers to which this adjustment was granted, the price adjustment was indirect rather than direct in nature because the quantity sold does not vary and this adjustment was not related to a particular sale, citing Zenith Elecs. Corp. v. United States, 77 F.3d 426, 431 (Fed. Cir. 1996), and Torrington v. United States, 68 F.3d 1347, 1353 (Fed. Cir. 1995). The petitioner also states that the sale related to the particular customer type pertains to the downstream sale and. thus, this post- sale price adjustment should be categorized as an indirect expense, citing Antifriction Bearings (Other than Tapered Roller Bearings) and Parts Thereof from France, et al.; Final Results of Antidumping Duty Administrative Reviews, 62 FR 54043, 54054 (October 17, 1997). CDC contends that the Department made an adjustment to normal value properly for the amounts reported in the OTHADJGH field. CDC asserts that the record demonstrates that this post-sale price adjustment was granted on sales of scope merchandise. CDC also states that this discount pertained to sales made between CDC and its ready-mix customers and does not pertain to downstream sales. CDC comments that the Department examined this discount at verification and found no discrepancies with the manner in which it was reported. For these reasons, CDC believes that, as in prior reviews of this order, the Department should grant an adjustment to normal value for the amounts reported in the OTHADJGH field. Department’s Position: As a general matter, the Department only accepts claims for other price adjustments as direct adjustments to price if actual amounts are reported for each transaction. Other price adjustments based on allocations are not allowable as adjustments to price unless they are based on a fixed and constant percentage of sales price. As demonstrated in CDC’s responses, CDC reported these adjustments on a transaction-specific basis. For example, CDC reported billing adjustments that corrected errors in the invoice price or adjusted prices for certain customers when a price change occurred or reported adjustment for unusual delivery conditions such as when a customer incurred additional delivery charges due to a difficult delivery location. These examples demonstrate that the adjustments were transaction-specific in nature. Another of CDC’s adjustments was calculated on a fixed-percentage basis and was also product- and customer-specific. As stated above the Department allows price adjustments when they are granted as a fixed and constant percentage of sales price on all transactions for which they are reported. The petitioner’s comments regarding a direct expense being an expense which "does not vary with the quantity sold," "is related to a particular sale," and is not based on "downstream sales" are misplaced. The issue is whether the adjustment is a post-sale adjustment to price rather than whether it is a direct or indirect expense. The criteria the Department uses to distinguish between direct and indirect expenses and adjustments to price are different. The amounts reported in the other-adjustments field pertain to adjustments to price, not expenses. Moreover, we examined these adjustments at verification and found no discrepancies. Therefore, consistent with our treatment of such adjustments in the last review, we have allowed these adjustments. 13. Financing Cash Deposits Comment 25: CDC disagrees with the Department decision to deny CDC’s adjustment to U.S. indirect selling expenses which CDC made to offset the cost of financing antidumping cash deposits. CDC argues that the Department has been inconsistent with its practice of and its reasons behind treating the cost of antidumping cash deposits as selling expenses. CDC points out that, since October 1997, the Department has followed the practice of denying the offset for antidumping cash deposits to its indirect selling expenses. CDC indicates that the CIT has upheld the position that the Department may allow respondents an adjustment to their indirect selling expenses and points to recent decisions made by the CIT with respect to NTN Bearing Corp. of America v. United States, Slip Op. 99-71 at 9 (CIT, July 29, 1999), and Timken Co. v. United States, Slip Op. 98- 92 at 6 (CIT, July 3, 1998). The petitioner argues that the Department should continue to deny CDC’s request because there is no basis under the statute to grant it. The petitioner states that CDC does not explain how an item that does not qualify as a U.S. indirect selling expense should nevertheless form the basis for an adjustment under the statute. It argues that, if CDC’s reported U.S. indirect selling expenses included the claimed cash-deposit opportunity cost, then it is conceivable that the Department should compensate the respondent by making an upward adjustment to CEP. However, the petitioner argues that CDC has not brought forward any evidence that its reported U.S. indirect selling expenses included the claimed cash deposit opportunity cost. The petitioner argues that CDC’s imputed claimed does not necessarily mean that it is included in the reported U.S. indirect selling expenses as an actual cost. The petitioner states that, in the absence of evidence that the reported U.S. indirect selling expenses were distorted in some way by cash deposits, the Department made the correct decision to deny CDC’s claim. Department’s Position: We have not adjusted to CDC's U.S. indirect selling expenses for imputed expenses which CDC claims are related to financing of cash deposits. The statute does not contain a precise definition of what constitutes a selling expense. Instead, Congress has given the Department discretion in this area. Current Department practice does not treat antidumping financial expenses associated with cash deposits as a direct, inevitable consequence of an antidumping duty order. Our denial of an adjustment to indirect selling expenses for the financing of cash deposits is consistent with our treatment of such expenses in the last two administrative reviews of this order in addition to reviews of other orders. See Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof from France, et al., 62 FR 54043, 54079 (October 17, 1997), and Tapered Roller Bearings and Parts Thereof, Finished and Unfinished, From Japan, and Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and Components Thereof, From Japan, 63 FR 63860, 63866-67 (November 17, 1998). As our position is unchanged from the two prior reviews, we adopt the discussion with respect to this issue in our 95/96 Final Results, 63 FR at 1278. 14.Duty Absorption Comment 26: CEMEX contends that the Department’s finding of duty absorption is unreasonable and contrary to law. CEMEX argues that the Department never analyzed whether duty absorption exists, as required by the URAA and the SAA. Instead, CEMEX maintains, the Department presumed duty absorption from the existence of a dumping margin. CEMEX argues further that the Department’s standard of allowing respondents to rebut the presumption of duty absorption by demonstrating that the unaffiliated purchasers paid the dumping duty is unreasonable because the Department has provided no guidance regarding this standard. CDC argues that it cannot provide an irrevocable agreement between RGPCC and the first unaffiliated customer that demonstrates that RGPCC did not absorb duties during the instant review period because the Department’s standard was developed after imports began entering the United States. CDC contends, however, that, based on information the Department examined at verification, the record establishes that RGPCC did not absorb duties during the instant review period. The petitioner argues that CEMEX and CDC have no foundation for objecting to the Department’s duty-absorption determination. The petitioner maintains that the Department’s determination is consistent with its practice. In addition, the petitioner asserts that the Department found correctly that CEMEX and CDC provided no evidence to rebut the presumption of duty absorption. The petitioner asserts that CEMEX’s suggestion that the Department’s duty-absorption presumption is not possible to rebut is entirely unfounded. The petitioner states further that there is no evidence on the record to indicate that CEMEX ever sought the Department’s guidance or attempted to rebut the Department’s presumption. The petitioner also rebuts CDC’s claim that the Department did not set a standard for proving that duties are not absorbed by an importer until after imports began entering the United States in the current administrative review. The petitioner argues that, like CEMEX, CDC knew that the duty-absorption issue would arise in this review and it had sufficient time and opportunity, prior to the initiation of the review, to develop the evidence demonstrating that it did not absorb duties. Department’s Position: We disagree with the respondents that our duty-absorption finding for the preliminary results of review was erroneous. Our final results of review indicate that both CEMEX and CDC absorbed duties. We made this determination in accordance with our statute, regulations, and practice. Section 751(a)(2)(A)(ii) of the Act and 19 CFR 351.213(j) indicate that we will consider dumping margins calculated during a review in determining whether duty absorption exists. As we explained in Antifrication Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From France, et al; Final Results of Antidumping Duty Administrative Reviews and Partial Termination of Administrative Reviews, 62 FR 54043, 54074 (October 17, 1997), the "existence of a margin raises an initial presumption that the respondent and its affiliated importer(s) are absorbing the duty." This is a reasonable presumption because the continued existence of dumping duties indicates that the producer and its affiliated U.S. importer have not adjusted their prices to eliminate dumping. If the producer has not set its price to the first unaffiliated U.S. customer high enough to eliminate dumping, it is reasonable to presume that the producer is also absorbing the dumping duties. The reasonableness of this presumption is also reflected in the SAA at 885, which states that "the affiliated importer may choose to pay the antidumping duty rather than eliminate the dumping..." (emphasis added). In sum, the existence of dumping is a reasonable presumption that the affiliated importer is absorbing duties. In addition, we disagree with CDC’s argument that we examined information at verification that establishes that its affiliated importer did not absorb duties during the instant review period. The information in the exhibit to which CDC refer was an exhibit that we used in the sales-trace portion of the verification and indicates the profit earned on the sale of one transaction. The exhibit does not indicate, however, that the unaffiliated purchaser absorbed duties nor is it representative of all sales made by the affiliated importer. Furthermore, we have articulated in numerous administrative reviews that, in order to satisfy our standard, the respondent must provide an irrevocable agreement between the unaffiliated purchaser and the affiliated importer that demonstrates that the unaffiliated purchaser will pay the antidumping duties. The verification exhibit does not satisfy this standard. Therefore, in accordance with the statute and regulations, we found that CEMEX and CDC absorbed duties during the instant review period. 14. PROMEXMA Sales Comment 27: CEMEX disagrees with the Department’s inclusion of sales made by PROMEXMA in the calculation of normal value. CEMEX argues that sales were made at a different levelof trade from the levels of trade identified in the Department’s level-of-trade memorandum and maintains that the Department’s level-of-trade memorandum did not address such sales. CEMEX argues that, if the Department includes sales made by PROMEXMA in its calculation of normal value, it must exclude sales made to PROMEXMA in order to avoid double-counting them. The petitioner asserts that, in this review, CEMEX has not provided the Department with any new information regarding PROMEXMA’s selling functions to warrant the Department’s departure from its determination on this issue in the 95/96 review. The petitioner contends that, contrary to CEMEX’s argument that PROMEXMA sales were made at a different level of trade, the Department should, as it did in the 95/96 review, determine that all of CEMEX’s home-market sales were at one level of trade. The petitioner argues that the Department’s decision in this review was based, as it was in the earlier review, on the fact that CEMEX did not establish that there is a substantial difference between the selling functions undertaken by CEMEX and those undertaken by PROMEXMA. The petitioner also argues that, even if CEMEX had demonstrated differences in home- market levels of trade, CEMEX has not attempted to provide a legal argument that such differences require the removal of PROMEXMA’s sales from the margin calculation. The petitioner argues that the statute merely requires that the Department attempt to compare U.S. and comparison-market transactions at the same level of trade. The petitioner contends that, while CEMEX maintains that it does not have a matching level of trade in the home market for PROMEXMA sales, it cannot invoke this provision as a basis for excluding PROMEXMA’s sales. Moreover, the petitioner argues that CEMEX has not cited any legal authority for preferring one allegedly non-matching level of trade to another allegedly non-matching level of trade. Department’s Position: In accordance with section 351.403(d) of the regulations, we do not calculate normal value based on the sale through an affiliated party if sales of the foreign like product by an exporter or producer to affiliated parties account for less than five percent of the total quantity or value of the exporter’s or producer’s sales of the foreign like product in the market in question. The sales CEMEX made to PROMEXMA account for less than five percent of the total quantity of CEMEX’s sales made in Mexico. Accordingly, for these final results of review, we have excluded from the margin calculation those sales made by PROMEXMA to unaffiliated customers while retaining CEMEX’s sales made to PROMEXMA. Since we have excluded from the margin calculation PROMEXMA’s sales, CEMEX’s argument that such sales were at a different level of trade is moot. Since PROMEXMA is an affiliated customer, we included all of the reported sales CEMEX made to PROMEXMA in the arm’s-length test. We excluded from the margin calculation those sales made to PROMEXMA that did not pass the arm’s-length test. 15. Contrucentro Employee Sales Comment 28: CDC contends that the Department did not eliminate employee sales of its affiliated reseller, Construcentro, for purposes of calculating normal value. CDC argues that, for the final results of review, the Department should exclude such sales from the calculation of normal value. The petitioner agrees with CDC. Department's Position: We agree that we should have eliminated Construcentro’s employee sales from the our calculations. For these final results of review, we have ensured that such sales are not part of our calculation of normal value. 16.Further-Manufactured Sales Comment 29: CEMEX and CDC contend that, while they met the Department’s special rule for further-manufactured sales in accordance with section 772(e) of the Act, the Department excluded information on further-manufactured sales from the assessment-rate calculation. CEMEX contends that this error resulted in an overstatement of the per-ton assessment rate. CDC requests that the Department include in its assessment-rate calculation sales of RGM’s further-manufactured cement that it sold to unaffiliated customers during the review period. CEMEX requests that the Department include all of its further-manufactured sales in the assessment-rate calculation and provides suggested methodology for the Department’s calculations. The petitioner agrees with both CEMEX and CDC that the Department’s assessment-rate calculation did not take further-manufactured sales into account and requests that the Department include such sales into its assessment-rate calculation for the final results of review. It provides a methodology to implement this recommendation which varies slightly from CEMEX’s suggested methodology. Department's Position: We disagree with the respondents and the petitioner and have not incorporated their suggestions into our final calculations. Instead, we have calculated a per-unit duty-assessment value using the reviewed sales and will instruct the U.S. Customs Service to apply this value to all merchandise entered during the review period regardless of whether value was added to it after importation. This approach will not cause any distortions with respect to the merchandise that was entered and subsequently further manufactured because, as indicated in CEMEX’s and CDC’s responses, all merchandise, regardless of whether it was further manufactured, was identical. As we mentioned above, the statute is silent with respect to the methodology by which we are to calculate assessment rates. See Federal Mogul Corp. v. United States, 918 F. Supp. 386, 404 (CIT, 1996). Nonetheless, our general practice, while not challenged in prior proceedings, has been to calculate an assessment rate, or assessment value as applicable, based on the sales that were reviewed during the period of review and apply this rate, or value, to all entered merchandise, including those circumstances where the company met our special rule for further-manufactured merchandise. 18. Ministerial Errors a. Model-matching Comment 30: CEMEX and CDC state that the Department compared all U.S. sales to a weighted-average normal value of CDC's sales. CDC contends that, for purposes for making price-based comparisons, the Department should calculate normal value on a company-specific basis, as the Department did in prior reviews of this order. CEMEX requests that the Department compare all U.S. sales to the weighted-average normal value of both CEMEX and CDC. The petitioner agrees that the Department matched CEMEX's and CDC's U.S. sales only to the weighted-average normal value of CDC incorrectly and requests that the Department correct the error. Department's Position: We agree with the respondents and the petitioner that, in our preliminary calculations, we neglected to include CEMEX's home- market sales in our calculation of weighted-average normal values. Consistent with our decision to collapse CEMEX and CDC, we have made the necessary modifications to the calculations to ensure that both CEMEX's and CDC's U.S. sales are compared to the respective weighted- average normal value of home-market sales which both CEMEX and CDC made during the POR. Furthermore, we do not believe that we should calculate normal value on a company-specific basis. Calculating normal value on a company-specific basis is inconsistent with our decision to collapse CEMEX and CDC and would, in essence, result in a company-specific margin analysis. In addition, our decision to calculate a single weighted-average normal value for both companies was not influenced by prior reviews of this order. Rather, in accordance with 19 CFR 351.104, our decision was based on and limited to the facts on the record in this administrative review. b. CDC’s Sales to Employees Comment 31: The petitioner contends that the Department did not eliminate CDC’s sales to employees from its calculations, as it intended, due to the spacing of the identifier in the customer- category field. For the final results, the petitioner requests that the Department insert language into its calculations to compensate for spacing irregularities in the customer-category field. CDC agrees with the petitioner and offers language to correct the problem. Department’s Position: We agree with the petitioner and CDC and have modified our calculations to ensure that we eliminated CDC’s sales to employees from the calculation of normal value. c. U.S. Direct Selling Expenses Comment 32: The petitioner argues that, in calculating the foreign unit price expressed in U.S. dollars (FUPDOL) for CDC’s export-price sales, the Department erred by adding U.S. direct selling expenses valued in dollars per short ton to the sum of normal value, DIFMER, and packing expenses valued in dollars per metric ton. The petitioner requests that the Department convert the relevant U.S. direct selling expenses to metric tons so that all of the expenses in the FUPDOL calculation are on a metric-ton basis. Department’s Position: We agree have made the appropriate conversion. Recommendation Based on our analysis of the comments received, we recommend adopting all of the above positions and adjusting all related margin calculations accordingly. If these recommendations are accepted, we will publish the final results of review and the final weighted- average dumping margins for all reviewed firms in the Federal Register. AGREE____DISAGREE____ ________________________ Robert S. LaRussa Assistant Secretary for Import Administration _________________________ Date