(65 FR 16877, March 30, 2000) A-583-833 Investigation PUBLIC DOCUMENT MEMORANDUM TO: Richard W. Moreland Acting Assistant Secretary for Import Administration FROM: Susan H. Kuhbach Acting Deputy Assistant Secretary for Import Administration SUBJECT: Issues and Decision Memorandum for the Investigation of Certain Polyester Staple Fiber from Taiwan ------------------------------------------------------------------------ SUMMARY We have analyzed the briefs and rebuttals of interested parties in the investigation of Certain Polyester Staple Fiber (PSF) from Taiwan. 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 Issues section of this memorandum. Below is a complete list of the issues in this investigation for which we received comments and rebuttals by parties: ISSUES APPLICABLE TO BOTH RESPONDENTS Comment 1: Adverse Facts Available Comment 2: Errors in Computer Programing ISSUES SPECIFIC TO FAR EASTERN TEXTILES, LTD. A. General Issues Comment 3: Pre-verification Revisions and Minor Errors Comment 4: Product Coding Comment 5: Antibacterial and Flame-Retardant Products B. Sales Issues Comment 6: Movement Expenses and Bank Charges on U.S. Sales Comment 7: Commissions Comment 8: Sales to Affiliate Comment 9: Verification of Surprise Sales C. Cost of Production/Constructed Value Issues Comment 10: Major Inputs-PTA Comment 11: Major Inputs-EG Comment 12: Material Costs - Scrap Consumption Comment 13: Foreign Exchange Gains and Losses Comment 14: G&A Expenses III ISSUES SPECIFIC TO NAN YA PLASTICS CORPORATION A General Issues Comment 15: Mis-coding of Regenerated and Virgin Products Comment 16: Recoding of Sale B Sales Issues Comment 17: Exchange Rates Comment 18: Inland Freight-General Issues Comment 19: Inland Freight-Adjustment for Affiliated Expenses Comment 20: Inland Freight-Additional Freight to Factory Comment 21: Inland Freight-Affiliated Transactions at Arm's Length Comment 22: Indirect Selling Expenses Comment 23: Imputed Credit Expenses on Certain U.S. Sales Comment 24: Bank Charges Comment 25: Commission and Marine Insurance Comment 26: U.S. Short-Term Interest Rate Comment 27: Home Market Short-Term Interest Rate C. Cost of Production/Constructed Value Issues Comment 28: Recovery of Inputs Comment 29: Exchange Gains Comment 30: Minor Verification Corrections Comment 31: Product-Specific Costs Comment 32: General and Administrative Cost Comment 33: Long-term Interest Income Comment 34: Packing Expenses Comment 35: Unreported Costs Comment 36: Revised Yields Comment 37: Positive Yields Comment 38: Scrap Credit Comment 39: Inputs from Affiliates BACKGROUND On November 8, 1999, the Department of Commerce (Department) published the preliminary determination in this investigation.(1) The period of investigation (POI) is April 1, 1998 through March 31, 1999. We invited parties to comment on our preliminary determination. At the request of certain interested parties, we held a public hearing on March 10, 2000. DISCUSSION OF ISSUES I. ISSUES APPLICABLE TO BOTH RESPONDENTS Comment 1: Adverse Facts Available Petitioners' Comments: The petitioners(2) state that in "the guise of responding to limited, supplemental questionnaires issued by the Department," the respondents have taken the opportunity to submit completely revised sales and cost databases on six different occasions in the case of Far Eastern Textiles, Ltd (FETL) and on five occasions in the case of Nan Ya Plastics Corp. (Nan Ya), with a certification of accuracy accompanying each submission. The petitioners further argue that each one of these new databases contained major revisions to previously submitted data that were neither requested by the Department nor explained by the respondents. According to the petitioners, the information currently on the record for both companies bears virtually no resemblance to the original questionnaire responses. Moreover, the petitioners continue, even after the numerous revisions to their submitted data, the Department nevertheless identified still more, significant aspects of both respondents' revised databases that still did not verify.(3) With regard to FETL, the petitioners cite five instances, in particular, where FETL's actions failed to satisfy the requirements of section 782(e) of the Act: (1) FETL failed to submit accurate control numbers; (2) FETL failed to provide correct cost of production data; (3) FETL's sales and cost data are so incomplete that they cannot serve as bases for reaching a final determination; (4) the verification demonstrated that FETL had not acted to the best of its ability in providing the requested information; and (5) these deficiencies cannot be corrected without imposing undue difficulty on the Department. With regard to the other respondent, during Nan Ya's sales verification the Department discovered that the reported dates of sale (based on the sales confirmation date) for several of the examined sales were incorrect. Judging from the frequency throughout the U.S. sales database of a single, apparently erroneous date, the Department further noted that it appeared that the dates of sale for many additional sales observations that the Department was not able to examine at verification likewise were probably incorrect and that, in many instances, the incorrect dates were potentially off by up to several months. Because the date of sale is used to determine the universe of sales to be included in the margin calculation, there is "real doubt," according to the petitioners, that the database reported by Nan Ya is accurate and complete in containing all sales that fall within the POI. The petitioners argue that this error alone justifies the use of facts otherwise available and that the Court of International Trade has expressly upheld the Department's reliance on best information available where the Department was unable to verify the total quantity of U.S. sales identified in the response.(4) The petitioners note several other things that they believe further indicate that Nan Ya's sales listings are incomplete. To begin with, according to the petitioners, the Department did not verify that certain sales included in Nan Ya's original home market sales listing, but deleted from subsequent versions were sales of merchandise falling either outside the scope of this investigation or outside the period of investigation. Without verified data showing that these sales should not have been reported as home market sales by Nan Ya, the petitioners question the accuracy and completeness of Nan Ya's most recently submitted database. Second, the petitioners claim that Nan Ya's failure to report home market sales of colored PSF also renders the sales listing incomplete. Even though Nan Ya sold only non-colored PSF in the United States during the POI, the petitioners state that home market sales of colored products could be possible matches for price-to-price comparisons since color was not part of the matching criteria. As possible evidence of further under- reporting by Nan Ya, the petitioners look to the Cost Verification Report where it lists certain products for which there were no reported costs. Finally, the petitioners claim that information submitted at verification by Nan Ya additionally supports the incompleteness of its reporting since total quantities and values of its sales listings do not comport with information in the reconciliation chart used at verification to link the sales listings to Nan Ya's general ledger and financial statement. The petitioners contend that the Department has been "most accommodating to FETL and Nan Ya in giving them every opportunity to correct submissions for the record." However, consistent with sections 776 and 782 of the Act, the petitioners argue that the Department should not rely upon either respondent's unverifiable and deficient responses but should resort instead to facts otherwise available. Moreover, given the failure of both respondents to cooperate by repeatedly submitting repeated database revisions with no justification or explanation, and given the Department's inability to verify significant aspects of the most recent submission, the petitioners urge the Department to find that the use of adverse inferences is warranted in this case for both FETL and Nan Ya. The appropriate adverse facts available rate, the petitioners suggest, is 70.70 percent, the highest rate in the petition. FETL's Comments: In response, FETL argues that the petitioners' request for the Department to apply adverse facts available in determining FETL's dumping margin is unsupported by the evidence on the record, by the Department's instructions, and by the history of this investigation. Moreover, the petitioners' arguments are untimely as they have failed to raise certain objections until the last stage of this investigation. Specifically, FETL argues that it did submit accurate control numbers in the manner and time-frame requested by the Department. The petitioners also have exaggerated, FETL continues, the scope and effect of certain relatively minor errors found at verification in the reporting of certain sales and expense data. FETL stated that for those reporting errors that were identified at verification, there is readily-available data at the disposal of the Department which could serve as reasonable partial facts available for any missing or incorrect data. Nan Ya's Comments: Nan Ya counters that the petitioners' arguments in this regard are unsupported by the record and, thus, should be rejected for the following reasons. First, Nan Ya argues, the Department fully verified that Nan Ya's reported home market and U.S. sales were complete. Second, according to Nan Ya, the petitioners' arguments regarding its incorrect reporting of sales confirmation dates are irrelevant to the issue of sales completeness. In fact, other reported dates (e.g., dates of shipment, invoice) for these observations demonstrate that the actual confirmation dates of these sales were within the POI. Also, for all of the observations with possibly incorrect sales confirmation dates and for which sales documentation was examined, the Department verified that the correct date of sale was later in time, and there is no record evidence suggesting that the actual dates for these sales were prior to the reported date. Moreover, Nan Ya argues that there is no evidence to support the petitioners' contention that it is quite likely that other sales not reported could fall within the POI. To the contrary, Nan Ya contends, the Department verified that all other confirmation dates had been correctly reported and that all sales that were ordered prior to the POI but shipped during the POI were correctly reported. Nan Ya concludes that, contrary to the petitioners' assertions, errors in sales dates do not affect the completeness of its sales reporting. Rather, this error, Nan Ya suggests, only affects the selection of the proper exchange rate. (See Exchange Rate comment for Nan Ya below.) Therefore, applying adverse total facts available to Nan Ya is unwarranted. Furthermore, Nan Ya continues, the cases to which the petitioners cite in making their arguments for adverse facts available are irrelevant to this instance because those cases pertain to situations where the respondents failed to report certain sales. In particular, the respondent in Circular Welded Non-Alloy Steel Pipe from Brazil failed to report sales in a purchase order. Likewise, the respondents in Certain Circular Welded Non-Alloy Steel Pipe from Mexico did not report certain home market sales. With respect to the sales that were in the original home market sales listing but subsequently deleted, Nan Ya points out that prior to verification as part of its October 26, 1999 submission, it provided copies of the invoices for these sales showing that they were sales of products outside the scope of this investigation. Similarly, Nan Ya clarifies that while it did not at first report sales of colored PSF in the home market, it did subsequently report these sales as part of its pre-verification submission of corrections. Nan Ya then states that there is evidence on the record indicating that the products for which it did not report costs were sold only to third countries. As for the petitioners' allegation that the information in the reconciliation chart used at verification does not support the completeness of its sales reporting, Nan Ya states that the petitioners offered no evidence to support their allegation. Department's Position: We do not find the use of total facts available with respect to either FETL or Nan Ya to be warranted in this case. Though in specific instances we have resorted to partial facts available for information that we found to be incomplete or inaccurate (see discussion of specific issues below), we have found no grounds, consistent with the statute, on which to reject the reported data of either respondent in its entirety. The Act is clear about the circumstances under which the Department may resort to facts available. Specifically, section 776(a)(2) of the Act states that the Department shall apply facts available when an interested party (1) withholds information that has been requested by the Department, (2) fails to provide such information in a timely manner or in the form requested, (3) significantly impedes a proceeding under the antidumping statute, or (4) provides information that cannot be verified. However, the Act also states that the Department shall not decline to consider information that is submitted by an interested party and that is necessary to the determination but does not meet all of the applicable requirements established by the Department if (1) the information is submitted by the deadline established for its submission, (2) the information can be verified, (3) the information is not so incomplete that it cannot serve as a reliable basis for reaching the applicable determination, (4) the interested party has demonstrated that it has acted to the best of its ability in providing the information and meeting the requirements established by the Department with respect to the information, and (5) the information can be used without undue difficulty.(5) In view of these statutory provisions governing the use of facts available, we turn first to petitioners' allegations regarding FETL. The petitioners state that throughout the course of this investigation, FETL has revised its reported data on several occasions, and that "the information currently of record bears virtually no resemblance to that originally submitted."(6) Moreover, the petitioners continue, " . . . every one of these new databases contained major revisions to previously submitted data that were neither requested by the Department nor explained by FETL . . . ."(7) We do not agree with this characterization of the revisions in FETL's data, for the following reasons: First, with regard to the allegation that certain of these revisions were unsolicited by the Department, it is important to note that FETL's questionnaire and supplemental responses were all submitted within the time limits established by the Department. FETL's submission prior to verification of corrections to these responses were likewise within the time limits established under the Department's regulations(8) and, therefore, were also timely. Moreover, the Department's acceptance of and reliance on these revised data are reasonable. The statute provides that whenever the Department determines that foreign merchandise is being sold in the United States at less than fair value, "there shall be imposed upon such merchandise an antidumping duty . . . in an amount equal to the amount by which the normal value exceeds the export price (or the constructed export price) for the merchandise.(9) Consistent with this statutory directive, the U.S. Court of Appeals for the Federal Circuit and the U.S. Court of International Trade have emphasized that the purpose of the antidumping statute is to determine margins as accurately as possible.(10) It would, therefore, be incongruous with the express intent of the statute to rely on data that are clearly inaccurate by a respondent's own admission. Accordingly, it is the Department's general practice to allow respondents to revise their data upon identification of errors when such revisions are done in a timely manner.(11) Timely revisions to respondents' submissions are neither unusual nor inconsistent with the Department's standard practice. Contrary to the petitioners' allegation, many of the revisions in FETL's databases were done at the request of the Department and were, in fact, responsive to concerns raised by the petitioners. For instance, in their July 30, 1999 submission, the petitioners claim that " . . . FETL's responses are incomplete and incorrect, and require additional information be placed on the record." In a subsequent August 19, 1999 letter (at 1), the petitioners requested that the Department instruct the respondents "to resubmit their U.S. and home market computerized sales listings based upon the date of order confirmation as the date of sale." On November 10, 1999, after the preliminary determination, the petitioners again submitted another request asking that the Department solicit even more information from FETL.(12) Clearly, FETL should not be penalized for submitting information at the direction of the Department and which was also responsive to the petitioners' concerns. The petitioners further allege, however, that many of the revisions to FETL's databases were never explained by FETL. The petitioners first made this argument in a September 23, 1999 letter to the Department. In response to a subsequent request by the Department to "provide a complete and detailed narrative explanation" of its previous database revisions,(13) FETL submitted an extensive narrative with supporting attachments detailing each of its prior revisions.(14) In a memorandum analyzing FETL's explanation of its revisions, the Department found that many of the revisions appeared to have a relatively small impact on the dumping margin calculation.(15) We further found that, "FETL has only provided a summary of certain revisions and, therefore, the full extent of the changes is not entirely clear in every instance. However, FETL's explanation overall provides a sufficient basis for understanding the basic changes incorporated into the revised databases."(16) Thus, for the preliminary determination, we relied upon FETL's revised data on the finding that FETL had adequately explained its revisions to date. Likewise, we find that all of FETL's additional revisions subsequent to the preliminary determination were explained in sufficient detail. Moreover, even after significant revisions, continue the petitioners, the Department still identified serious errors in FETL's revised databases at verification. Errors discovered at verification are not, however, automatic grounds for the rejection of the whole of a respondent's reported data. As detailed in subsequent comments below, the errors discovered during the verification of FETL's sales and costs were limited in scope and their impact on any potential dumping margin was small. In place of any missing data, or data that were found to be incorrect at verification, the Department can use as partial facts available other readily available data on the record, as it frequently does in instances where data "gap-fillers" are necessary. Thus, FETL's data can be used without undue difficulties. Moreover, pursuant to section 782(e) of the Act, we do not find that FETL's information is so incomplete that it cannot serve as a reliable basis for reaching a final determination. To the contrary, FETL's data-in all essential respects-were verified to be sufficiently complete and accurate. Contrary to the petitioners' implied disbelief that "typos" and "arithmetic errors" were the reason for many of FETL's revisions, we find FETL's explanation to be credible. In its October 28, 1999 submission (at 2), FETL explained: Based on the hard copies, FETL manually typed the data into separate U.S. and home market sales databases to prepare the reported U.S. and home market sales listing for this response. The computer-generated hard copies however include sales of polyester staple fibers of less than three denier. These hard copies of home market sales also include both home market sales and export sales through trading companies in Taiwan. FETL thus manually excluded these non- subject merchandise sales from both U.S. and home market sales listing, excluded indirect export sales from the home market sales listing, and added indirect export sales for the U.S. market into the U.S. sales listing. FETL officials further detailed at verification the difficulties they had faced in compiling their databases: Company officials explained their methodology for identifying the sales to be reported in their sales databases. Officials started by noting that they had great difficulty at first in reporting sales dates according to purchase order dates for export sales because the purchase order date is not normally a key date that they track in their main accounting system. Moreover, company officials had difficulty searching and sorting their databases according to the matching criteria used in this investigation because at least some of the criteria are not distinctions that FETL tracks on a detailed basis through its accounting system. Deriving a completely accurate home market and U.S. sales database ultimately required a manual, transaction-by-transaction review of the documentation for each sale. Officials stated that some of the sorting and selection of sales was, however, accomplished through use of their computer system. They continued by showing us the approach they adopted for excluding certain sales based on product characteristics and other factors. We noted that the methodology adopted appeared consistent and reasonably comprehensive given the structure of FETL's computerized accounting system.(17) It is clear that FETL officials went to great lengths to provide their data in the form and manner requested by the Department. Although their initial submissions contained several errors, FETL continued throughout the proceeding to re-examine its data, and brought any subsequently identified deficiencies to the attention of the Department in a timely manner. Moreover, there is no indication that the errors in FETL's submissions, including the errors identified at verification, comprise a pattern of misleading or contradictory responses on the part of the respondent. Neither is there any evidence to suggest that FETL officials willfully withheld pertinent information from the Department. Rather, while some of the errors identified appeared to be in the respondent's favor, others were adverse to FETL's interests. Finally, the petitioners have further suggested that the Department did not adequately follow through on "the pledge in its preliminary determination to thoroughly check the completeness of respondents' sales reporting and in particular the database revisions."(18) The petitioners refer to the Preliminary Determination where the Department stated (at 60775), "At verification, we will closely examine changes between the most recently submitted data sets and prior data sets including, but not limited to: cost allocations, materials usage, unit prices, and expenses. We will also thoroughly check the completeness of respondents' sales reporting." The petitioners' allegation is wrong, and is flatly refuted by the verification reports. For example, FETL's Sales Verification Report (starting at 28) clearly establishes that the Department closely examined FETL's exclusion of third- country and non-subject merchandise sales from its September 20, 1999 sales databases. The report continues by detailing our examination of the revisions to freight expenses in the databases submitted prior to the preliminary determination. It is also clear from the verification report that we closely examined the completeness of FETL's sales reporting (e.g., at 6) and FETL's methodology for coding its merchandise (e.g., at 8). These were all areas in which the petitioners alleged that "extensive changes" had been made.(19) Thus, our verification procedures were fully consistent with the above statements from the Preliminary Determination, and reasonably addressed the petitioners' concerns in this regard. As with FETL, the petitioners have also taken issue with the information submitted by Nan Ya during the course of this investigation by stating that: Nan Ya has revised its reported data on so many occasions that "the information currently of record bears virtually no resemblance to that originally submitted;" and that the new databases contained major revisions to previously submitted data that were neither requested by the Department nor explained by Nan Ya. These characterizations of the revisions to Nan Ya's data are incorrect for the following reasons: As for the petitioners' complaint against Nan Ya's unsolicited revisions to its submitted data, we note that all of these changes were submitted prior to verification and were timely within the guidelines set forth in the Department's regulations. As stated above regarding similar submissions of data by FETL, the acceptance by the Department of such changes is not unusual nor inconsistent with the Department's standard practice. We further note that many of the revisions in Nan Ya's data were in response to concerns about that data raised by the petitioners. For instance, in their letter of July 14, 1999, commenting solely on Nan Ya's original Section A response the petitioners make numerous requests to the Department for further information from Nan Ya, including: financial and other information on all its affiliated parties; details on major inputs; explanations on a large number of items in its chart of accounts, as well as information on increases and decreases in these items during the POI; descriptions of share transactions and guaranteed endorsements between Nan Ya and its affiliates; detailed information on financial transactions between Nan Ya and affiliated companies; and, further information regarding its sales and distribution processes. In the same submission, the petitioners specifically request that Nan Ya be required to alter its submitted data by changing its date of sale methodology for both home market and U.S. sales, and by designating/including sample sales in its sales listings and providing supporting documentation on the same. Similar and even more extensive requests for information from Nan Ya were received from the petitioners on August 6, 1999, regarding Nan Ya's original responses for sections B, C and D of the questionnaire. In this submission, the petitioners specifically requested that Nan Ya be required to adjust its submitted data as follows: eliminate two forms of antibacterial finishes as separate finish codes; change payment dates and credit periods; shift expenses between indirect, and general and administrative; change the calculation of general and administrative expenses; alter the calculation of inventory carrying costs; segregate packing labor and overhead expenses; include U.S. inland freight expenses where necessary; revise the interest expense calculation. The petitioners' submission of August 6, 1999, also requested many detailed explanations, supporting documents and calculation worksheets, and hinted that further changes to Na Ya's data may be required based on the content of the information to be received. Some of the further changes that were envisioned by the petitioners included: a change in the coding for matching purposes, especially regarding regenerated versus virgin material, based on explanations regarding Nan Ya's product codes; the possible elimination of high shrinkage and PE/PET low melt products as separate composition codes; and the use of price to broker instead of price to customer for certain U.S. sales. Also, in their letter of August 19, 1999, as mentioned above regarding FETL, the petitioners requested that the Department instruct Nan Ya to resubmit its U.S. and home market sales listings based upon the date of order confirmation as the date of sale. As can be surmised from the foregoing, the petitioners clearly expected that Nan Ya would be providing additional information following its original questionnaire responses and that they wanted to influence the Department's decision on what information to request from Nan Ya. The petitioners further alleged on October 13, 1999, that many of the revisions to Nan Ya's databases were not explained adequately. On October 19, 1999, the Department asked Nan Ya to explain the changes it had made to its databases and on October 25, 1999, Nan Ya responded. In a memorandum dated October 25, 1999, the Department analyzed Nan Ya's explanations and found that the revisions were adequately explained and appeared to be reasonable. Accordingly, we used Nan Ya's revised data for the preliminary determination. We also find that Nan Ya's additional revisions subsequent to the preliminary determination were explained in sufficient detail. We note that, with the exception of minor errors discovered during verification, the information in Nan Ya's revised databases was supported by the findings of verification. It was also made clear at verification that Nan Ya expended large amounts of time and effort in putting its data into the form and manner requested by the Department. As noted in the verification report, the way in which Nan Ya records information in its normal accounting system does not correspond to the date of sale methodology adopted by the Department nor to the Department's model matching criteria. In addition, we found at verification that Nan Ya does not track its expenses on a transaction-specific basis and had to manually calculate per-unit amounts before preparing its data. Although Nan Ya's initial submissions contained errors, the company continually examined its data during the course of the proceeding in an effort to provide accurate data to the Department in a timely manner. The petitioners' arguments regarding the respondents' multiple submissions of data are based upon the requirement that parties to an antidumping proceeding certify the accuracy and completeness of the factual information they submit. In essence, the petitioners argue that an original questionnaire response should be absolutely complete and accurate, i.e., no further revision of that data should be required to calculate a dumping margin, because the submitting respondent certified its accuracy and completeness In response, we point out that the statute envisions the correction of submitted data.(20) Indeed, in the overwhelming majority of cases, supplemental questionnaire responses are required. The informational requests by the Department are quite demanding and specific. Given the foregoing, it is not unusual, if not common, for respondents to make errors in interpreting the Department's requests and in submitting their data. We note that the certifications we require on factual information, according to 19 CFR 351.303(g) state that the company official presenting the information must certify that, " the information contained in this submission is, to the best of my knowledge, complete and accurate." We interpret the certification to mean that the submitting party is not knowingly submitting false information and that, as far as that party knows at the time, the information is complete. We do not interpret the certification to imply that the information being submitted must be guaranteed to be complete and totally free of errors. Beyond the argument that an original questionnaire response should be fully ready for a dumping calculation, the petitioners then take issue with the respondents' corrections to their databases in absence of a specific request by the Department. For example, in their submission of September 23, 1999, the petitioners state, " . . . [the respondents] took advantage of the agency's request for specific data by substantially altering the database originally submitted in numerous respects that had nothing to do with the agency's requests for information." This argument by the petitioners implies that respondents, upon discovering errors in their submitted data, should not be allowed to correct the errors unless the Department specifically requests the correction. The Department's goal in conducting an antidumping proceeding is to calculate the most accurate dumping margins possible. Since respondents are in possession of the data needed for the calculations, the Department may not be aware of inaccuracies in the reported information. Therefore, relying only upon Department requests for specific information may not, and most probably would not, give rise to the most accurate data possible. As can be seen from the foregoing, such a policy would be contrary to the goal of calculating accurate dumping margins. Accordingly, we find the respondents' self-correction of data to be appropriate. We also find, however, that such changes should be explained. While many of FETL's and Nan Ya's corrections were not immediately explained, they were eventually clarified, and verified with minor exceptions. The petitioners also state that, at verification, the Department failed to examine closely Nan Ya's changes to its data over time. The petitioners' main complaint in this regard seems to be centered on a relatively small amount of sales that were in Nan Ya's original database, but were deleted from subsequent databases. Nan Ya, however, provided sufficient documentation prior to verification showing that these were indeed sales of merchandise not covered by the scope of this investigation. In particular, Nan Ya provided invoices and shipping notices showing that these were sales of products with a denier 1.2 (less than the minimum denier of 3 covered by the scope of this investigation) and, therefore, should not have been reported. As a result, given the time constraints at verification, we found it unnecessary to pursue this issue any further. Prior to verification, the petitioners made a submission detailing a large number of items they were specifically interested in having the Department address at verification. The topics of interest to the petitioners, as specified in the pre-verification submission of January 7, 2000, were: product coding of virgin and regenerated materials; pricing of conjugated versus single- component PSF; finish coding of antibacterial products; transportation provided by affiliated companies; sales in the original database but subsequently deleted; payments terms for U.S. sales; the calculation of indirect selling expenses; factory-to-port freight and international freight expenses on U.S. sales; sales processes for U.S. sales to and through brokers; and, the calculation of inventory carrying expenses. All of these topics were covered at length at verification, with the exception of the sales in the original database but subsequently deleted, as explained above. With regard to the petitioners' arguments regarding errors in Nan Ya's reported dates of sale, we disagree that there is real doubt that the U.S. sales database reported by Nan Ya is accurate and complete in containing all sales that fall within the POI. To the contrary, for the reasons cited in a separate Dates of Sale memorandum(21) on the subject, we find that Nan Ya's U.S. sales database is sufficiently complete and reliable for the purposes of our analysis and, as such, should not be summarily rejected because of certain erroneous dates of sale. This finding would be consistent with the fact that in verifying the completeness of sales quantity and value at verification, we found no major errors.(22) Our discussion of the impact on these dates of sale errors on exchange rate conversions is included in Comment 17 below. Aside from Nan Ya's misreporting of its dates of sale and the sales in its original database but subsequently deleted, as mentioned above, the petitioners raise other issues regarding the completeness of Nan Ya's reporting in support of the use of adverse facts available, specifically (1) Nan Ya's failure to report home market sales of colored PSF; (2) the existence of certain products for which there were no reported costs; and (3) total quantities and values in Nan Ya's sales listings do not comport with information in the reconciliation chart used at verification to link the sales listings to Nan Ya. Regarding the small number of sales of colored PSF in the home market, these sales were only reported at the commencement of verification. However, the information provided included prices, quantities, freight amounts, products codes, and was complete with the exception of information on the finish of some of these products. An analysis of Nan Ya's sales of colored PSF shows that only a few if any of these transactions would be included in the calculation of weighted-average home market prices for comparison to U.S. sales. Because of the minuscule quantity represented by these sales, they would have little or no effect on the overall estimated weighted-average dumping margin. As a result, we have decided not to include these sales in our final analysis. As for the petitioners' claim that Nan Ya failed to report costs on certain products, we concur with Nan Ya that these products were sold only in third countries during the POI. Accordingly, Nan Ya was correct in not reporting costs for these products. Finally, the petitioners point to differences between the amounts in the reconciliation chart and the amounts in the sales listings as evidence that the Department could not verify the completeness of Nan Ya's sales. As stated in the verification report, the computer language Nan Ya used to generate the amounts in the reconciliation chart for sales of merchandise within the scope of this investigation was overly inclusive and picked up sales of merchandise that were not in the scope. While some of the erroneously included categories of sales were explicitly mentioned in the report, others were not (e.g., fibers 10 to 18 denier that are cut to lengths of 6 to 8 inches). We note that the technique used to verify completeness is a valid and commonly used technique when a respondent does not keep its books and records in a manner which corresponds to the date-of-sale methodology employed by the Department, or to the scope of the investigation - as was the case here. The verification of completeness involved examining many transactions among the thousands of sales made by Nan Ya of both subject and non-subject merchandise. Since the number of transactions with possibly incorrect dates of sale was relatively small compared to the totality of Nan Ya's sales, it is not likely that the completeness exercise would have stumbled across one of these transactions. We found all transactions examined as part of the completeness exercise to be properly categorized as either U.S. sales of PSF, home market sales of the same, or sales not subject to the investigation, with the exception of two transactions that were improperly included in the sales listings. As for the transactions with the improper dates of sale, as stated in the Dates of Sale Memorandum, we note that any reasonable construction of the proper date of sale leaves these transactions in the POI, with only minor possible exceptions. Thus, we are satisfied that Nan Ya's sales databases are complete. In light of the statutory provisions regarding the use of facts available and the Department's practice regarding the acceptance of revised data, we do not find that the application of total facts available for either FETL or Nan Ya is warranted. With respect to FETL, none of the errors in its data are of such scope as to render the data in their entirety useless for purposes of calculating an estimated dumping margin. Likewise, the error in Nan Ya's reporting of its dates of sale is not so significant as to warrant rejecting Nan Ya's data in its entirety. Moreover, as stated above, the other issues raised by the petitioners regarding the use of total facts available for Nan Ya (i.e., its failure to report home market sales of colored PSF, products for which there were no reported costs and the invalid nature of the completeness test) are either without merit or have no significant impact on our analysis. Comment 2: Errors in Computer Programming Petitioners' Comments: According to the petitioners, the computer program the Department used in the Preliminary Determination contained the following three errors that should be corrected: (1) imputed credit expenses were not included in the circumstance-of-sale adjustment to normal value; (2) U.S. packing expenses were erroneously converted, and (3) the Department did not differentiate between grades in the cost test as it had in the comparison of sales. Respondents' Comments: Neither FETL nor Nan Ya commented on these issues. Department's Position: We agree with the petitioners that we did improperly exclude imputed credit expenses in making the circumstances-of-sale adjustment and double-converted U.S. packing expenses before adding them to normal value. With respect to the use of grade in defining a product for purposes of the cost test, we agree that separate grades constitute separate products, notwithstanding that different grades of the same merchandise will have the same cost. Accordingly, we have corrected all three of these errors for this final determination. ISSUES SPECIFIC TO FETL General Issues Comment 3: Pre-verification Revisions and Minor Errors FETL's Comments: At the commencement of verification, FETL reported several "minor" revisions to its cost data as a result of errors it discovered in preparing for the verification.(23) FETL likewise submitted two pre- verification corrections at the start of the sales verification.(24) One of these pre-verification revisions, identical to a revision reported at the cost verification, was to correct for the misidentification of the fiber type for certain control numbers. The other minor revision was to correct for the fact that an incorrect amount for FETL's short term credit expense was inadvertently included in its most recent submission. FETL asks that, because these cost and sales revisions were submitted in a timely manner and were verified by the Department, that the Department incorporate these revisions into the final determination. FETL further requests that the Department incorporate corrections to minor errors found during the course of the sales and cost verifications in FETL's reporting of raw materials discounts and its indirect selling expense ratio. FETL argues that it is the Department's established practice to incorporate verification corrections into a final determination when such errors are inadvertent, minor, isolated to particular areas, and do not affect the integrity of the data. Petitioners' Comments: The petitioners counter that FETL's pre-verification revisions were, in fact, substantial and extensive, including "yet another round" of significant changes to a substantial portion of the sales and cost databases. Moreover, despite these revisions, the petitioners note, the Department still discovered several additional errors in the reported data during the verifications. The cumulative effect of these changes, the petitioners conclude, raises serious questions as to the accuracy and integrity of the entirety of FETL's databases. Accordingly, the petitioners contend, the Department should reject FETL's databases and resort to total adverse facts available. Department's Position: We have incorporated into our final determination revisions due to minor errors that were identified in FETL's pre-verification submissions as well as certain errors identified during the course of the verifications. As noted by the respondent, it is the Department's practice to correct for minor errors found during the course of verification where, among other considerations, the errors are found to be isolated in nature and are determined not to affect the integrity of the data overall.(25) We find that the pre-verification revisions submitted by FETL at the start of the cost and sales verifications were minor and were of the type typically identified by the respondent during preparation for verification. Regarding the cost of production and constructed value databases, the pre- verification corrections were minor in that they affected only specific accounts, did not change the reporting methodology, and corroborated, supported, and clarified information already on the record.(26) We noted that due to the nature of FETL's accounting system and in order to conform with the Department's reporting classifications, the respondent developed its cost of production and constructed value figures through detailed allocation methodologies in which one change would affect many figures and many or all control numbers. We have determined the corrections to be minor even though they have affected many control numbers for the reasons stated, and, therefore, have included them in FETL's cost of production and constructed value for purposes of the final determination. In the case of FETL's discounts on raw material purchases, we verified that the discounts related to raw materials were used in the operating division which produced the subject merchandise.(27) However, we were unable to determine the specific lines to which the discounts on raw materials related (i.e., subject merchandise or non-subject merchandise). Since we are unable to associate these discounts specifically with inputs used to produce the subject merchandise, for the final determination, we have included the discounts on materials as an adjustment to G&A expenses, which relate to operations of the company as a whole. Likewise, we have incorporated the correct credit expenses into our final determination, because these data were submitted in a timely manner, and the Department found no indication during the course of verification that this was anything more than an inadvertent error. With respect to the errors first identified during the course of verification, we have addressed each specific error raised by all parties in separate comments below. Comment 4: Product Coding Petitioners' Comments: The petitioners argue that FETL has not complied with the Department's established hierarchy for model matches, particularly with regard to the matching category "fiber type." According to the petitioners, the following are specific instances relating to this alleged mis-coding of FETL's control numbers. The petitioners start by stating that FETL initially reported only two fiber types (virgin and regenerated), and it was not until after the preliminary determination that FETL divulged that it, in fact, sold blended rather than regenerated product. The petitioners allege that not only was this a "radical coding revision," but that this revision was astounding given that FETL had earlier described its regeneration production process in detail. Moreover, the petitioners continue, even after FETL revised its submissions to reflect the fact that it produces blended PSF, FETL still failed to specify the fiber types and percentages for all products coded as blended. FETL instead coded products with the same fiber composition, cross section, product finish, and product denier as a single control number regardless of the blending percentage. The petitioners furthermore maintain that a given control number can encompass a variety of blends. According to information collected at verification, FETL's reported fiber types (virgin versus blended) are determined by the broad groupings of the actual PET chips used to make those fiber types. However, the petitioners contend, the parameters and characteristics of these broad chip groupings have not been clearly identified. According to the verification report, the petitioners suggest, these PET chip groupings can encompass "a wide range of materials, including 100 percent virgin chips or 100 percent recycled chips." Moreover, in addition to fact that the blend of materials within a broad chip group is not clearly defined, according to the petitioners different recipes of broad chip groupings can be used to produce the products within a given control number. Consequently, there is not one unique chip blend for each reported control number. Moreover, continue the petitioners, by coding all of its products as either virgin or blended, FETL has not reported any products as 100 percent regenerated or 100 percent recycled. The petitioners contend that this is clearly incorrect given that FETL has also reported that it has at least one production line that produces PSF "exclusively from waste inputs." Therefore, conclude the petitioners, FETL has coded both blended and 100 percent regenerate or recycled products as a single control number despite the explicit instructions to the contrary in the Department's original questionnaire. Finally, the petitioners contend, rather than submitting a well-supported, timely request that the Department revise the model matching hierarchy, FETL simply substituted its own interpretation for the various codes defined by the Department. The petitioners insist that this is impermissible under the law. Therefore, according to the petitioners, FETL's control numbers fail to differentiate adequately product characteristics under the model match approach devised by the Department. The extent of the potential distortion to the Department's dumping analysis as a result of these errors in coding, according to the petitioners, is evidenced in the range of disparities in particular prices and costs for products under the same control number. As a result of these coding errors, the petitioners contend, the Department is unable to perform a reasonable model match, apply the difference in merchandise adjustment, or conduct the below-cost test. Moreover, at this late stage in the investigation, it is not possible to ask FETL to revise its reporting of the U.S. sales, home market sales, and costs, separating these product codes into separate control numbers based on the PET chip recipe. Therefore, the petitioners argue, the only available alternative to the Department short of applying adverse facts available to all of FETL's sales (comment on adverse facts available above) is to apply a partial adverse facts margin of 70.70 to the U.S. sales affected by these coding problems. FETL's Comments: FETL counters that the petitioners' arguments are unsupported by the evidence on the record, the Department's instructions to FETL, and the case history. Moreover, the petitioners arguments are untimely and, as such, should be rejected. Specifically, FETL maintains that it correctly identified the fiber type of each observation in its sales and cost databases, and that these databases were fully verified as accurate in this regard by the Department. The respondent continues by arguing that its corrected fiber types, blending ratios, and raw materials consumption were reported well within the time limits (i.e., seven days prior to the commencement of verification) established under the Department's regulations. With regard to FETL's broad groupings of PET chips, the respondent argues that these broad groupings, along with their defining characteristics, were reported in the August 27 supplemental response. Moreover, FETL continues, the Department verified that the waste content of chips within each one of these groups falls within a very narrow range. The respondent maintains, therefore, that it has correctly grouped its chip types on the basis of virgin and blended products. FETL further argues that it has fully complied with the Department's instructions on how to report blended fiber by coding all blended products under the same digit in the fiber type category. No additional breakout was requested by the Department. FETL contends that the Department decided on this product characteristic coding based on extensive arguments from the petitioners, and that the petitioners never objected to this coding methodology until its case brief. Moreover, the respondent adds, the examples the petitioners cite of where they believe FETL should have coded the fiber type as recycled or regenerated are wrong; the raw materials used to manufacture those products include a combination of waste, recycled and regenerated inputs and therefore, consistent with the Department's coding definitions, those products have been coded as blended. With regard to the petitioners' argument that considerably different products have been combined under a single control number, FETL contends that the evidence (i.e., widely divergent prices and costs) the petitioners offer in support of this point is simply wrong, and is based on an erroneous matching of cost data and control numbers and an incorrect exclusion of other relevant cost data. FETL maintains what when the total costs of various products within the same CONNUM are compared, the differences are minor and largely due to other additives (such as color) that are not related to Fiber Type. FETL concludes that it has correctly weight-average under a single control number all products with identical characteristics (according to the Department's matching criteria) for which a further differentiation of characteristics was not requested by the Department. Department's Position: We disagree with the petitioners' contention that FETL's control numbers fail to adequately differentiate product characteristics under the model match approach devised by the Department. The petitioners' arguments can be summed into essentially two allegations. The first is that FETL failed to report, in a timely manner and at the level of detail requested by the Department, the Fiber Type of its U.S. and home market sales. The second argument is that, to the extent that FETL did attempt to follow the Department's Fiber Type coding instructions, it did so in an arbitrary, inconsistent and inaccurate manner. We address these in turn. To put this discussion in context, it is useful to describe the product matching criterion at issue. After allowing interested parties to comment on the characteristics that should be considered in matching home market and U.S. sales, the Department adopted the following criteria for classifying or coding "Fiber Type:" FIELD NUMBER 3.2: Fiber Type FIELD NAME: TYPEH DESCRIPTION: 1= Virgin 2=Recycled 3=Blended (please specify fiber types and percentages) 4=Regen (regenerated) Virgin refers to PSF made from terephthalic acid and ethylene glycol which are polymerized into a compound called polyethylene terephthalate ("PET"). Recycled refers to PSF which is made from PET derived from recycled materials such as soda bottles. Regen is made from different types of polyester waste materials such as fiber and filament waste and scrap polyester materials. Blend refers to fiber which is made up of a combination of fiber types. Please specify the different types and the percentages of the components, i.e. virgin 20%/recycled 80%. Two aspects of these criteria are noteworthy. First, all the categories are defined by reference to the materials used to make them.(28) Second, although respondents are asked to provide the percentages of different input materials in the blends, they are not asked to create a separate code for each blend. For example, if one product is produced with 25 percent virgin input and 75 percent regen, while another product is made with 35 percent virgin input and 65 percent recycled input, the Department's instructions do NOT require the respondents to assign different codes to those different blended products. At verification, FETL officials explained how the company classified its "fiber type" based on the types of chips it used to produce PSF:(29) We asked officials to explain how they identified and categorized the fiber type of each of their sales for determining which sales should be reported for each CONNUM in their submissions. In response, officials provided as verification exhibit 6(a) a table which shows the average waste contents and contents of other inputs of the 13 broad categories of PET chips used to make PSF. These 13 chip categories were created for use in the investigation, and are broad groupings of over 100 individual PET chip types that FETL produces and uses in its PSF production. FETL has grouped its chip types into these 13 broad categories according to certain characteristics. (Company officials noted that a description of the characteristics of the 13 categories was submitted in exhibit 6 of FETL's August 27 supplemental response.) The table in verification exhibit 6 shows the average percentage of the total waste content in a particular chip category. In order to classify a product/CONNUM as virgin or blended, FETL looked at the chip inputs to produce the particular kind of PSF. If the input chips types had a combined average of 0.1 percent waste or less, then they coded the resulting PSF as virgin. Chip types C1 through C6, and C10 through C11 are virgin chips (i.e. less than 0.1 percent waste). Chip types C8, C9, and C12 are blended/regenerated chips in that they contain more than 0.1 percent waste (chip type C7 is not used in producing merchandise within the scope).(30) The responses received from FETL are consistent with this explanation. Since its earliest sales listing, FETL has categorized all of its sales as being of two "Fiber Types." Originally, the two types were virgin and regenerated. (Subsequently, the second type was identified as blended.) The petitioners rightly point out that it was not until December 17, 1999, that FETL clarified that its sales that were previously reported as regenerated were actually blended. However, we disagree with the petitioners' contention that this represents a radical coding revision. As FETL suggests in its January 4, 2000, submission, the revision does not affect FETL's product control numbers, sales listings, or the cost databases. In other words, rather than having any tangible impact on any potential dumping margin, the revision amounts to little more than a change in the label for one subset of data within FETL's databases. The record evidence, viewed in light of the potentially unique facets of FETL's operations, does not suggest that incorrectly coding blended products as regenerated was a willful or coordinated error on the part of the respondent. For instance, one contributing factor to this error is that the Department has applied a more narrow definition to the terms recycled, regenerated, and blended than FETL observes in the course of its normal operations and record- keeping.(31) Moreover, rather than being confined to the production of a few, easily distinguishable types of products, non-virgin inputs are ubiquitous across many of FETL's manufacturing lines. Also, with the exception of FETL's "condux line," these non-virgin materials are generally reintroduced at the beginning stages of the production process and, therefore, become largely intertwined with the virgin raw materials. The following is an excerpt from the Sales Verification Report (at 11): Officials explained that in their submissions, they have used the term blended to refer to products that are produced using a mixture of virgin and recycled ("waste") inputs. In other words, blending here refers to a mixture of inputs and not a blend of outputs (i.e., different types of fiber). Officials also explained that FETL does not differentiate in the ordinary course of business between regenerated and recycled inputs. Rather, they refer to all non-virgin inputs as "waste." The waste that FETL uses is generated at various stages in FETL's production process. Referring to exhibit 6 of their August 27 submission, officials provided an overview of their production process, highlighting the stages at which waste is generated. For example, a small amount of very pure waste is recaptured at the purified terephthalic acid (TPA) stage. Officials noted that this is not really waste in the normal sense, but merely left over TPA that was not used up in the previous batch. There is also "solid waste" generated after the polymerization process. Officials explained that all solid waste is broken down into its constituent chemicals (i.e., back down into monomers) while mixing it with glycol. This liquid waste is then reintroduced into the estrification process. Finally, some waste is generated at the spinning and drawing stages. Much of this waste is used at the "recycle line" and introduced at the polyester (PET) chip stage. These are referred to as "condux chips." The remaining amount of the fiber waste is broken down, mixed with glycol and reintroduced as liquid waste at the estrification stage. Officials stated that all of FETL's PSF production lines except one generally use some blend of virgin and waste inputs. The one exception is FETL's recycling plant or "condux line" that produces PSF exclusively from waste inputs. Officials reiterated that FETL has no 100 percent virgin lines,(32) only one 100 percent recycled line. It is also clear from the record that FETL made a good faith effort to comply with the Department's instructions, and provided a reasonably complete response with regard to product matching. In exhibit 6 of its August 27, 1999 supplemental response, FETL provided a general description of the major categories of PET chips used in its production process. In its December 6, 1999 supplemental response (at Exhibit 13), FETL reported the percentages of each chip category type, as well as the percentage of other raw materials, used in making the products within a particular control number. These chip categories, including the physical characteristics of the chips that comprise each category, were then closely examined at verification. Therefore, FETL complied with the Department's instructions in specifying the types and percentages of the components of its blended merchandise. Also, as is clear from the dates of its submissions, FETL's compliance was timely.(33) Having concluded that FETL adequately reported the "fiber type" of its U.S. and home market sales in a timely manner, we address the petitioners claim that: " . . . it was important for the Department to know the proportions so that it could determine how to compare the blended products reported. Only by reporting blend percentages, as requested, could FETL provide the Department with information to enable it to compare products properly. And only by reviewing the blend ratios could the petitioners have the opportunity to comment on proper product comparisons based on the merchandise described." As we understand this argument, the petitioners are asking for the opportunity to argue that the product matching criteria should be changed, i.e., although the Department did not ask respondents to match their sales of blended product based on the percentages of input types, such matching would, in fact, be appropriate. Clearly, the Department has the authority to amend its matching criteria throughout the course of a proceeding (see Rautaruukki Oy v. United States, 22 CIT , Slip op. 98-112 (August 4, 1998). Furthermore, we acknowledge that our understanding of PSF has increased greatly since we initiated this case and developed the matching criteria. Therefore, we have reexamined the information on the record of this proceeding regarding whether different blends should be assigned different product codes so that similar blends would be matched in the U.S. and home markets. The most in-depth discussion of the use of blends and other fiber types as matching criteria was submitted by the petitioners in response to our request for comments at the outset of this investigation. The petitioners argued that the Department should not use "fiber type" (or as they would have labeled it, "Type of Raw Material Used to Produce PSF") as a matching criterion at all. The petitioners contended that the matching exercise should focus on the physical characteristics of the merchandise being investigated, and not on the characteristics of the input used to make that merchandise. With respect to the blended category, in particular, the petitioners argued that if the Department decided, notwithstanding their arguments to the contrary, to include Fiber Type as a matching criterion, the Department should add a blended category because the three categories proposed by the Department (virgin, regenerated and recycled) were incomplete since PSF could be produced from a variety of input types. Although the petitioners asked us to include a fourth product category of Fiber Type, there is no discussion that blends should further be matched based on the proportions of component materials. We acknowledge that the petitioners requested that blend percentages be specified (information which we did request), but beyond that, there is no indication that blend percentages would be important as matching criteria. Therefore, we have no basis to say that, even if the respondent's data would permit it, that we should match home market and U.S. sales on the basis of the blend percentages. We turn now to the petitioners' second broad claim, i.e., to the extent that FETL did attempt to follow the Department's Fiber Type coding instructions, it did so in an arbitrary, inconsistent and inaccurate manner. With regard to the accuracy of FETL's Fiber Type data, the petitioners object to FETL's methodology for classifying the Fiber Type of a particular PSF product based on the chip recipe used to manufacture that product. The petitioners specifically allege that FETL has not adequately identified the parameters of the 13 broad chip groupings of its more than 100 chip types. The petitioners further allege that the actual waste contents of the individual chips comprising a chip group may diverge widely from the average waste content of that group. The petitioners' allegations are directly refuted, however, by FETL's submissions and our findings at verification. As noted above, FETL first reported the characteristics of its chip groups in Exhibit 6 of its August 27 response. Subsequently, at verification the Department closely examined (and took exhibits documenting)(34) the physical characteristics and chemical contents of the specific PET chips and verified that the waste percentages of the chips within a particular chip group vary only within a narrow range. As stated in FETL's sales verification report (at 13): To better understand the average waste percentages of each chip category, we asked to what extent there were ranges in the percentages of PET chip types (C1, C2, etc.) used to produce a given CONNUM. We randomly selected CONNUM 2422150, and asked to see the specifications of each of the sub-types of chips used to produce that product. According to the detailed fiber recipe (in verification exhibit 6(b)), the following broad categories of chips were used to produce this CONNUM: C1, C9, and C13. Verification exhibit 6(c) shows the specific chip types included within each of these chip categories. We asked to see the specifications of each of the chip types. Officials provided the chip specifications for each individual chip type. We selected a sample of these specifications as verification exhibit 6(d). Reviewing the specifications for all of the chips within the C1 category confirmed that the range of waste in C1 was between 0 and .04 percent. We then reviewed the specifications for each of the chip types in the category C9. For all of the chips, the specifications noted a total waste content between 7.0 and 7.25 percent. To further confirm that FETL's methodology for distinguishing virgin from blended material was not arbitrary but, rather, based on some reasonable and objective criteria, at verification we asked company officials to explain why they had selected a 0.1 percent waste content as the threshold between virgin and blended. Officials explained as follows: Officials stated that 0.1 percent was a reasonable cut-off for classifying virgin from "regenerated" because a waste content this low really is not "waste" at all, but merely reflects the fact that sometimes in the esterification/polymerization process there is some trace of residue on the inside of the containers from previous batches that gets blended in with the current batch. Moreover, officials noted, there is a considerable difference between the average 0.1 percent waste threshold for virgin, and the next highest waste average percentage in the blended categories of PET chips (i.e., there is a natural dividing line.) Officials also presented verification exhibit 6(b), which details the fiber recipe, i.e., the percentage amount of each chip category used to produce a particular fiber product/CONNUM. Based on this recipe, FETL determines the precise chip types used to make each product/CONNUM. Referring to the exhibit, officials noted that only the following CONNUMS are made from virgin chips and, therefore, have been coded as virgin: 2112030, 2122060, 2122030, 2122040, 2122041, 2122060. Based on a brief scanning of the sales list, we confirmed that no other CONNUMS in the U.S. sales database were coded as virgin. (Sales Verification Report at 12.) Also, implicit in this and many of the petitioners' other arguments is the suggestion that the Department should have conducted a cost test and matched sales according to individual blend percent. In their December 22, 1999 comments on FETL's reporting of its fiber types, the petitioners argue that absence of reported blend percentages for individual sales "precludes the Department from knowing whether it is making a fair comparison in its selection of model matches for U.S. and home market products," because "a product with an 80 percent virgin/20 percent regenerated blend could be of significantly different cost than a product of 20 percent virgin/80 percent regen. blend." To a large extent, however, the petitioners' concerns have already been addressed by the respondent. At verification, FETL officials explained that: . . . in producing most products, they are restricted to using a very narrow range of blending proportions (i.e., ratios of virgin to waste inputs). This is because the blending proportions are dictated by the standards of physical and chemical properties established for each product. In other words, there is generally a set blending ratio for each product. See CONNUM Derivation section below. Therefore, according to officials, FETL's customers generally do not focus on what particular blend was used to produce a given product, because the blending proportions are implicit within the stated product specifications. It is also because these blending proportions are already implicit in the product characteristics themselves that FETL's product codes do not explicitly identify these proportions. (Sales Verification Report at 11.) In other words, matching sales (and conducting a cost test) based on the blending ratios for individual sales would be redundant. As FETL officials explained, the differences in blending ratios between different products largely mirror the differences in other physical characteristics.(35) To the extent those other physical characteristics are separately coded, the differences in blending ratios are implicitly taken into account. The petitioners counter that FETL must have combined "vastly different products" under one control number because, according to information examined at verification, there are different chip recipes and material costs reported for the same control number. The petitioners are incorrect, however, that different costs and recipes prove that there are vastly different products combined into one control number. Rather, FETL's submission and information examined at verification make it clear that FETL has several different production lines, each with different production costs.(37) Certain production lines can use certain types of raw materials (including types of waste) in producing a given product that cannot be used at other production lines.(38) The cost of a non-virgin input depends on the type of waste it is (e.g., bottle flake, liquid waste). Thus, the particular raw material recipe used to produce a product depends on the particular production line used to manufacture that product. The result is that a given product (and control number) can have varying production costs depending on the raw materials (e.g., chip recipe) and the production line used in its manufacture. It does not hold, however, that different recipes (used to produce the same or similar products) must have widely varying blending ratios. As is clear in exhibit 6(a) of the sales verification report, different chip groups can have roughly similar average waste contents.(39) Therefore, different recipes of combination of chips do not necessarily result in vastly different waste contents. In other words, the fact that different recipes can be used to produce a given control number does not undermine FETL's methodology for classifying fiber type of its sales according to the chip recipe. The petitioners also argue that FETL has incorrectly coded 100 percent regenerate or recycled products as blended products. To make their point, the petitioners cite the following excerpts from the sales verification report: Officials stated that all of FETL's PSF production lines except one generally use some blend of virgin and waste inputs. The one exception is FETL's recycling plant or "condux line" that produces PSF exclusively from waste inputs. Officials reiterated that FETL has no 100 percent virgin lines, only one 100 percent recycled line (at 11). According to officials, chip category C13 is 100 percent waste and, consequently, not a blend. (at 13). These statements made by FETL officials at verification must be understood in light of the point noted above, however. This point is that FETL does not differentiate in the ordinary course of business between regenerated and recycled inputs but, instead, generally refers to all non-virgin inputs as waste. Keeping in mind that blended fiber contains any combination of virgin, regenerated or recycled inputs, a product made of "100 percent" waste is correctly coded as blended when manufactured from a mixture of regenerated and recycled "waste." In conclusion, as a result of these coding errors, the petitioners contend, the Department is unable to perform a reasonable model match, apply the difference in merchandise adjustment, or conduct the below-cost test. For the above reasons, however, we find the petitioners' argument to be without merit. We find that FETL provided sufficiently complete, consistent and accurate data with regard to its product coding, and were responsive to the Department's instructions. Comment 5: Antibacterial and Flame-Retardant Products Petitioners' Comments: The petitioners contend that contrary to the Department's instructions, FETL did not modify its control numbers to distinguish antibacterial and flame retardant properties, nor did it provide the associated costs of production or the difference in merchandise adjustment data for these products. The petitioners argue that according to information obtained at verification, FETL distinguishes these properties in its own internal product coding system and, therefore, could have readily complied with the Department's request. As a result of this failure, the Department is unable to properly match FETL's sales of antibacterial and flame retardant PSF and, consequently, unable to calculate the actual dumping margin for these products. Moreover, according to the petitioners, FETL's behavior in this regard has prevented the petitioners from fully and meaningfully commenting on FETL's sales of these products. For these reasons the Department should rely on the facts available rate of 70.70 percent for determining the dumping margin on sales of these products. FETL's Comments: FETL urges the Department to reject the petitioners argument to rely on a facts available rate in this instance on the grounds that any sales of merchandise infused with antibacterial or flame retardant agents can be readily identified within the sales databases. Moreover, according to FETL, the quantities of these products that were sold were relatively small. However, if the Department does choose to adjust the antibacterial costs, FETL argues that it should use as fact available Nan Ya's reported antibacterial finishing costs. This would not create any undue difficulties for the Department, FETL continues, because any sales of PSF infused with antibacterial material are readily identifiable in the database and represent only a very small fraction of FETL's total U.S. sales. Department's Position: We begin by noting that the Department's, as well as the parties', understanding of these antibacterial and flame-retardant properties has evolved throughout the course of this investigation. In the initial stages of this proceeding, the Department provided parties the opportunity to comment on the importance of various product characteristics and appropriate matching criteria.(40) Though interested parties commented extensively on several categories of physical characteristics, unique coatings or raw materials that impart special properties to the PSF were generally not discussed in depth by either the petitioners or the respondents.(41) Reflecting the relative unimportance most parties appeared to attach to these special properties, the Department did not incorporate a separate category for these properties into the matching criteria in the initial questionnaire. Rather, based on our understanding at the time that these special characteristics were imparted in the form of a special coating or finish, we intended these characteristics to be subsumed within the "product finish" category. We note that this appeared consistent with the petitioners' proposed matching criteria and hierarchy of physical characteristics.(42) In the matching control number section of its initial (July 23, 1999) sections B and C questionnaire response, FETL provided a separate breakout of its flame retardant and antibacterial products under the category "Fiber Composition" (at B-5). The petitioners, in their July 30, 1999 submission, explicitly objected to FETL's breakout of these products, as follows: FETL has significantly modified the Department's reporting of fiber composition by reporting not only new subcodes for the requested fiber composition in the first digit but by adding a second digit that described either the color of the PSF (white color or with color) or additives to the PSF (flame retardant or anti-bacteria). The supplemental information reported by FETL is either repetitive of the codes already established by the Department, or is not a significant product characteristic that warrants consideration in the control number. Therefore, the Department should reject FETL's attempt to distort the model matching and instruct FETL to resubmit its home market and U.S. sales data in accordance with the Department's characteristics and coding. (at 2) Similarly, the Department should reject FETL's attempt to segregate additives (either flame retardant or antibacterial) as these characteristics are not a definition of fiber composition, nor are they significant enough to warrant consideration in the Department's model match criteria. (at 4) Mindful of the petitioners' objection, in its subsequent supplemental questionnaire the Department stated that it appeared that FETL's separate breakout of these properties was inappropriate, and that these processes require either a coating or additive that would better be captured under the Product Finish category. Accordingly, we instructed FETL to explain the process required to add flame retardant or anti-bacterial qualities to PSF and, also, to modify its databases to include these characteristics under the Product Finish category. In response, FETL stated that "respective CONNUM digits are modified according to the Department's instructions in the cost dataset." (August 27, 1999, at 7.) It was only at the sales verification of FETL that it was first explained to the Department that, in fact, antibacterial properties can be imparted either by coating the PSF or by infusing the antibacterial agent into the raw material PET chips. FETL officials explained at verification that: US 2 was a sale of a product made from a chip that is infused with an antibacterial material. This product is distinctly different from products with antibacterial finish. This type of antibacterial product is distinguished by an "AB" in the 7th and 8th digits of the product code. In its original response, FETL reported this as a separate fiber composition code. However, according to the Department's instructions, they reported these sales as virgin sales with no distinction for the added antibacterial agent.(43) Since the antibacterial agent is embedded in the chip and not added as a finish, the finish coding is listed as "2" for non-silicone coding. According to FETL officials, this is a very special product and commands a higher price. FETL officials explained that they also produce PSF with antibacterial coatings which would have a product code with "HL" in the 7th and 8th digits.(44) An examination of FETL's data indicates that it made sales of flame retardant products, products infused with antibacterial material and products with antibacterial coatings during the period of investigation. Since the application of facts available is not warranted with respect to these products, as stated above, the only question remaining is whether or not these products should be separated out through the use of additional matching criteria. The products with antibacterial coatings fit in well with our matching criteria. Accordingly, there is no need to make any changes with respect to these products. With respect to flame-retardant products and products infused with antibacterial material, we note that the total quantities of sales of these products are so small that they will have no significant affect on the overall estimated weighted-average dumping margin; therefore, we find it unnecessary to contemplate whether the matching criteria are fully appropriate for these products. Based on the foregoing, we have removed the sales of infused antibacterial and flame retardant products from the sales databases. B. Sales Issues Comment 6: Movement Expenses and Bank Charges on U.S. Sales Petitioners' Comments: The petitioners claim that FETL's simultaneous use of quantity-based allocations on some transactions and value-based allocations on others in calculating per-unit amounts for certain expenses (i.e., bank charges, ocean freight, domestic inland freight, brokerage and marine insurance) on U.S. sales resulted in significant under-reporting of these expenses. Since the Department is unable to correct this under-reporting, the petitioners argue that the highest reported per-unit amount for each of these expenses should be used on all U.S. sales as facts available. FETL's Comments: FETL argues that the petitioners' characterizations of the misreporting of these expenses are over-exaggerated. In fact, states FETL, errors with respect to these expenses impacted only one sale examined as part of a sales trace. Since the errors were so limited, FETL argues that the application of adverse facts available is not warranted. Department's Position: Since FETL does not track bank charges, ocean freight, domestic inland freight, brokerage or marine insurance on a transaction- specific basis in its accounting records, it had to calculate these expenses by pulling the source documents (e.g., supplier invoices) and calculating the per- unit amounts by hand. Unlike bank charges, ocean freight, brokerage and domestic inland freight which were sometimes allocated by volume and sometimes by value, marine insurance was consistently allocated correctly by value. Therefore, marine insurance expenses are not at issue here. In addition to the U.S. sales examined as complete sales traces, we also examined source documentation on these expenses for other transactions. We found that bank charges were the only expense at issue that was incurred on a value basis; therefore, bank charges should have been allocated based on value. The remaining expenses should have been allocated based on volume. We found that in some instances, the appropriate methodology was employed, but in other instances it was not. The use of the inappropriate allocation methodology sometimes resulted in a lower per-unit amount for an expense than was actually incurred, as the petitioners have stated, but also sometimes resulted in a higher per-unit amount on other transactions. The errors FETL made in reporting these expenses were generally of a small magnitude (either in the percentage deviation from the actual amount or in the absolute size of the difference in relation to the transaction price). While we find it appropriate to make an adjustment to FETL's reported amounts for these expenses based on our findings at verification, we do not believe that the use of an adverse inference in making this adjustment, such as proposed by the petitioners, is warranted. Instead, we are adjusting FETL's reported amounts for each of these expenses by the weighted-average percentage deviation between the reported amounts and the amounts actually incurred on transactions examined during verification. Comment 7: Commissions Petitioners' Comments: At verification, the Department noted that FETL did not report any commissions for its sales of PSF despite the fact that commissions were deducted from total selling expenses as a direct selling expense in the calculation of FETL's indirect selling expenses. The petitioners note that, according to the verification report, FETL officials could not offer any documentation supporting their explanation that commissions are only paid on export sales to certain export destinations other than the United States. The petitioners contend that FETL officials should have been able to document this explanation using various kinds of payment records, such as commission payment accounting vouchers. In lieu of any verified proof, the petitioners argue that the Department should conclude that there were commissions paid on U.S. sales and, accordingly, apply as facts available the result of dividing the total recorded commissions for FETL's Chemical Fiber Division by the total quantity of U.S. sales. FETL's Comments: In response, FETL urges the Department to reject the petitioners' arguments on the ground that they are not substantiated. According to the respondent, the Department found no discrepancies in the reporting of commission expenses in the U.S. database. However, FETL adds, if the Department does decide to adjust the respondent's sales to account for commission expenses, it should use a unit figure calculated by multiplying the total reported commissions in FETL's chemical fiber division by the ratio of FETL's verified total export sales of staple fiber to the total sales of the chemical fiber division, and then divide this amount by the total export sales of staple fiber. Department's Position: The Department found no evidence at verification, in the course of examining the sales documentation and accounting records for several of FETL's reported U.S. sales, that FETL had paid any commissions on those U.S. sales. Therefore, we do not find that there were unreported U.S. sales commissions and, accordingly, we have not applied facts available or made any adjustment to account for any allegedly unreported U.S. commissions. Comment 8: Sales to Affiliate Petitioners' Comments: The petitioners cite the failure by FETL to disclose the fact that it had sales to an affiliated party in the home market. FETL's Comments: FETL made no comment. Department's Position: In the course of reviewing the FETL's corporate structure and affiliations at verification, company officials identified one affiliated domestic trading company to whom FETL had sold a small amount of PSF during the POI.(45) In its submissions, FETL had failed to identify these as sales to affiliates. In examining the scope of this error, we verified that the amount of these sales was relatively small, and that most of the sales were of PSF outside the scope. We also further determined that there were only a few sales of merchandise within the scope to the affiliate and that these sales had been properly included within FETL's database (although they had not been flagged as affiliated sales). We found no information that suggested that FETL had any other PSF sales to affiliated parties that had been improperly recorded. We further saw no indication that FETL had deliberately withheld information on the nature of these sales. However, for the final determination we have removed these affiliated sales from the home market comparison pool.(46) Comment 9: Verification of Surprise Sales Petitioners' Comments: The petitioners state that " . . . the Department did not select any surprise sales to verify at its sales verification of FETL. Had any surprise sales been selected for verification, petitioners have little doubt that the numerous errors uncovered in the verification of Nan Ya's surprise sales would have been equally true for those of FETL (at 15). FETL's Comments: FETL did not comment on this issue. Department's Position: The petitioners' assertions regarding surprise sales are wrong. In its January 5, 2000 verification outline to FETL (at 10), the Department identified certain pre-selected sales that the respondent should prepare for the sales verification.(47) At verification, in addition to the pre- selected sales, we examined documentation for at least another five sales. Our findings regarding these surprise sales are contained in the home market sales trace and U.S. market sales trace sections of FETL's Sales Verification Report at 13 and 17, respectively. Cost of Production and Constructed Value Issues Comment 10: Major Inputs--PTA Petitioners' Comments: The petitioners note that FETL purchases a significant amount of its purified terephthalic acid (PTA) from an affiliate, Du Pont Far Eastern (Du Pont). The petitioners argue that Du Pont's cost of production of PTA was higher than both the market value and the transfer price to FETL. In addition, the petitioners claim that certain expenses were excluded from Du Pont's calculation which meant that the cost of production for this input was under-stated. The petitioners contend that since PTA is a major input, the Department should apply the major input rule and use the higher of market value, transfer price or the revised cost of production. FETL's Comments: FETL did not comment on this issue. Department's Position: We agree with the petitioners that, for the purposes of this final determination, PTA is a major input in the production of PSF. Therefore, in accordance with our normal practice we have valued the PTA purchased from Du Pont using the highest of transfer price, market price or the cost of production.(48) For purposes of calculating Du Pont's cost of production of PTA, we have revised the G&A expenses to include certain excluded costs which relate to the general operations of the company as a whole. Under the guidelines of section 773(f)(3) of the Act, for the final determination, we have used the cost of production as it was higher than both transfer price and market price. Comment 11: Major Inputs-EG Petitioners' Comments: The petitioners note that FETL purchases a significant amount of its ethylene glycol (EG) through a line of affiliated parties including Alberta & Orient Glycol Company (A&O), who was reported as the affiliated input producer. The petitioners argue that the cost of production of the EG which FETL purchased from A&O should be valued as the sum of A&O's cost of production plus the G&A and financial expenses of each party involved in the sale transaction. The petitioners recommend that the Department use the highest G&A and financial expense rates on the record where no rates are available for a particular party. As a result, the petitioners argue that the cost of production for EG was higher than the market value and the transfer price used to establish the cost of production and constructed value. The petitioners contend that since EG is a major input, the Department should apply the major input rule and use the highest of market value, transfer price or the cost of production. FETL's Comments: FETL argues that the Department should not calculate the cost of production of EG purchased from A&O based on the cost of all of the affiliated parties involved in the transaction. The respondent contends that such treatment would be a clear misinterpretation of the law, which indicates that the major input rule applies only to the cost of production of the immediate, affiliated supplier. FETL maintains that the law should be read with NSF Ltd. et. al. V. United States, 115 F.3d 965 (CAFC 1997) in mind where the United States Court of Appeals for the Federal Circuit explained that the "commonly accepted meaning of the term" must be followed. The respondent also argues that in Certain Cut-to-Length Carbon-Quality Steel Plate Products from Korea, 64 FR 73196 (December 29, 1999) the Department stated that transactions between the respondent and the affiliated party govern the standard. FETL rejects the petitioners' argument that the Department should use the highest reported G&A and interest rates to calculate the EG COP, claiming that it is unsubstantiated. The respondent suggests that if the Department decides to calculate the EG cost of production based on the costs incurred by all of the parties involved in the transaction, it should include the verified G&A and interest rates of the trading company who sells the EG directly to FETL. Department's Position: FETL does not purchase EG, the input at issue, directly from its affiliated producer. Rather, it is obtained through a line of affiliates, including the affiliated producer. Since the affiliated producer and all of the parties in the transaction between the producer and the respondent are affiliated with FETL, and the total value of the purchases of the EG from the affiliates is significant in relation to the total cost of manufacture of the subject merchandise, we have determined that section 773(f)(3) of the Act applies in this case. We disagree with FETL that the intent of section 773(f)(3) of the Act and the related regulations is only to be concerned with the cost of production of the immediate affiliated supplier. To do so in this case would mean to blindly base the cost of production computation of the affiliated supplier immediately preceding the respondent on the transfer price between the affiliated supplier and its affiliate, which in effect, contradicts the purpose of section 773(f)(3) of the Act. As discussed in Notice of Final Determination of Sales at Less Than Fair Value: Stainless Steel Round Wire from Canada, the intent of sections 773(f)(2) and (3) of the Act "and the related regulations is to account for the possibility of shifting costs to an affiliated party. This possibility arises when an input passes to the responding company through the hands of an affiliated supplier, regardless of the value added to the product by the affiliated supplier."(49) As such, we consider the appropriate method for computing the cost of production for a major input obtained from an affiliated producer through a chain of affiliated party transactions to be based on the actual cost of production incurred by the affiliated producer plus a portion of the general expenses of all of the affiliates involved in the transaction between the affiliated producer and the respondent. Thus, for the final determination, we computed the cost of production for the affiliated EG purchases based on the affiliated producer's cost of production plus a portion of the general expenses for each affiliate involved in the transaction. For those affiliates for which we do not have general expense information on the record, we determined the general expense amount to apply using data from the other companies involved in the transaction where such information is on the record. Section 776(a) of the Act allows the Department to use facts otherwise available when necessary information is not on the record. This partial facts available adjustment is consistent with our treatment in Notice of Final Determination of Sales at Less Than Fair Value: Certain Cold-Rolled Flat-Rolled Carbon-Quality Steel Products From Brazil, 65 FR 5554, 5581 (February 4, 2000). Since the resulting cost of production exceeded both the reported transfer price for EG FETL purchased from its affiliates and the reported market price, for the final determination, we have used the cost of production to value EG. Lastly, we consider FETL's reference to our decision in Certain Cut-to-Length Carbon-Quality Steel Plate Products from Korea, 64 FR 73196 (December 29, 1999) (Comment 12), to be misplaced. That case dealt with the "Transactions Disregarded" rule under section 773 (f)(2) of the Act, not the major input rule under section 773(f)(3) of the Act. Furthermore, the transactions at issue in Certain Cut-to-Length Carbon-Quality Steel Plate Products from Korea involved purchases of inputs through an affiliated trading company who purchased those inputs from an unaffiliated producer, whereas in the instant proceeding, FETL purchased the EG through several affiliated parties, the last of which was the actual producer. Comment 12: Material Costs - Scrap Consumption Petitioners' Comments: The petitioners note that in FETL's normal books and records it values the scrap consumed in the production process at an amount which is higher than that used in the cost of production and constructed value figures reported to the Department. Citing the Act, the petitioners claim that costs should be calculated based on the records of the exporter or producer if they are kept in accordance with the generally accepted accounting principles ("GAAP") of the exporting country and reasonably reflect the costs associated with the production and sale of the merchandise.(50) The petitioners note that there is no evidence on the record which indicates that FETL's valuation of scrap in its normal books and records is not in accordance with the Republic of China's GAAP or that it is not reasonable. As a result, the petitioners argue that the Department should adjust FETL's cost of production and constructed value data for the difference between the scrap as valued in the company's records and as valued for reporting purposes. FETL's Comments: FETL agrees with the petitioners on the fact that in its normal books and records they value their scrap in accordance with the Republic of China's GAAP. However, FETL contends that the Department has an established practice of valuing scrap at the market price and not the internally recorded value. As a result, the respondent revalued their scrap for reporting purposes. Therefore, FETL argues that their scrap costs should not be revised in the final determination. Department's Position: Section 773(f)(1)(A) of the Act provides that for purposes of calculating cost of production and constructed value, "[c]osts shall normally be calculated based on the records of the exporter or producer of the merchandise, if such records are kept in accordance with the GAAP of the exporting country (or the producing country, where appropriate) and reasonably reflect the costs associated with the production and sales of the merchandise." In the instant case, we agree with the petitioners that there is no evidence on the record which indicates that FETL's valuation of scrap in its normal books and records is not in accordance with the Taiwan's GAAP or that it is not reasonable. Therefore, for this final determination, we have adjusted the cost of production and constructed value data for the difference in costs resulting from the two scrap valuation figures. Comment 13: Foreign Exchange Gains and Losses Petitioners' Comments: The petitioners argue that FETL's exchange gains and losses related to accounts payable and debt for the POI, as identified in the Department's verification report, should be included in the company's G&A and interest expense rate calculations, respectively. FETL's Comments: FETL did not comment on this issue. Department's Position: We agree with the petitioners that exchange gains and losses related to accounts payable and debt should be included in FETL's G&A and interest rate calculations. As the Department has repeatedly stated, our normal practice is to include a portion of the respondent's foreign-exchange gains and losses in the calculation of cost of production and constructed value.(51) Accordingly, for the final determination, we included the exchange gains and losses generated from financial transactions in the calculation of the financial expense rate and included the exchange gains and losses generated from accounts payable in the calculation of the G&A expense rate. Comment 14: G&A Expenses Petitioners' Comments: The petitioners argue that the Department should exclude packing costs, which were excluded from the submitted cost of manufacturing, from the cost of goods sold denominator in FETL's G&A expense rate calculation. FETL's Comments: FETL did not comment on this issue. Department's Position: FETL calculated its G&A expense rate based on a cost of goods sold denominator that included packing expenses whereas the COM to which the rate is applied excludes packing expenses. Consistent with our general practice, in order to ensure that the denominator of the G&A rate and the COM to which it is applied have the same cost components, we have adjusted FETL's G&A rate by excluding packing expenses from the cost of goods sold denominator.(52) ISSUES SPECIFIC TO NAN YA General Issues Comment 15: Mis-coding of Regenerated and Virgin Products Petitioners' Comments: The petitioners take issue with Nan Ya's description of its product coding system with respect to second-quality regenerated products as reported in its pre-verification corrections submission and elaborated upon during verification. Given the late timing of this information, the petitioners argue the Department should not simply accept it for purposes of the final determination. Instead, the petitioners propose a product coding pattern for home market sales that would properly classify some second-quality merchandise as regenerated, but would also improperly classify as regenerated second- quality, C-grade virgin merchandise. Nan Ya's Comments: Nan Ya states that it timely submitted corrections to its description of its product codes at the beginning of verification and that the Department, for the most part, was able to confirm its product coding during the course of verification. While acknowledging that there were certain product coding errors uncovered during verification, Nan Ya argues that it is the Department's practice to correct minor errors found during verification and cites a number of cases including, Stainless Steel Round Wire from Taiwan (Round Wire) 64 FR 1733, 17340 (April 9, 1999), and Sulfanilic Acid From the People's Republic of China, 63 FR 63834, 63837 (November 17, 1998). Department's Position: Nan Ya's product coding was, for the most part, supported at verification. The few errors that were uncovered with respect to product coding pertained to the identification of regenerated, second-quality merchandise. These errors appeared to be isolated, identifiable using verified information, and easily correctable. Under these circumstances, we agree with Nan Ya that it is the Department's practice to correct such minor errors.(53) We disagree with the petitioners' proposal for product coding because it would result in the erroneous classification of certain sales as regenerated when, in fact, they were sales of virgin material, albeit of second-quality. Comment 16: Recoding of Sale Petitioners' Comments: The petitioners argue that the Department should recode the control number of one of Nan Ya's U.S. sales based on information the Department examined at verification while tracing through various documentation for that sale. Nan Ya's Comments: Nan Ya replies that, contrary to the petitioners' arguments, the documentation examined for this sale fully establishes that the respondent had coded this product correctly. Therefore, no recoding is necessary. Department's Position: We agree with Nan Ya that the control number for this sale is correct. Sales Issues Comment 17: Exchange Rates Petitioners' Comments: In addition to the issue of sales completeness (see Adverse Facts Available comment above), the petitioners contend that Nan Ya's failure to report proper sales dates will distort exchange rate conversions because the conversion rates for comparison home market sales are based on the reported U.S. dates of sale. According to the petitioners, given that the dates of sale for a substantial portion of Nan Ya's U.S. sales are likely incorrect, the incorrect exchange rates were applied to those sales in the preliminary determination. For all of the U.S. sales examined at verification and found to have incorrectly reported dates of sale, the correct dates of sale were later in time. Therefore, the petitioners contend, the correct exchange rates (based on the later dates) would have been less favorable to Nan Ya. If the Department does not apply total facts available to Nan Ya, the petitioners argue that the Department should at least apply partial facts available to the misreported sales. Specifically, the Department should apply the highest exchange rate during the POI after October 3, 1998 to all sales with reported dates of sale on that date. Nan Ya's Comments: Nan Ya suggests that the petitioners' argument to apply as adverse facts available to Nan Ya the highest exchange rate after October 3 is contrary to law. Citing Mannesmannrohen-Werke AG, et al v. United States,(54) Nan Ya argues that adverse facts available would only be appropriate in cases of failure to respond. The Department's practice is not to apply the worst facts available, Nan Ya states, to a respondent that is otherwise cooperative. Moreover, Nan Ya continues, the petitioners' argument that the conversion rates for home market sales are based on the reported U.S. dates of sale is incorrect and based on a misunderstanding of the Department's calculation methodology. According to Nan Ya, the Department's practice is to convert all foreign currency unit prices and expenses to U.S. dollars. However, Nan Ya states, it has reported its gross unit prices, commissions, indirect selling expenses, and credit expenses in U.S. dollars. Therefore, any errors in conversion based on the wrong dates of sales would only affect movement expenses (inland freight, international freight, and brokerage charges) and bank charges. Nan Ya states that the Department verified that Nan Ya recognizes these expenses in its accounting system when the merchandise is shipped or the brokerage invoices are received. Therefore, Nan Ya suggests, the Department should use the date of shipment to determine the appropriate exchange rate for these expenses given that the shipment date is the most proximate to the actual date on which the expenses were incurred. Department's Position: A detailed analysis of the problems found at verification in Nan Ya's reporting of its dates of sale for several of its U.S. sales observations is contained in the separate Dates of Sale memorandum. For the reasons noted in that memorandum, we do not find that application of adverse inferences is warranted when choosing among facts available for those dates of sale which were found to be incorrectly reported. We do agree with the petitioners, however, that using conversion rates based on inaccurate dates of sale may result in calculation errors. Because this information is not on the record, pursuant to section 776(a) of the Act, the Department must use facts available. As recommended in the Date of Sale memorandum, therefore, as facts available we have applied a constructed date of sale to those observations for which the date of sale was reported incorrectly. This constructed date of sale has been derived by deducting from the shipment date the average number of days (calculated based on all of the sales observations for which the date of sale was correctly reported) between confirmation date and shipment date. The exchange rates corresponding to these constructed dates of sale were then used in making any conversions from New Taiwan Dollars for those sales observations with incorrect dates of sale. Comment 18: Domestic Inland Freight-General Issues Petitioners' Comments: According to the petitioners, Nan Ya disclosed at verification, for the first time in this proceeding, three crucial pieces of information. First, Nan Ya explained that it has two affiliated companies that provide services in transporting subject merchandise during the POI. Second, Nan Ya disclosed that it uses a freight rate schedule separate from the one previously submitted for transporting export sales. Third, Nan Ya explained that its trucking affiliates only arrange and facilitate freight services for export sales whereas these affiliates actually provide the freight services directly for some home market sales. The petitioners contend that Nan Ya's failure to report this information previously reflects "fundamental problems deeply rooted in Nan Ya's reporting methodology" and a "selective disclosure of vital information." Accordingly, the petitioners urge the Department to reject the inland freight expenses as reported by Nan Ya and, instead, apply the lowest reported unit inland freight expense to all of the home market sales and the highest unit inland freight expense to all of the U.S. sales. If the Department chooses not to do this, it should at the very least adjust Nan Ya's reported inland freight to account for the new information obtained during verification. In particular, the petitioners argue that the Department should correct the inland freight costs reported for one of Nan Ya's U.S. sales which the Department found to be reported incorrectly based on information examined at verification while tracing through various documentation for this sale. Nan Ya's Comments: Nan Ya counters that the petitioners' arguments regarding inland freight are contrary to the record evidence and inconsistent with the law, regulations and the Department's practice, for the following reasons. First, Nan Ya states, the record evidence shows that Nan Ya used only the Formosa Plastics Trucking Co. for transporting PSF and, conversely, that there is no evidence suggesting that Formosa Fairway Corporation was used for transporting PSF. Second, Nan Ya maintains, the petitioners have mischaracterized the scope of business between Nan Ya and these two transportation affiliates. There is no evidence that Formosa Plastics Trucking Co. is involved in the manufacture of PSF, or that either affiliate is used for transportation services in connection with the production of PSF. Third, the freight schedule that Nan Ya provided for container transportation at verification merely supports its position that Nan Ya's transactions with these affiliates are at arm's length and, therefore, does not constitute new information. Finally, Nan Ya states, with the exception of a few minor errors, the Department verified that Nan Ya correctly reported its inland freight expenses in its home market and U.S. sales listing based on actual freight expenses charged by its freight company. Therefore, the petitioners' suggestion that the Department apply the highest inland freight rate reported to all U.S. sales is unwarranted. Rather, Nan Ya concludes, the Department should apply a "reasonable percentage" in any instances where the reported inland freight was found to be incorrect at verification. Department's Position: At verification, we closely examined Nan Ya's reported inland freight expenses as well as Nan Ya's freight providers. We found no evidence indicating that Nan Ya had not fully disclosed its affiliation with various transportation companies, or that an affiliate other than Formosa Plastics Trucking was involved in the transportation of Nan Ya's home market or U.S. sales of PSF. We therefore find that rejecting the inland freight expenses reported by Nan Ya on the basis that Nan Ya had not fully disclosed all affiliates involved in the transportation of subject merchandise is not warranted. We address the petitioners' other arguments regarding Nan Ya's inland transportation costs below. Comment 19: Domestic Inland Freight-Adjustment for Affiliated Expenses Petitioners' Comments: According to the petitioners, Nan Ya divulged at verification, for the first time, that its affiliated transportation company arranged for the domestic transport of export sales by unaffiliated transportation companies. In addition, state the petitioners, verification showed that the affiliated transport company was not compensated outright for its services as a facilitator of domestic transport and there is no evidence on the record indicating that Nan Ya has accounted for the affiliated transport company's costs in its questionnaire responses. The petitioners also point out that the original questionnaire requested information on the involvement of affiliated companies in the production or sale of subject merchandise and requested financial statements for those affiliates Since Nan Ya failed to report the nature of the affiliated transport company's involvement in U.S. sales and to provide information on its expenses, such as its financial statements, the petitioners argue that the use of partial facts available is justified. As facts available, the petitioners propose that the highest per-unit domestic freight amount on U.S. sales be assigned to all transactions. Should the Department not adopt this proposal, the petitioners alternatively suggest that amount be added to the reported domestic freight expenses to account for the affiliated transport company's costs. As a surrogate for such costs, the petitioners propose that Nan Ya's own selling, general and administrative expenses, as found in its audited financial statements, be used, consistent with Certain Cold-Rolled Carbon Steel Flat Products from Korea; Final Results of Antidumping Duty Administrative Review (Carbon Steel from Korea) 63 FR 781, 788 (January 7, 1998). Nan Ya's Comments: Nan Ya states that it has never portrayed its affiliated transport company as being involved in the manufacture of PSF. However, Nan Ya confirms that, it did report in its July 23, 1999 submission that the company was involved in domestic freight services. Nan Ya adds that the results of verification showed that it had accurately reported its freight expenses with the exception of minor errors. Should the Department find that the freight expenses should be adjusted upward to reflect the expenses of its affiliated transport company, Nan Ya requests that Department apply a more reasonable adjustment than that proposed by the petitioners. According to Nan Ya, the petitioners' proposal would be contrary to the law, regulations and the Department's practice. Department's Position: At verification, we found that Nan Ya's affiliated transportation company provided services with respect to the transportation of U.S. sales to the port, but that it was not compensated for these services. Since the affiliated transportation company provided its services without compensation, the transactions between it and Nan Ya cannot be considered to be at arm's-length. Accordingly, we agree with the petitioners that it is appropriate to add to the reported factory-to-port transportation costs an amount representative of the facilitation services provided by the affiliated company. Because this information is not on the record, pursuant to section 776(a) of the Act, the Department must use facts available. As facts available, we are adding to the reported domestic inland freight expense an amount based on Nan Ya's general and administrative expenses as the only company-specific general and administrative rate available on the record to reflect the value of the facilitation services provided by the affiliated company. Comment 20: Domestic Inland Freight-Additional Freight to Factory Petitioners' Comments: The petitioners note that, at verification, the Department identified at least one sale where Nan Ya had shipped merchandise from the factory where it was produced to another factory a few kilometers away, combined the shipment with another export sale to the same customer, and then trucked the combined load to the port. The petitioners argue that Nan Ya did not substantiate its assertion that no shipping cost was reported for this sale as it was accounted for elsewhere in its response because company officials could not identify the other sale with which this merchandise was combined. At a minimum, urge the petitioners, the Department should add an amount of inland freight expense, based on the freight rates in Nan Ya's freight tables, to account for transporting this merchandise from one factory to the other. The petitioners suggest applying a per-unit inland freight charge that they have calculated based on the distance between two of Nan Ya's factories that produce subject merchandise, and based on Nan Ya's reported containerized freight rate schedule. Moreover, the petitioners continue, the Department should treat this additional transportation cost as a minimum inland freight expense, and add it to any U.S. sale for which Nan Ya did not report any inland freight expense. Nan Ya's Comments: In response, Nan Ya states that it has reported inland freight for all other U.S. sales and, therefore, the petitioners' arguments should be rejected. Department's Position: Nan Ya officials provided at verification the following explanation of this sale: Officials explained that this particular sale was an exception. Because the quantity was so small, Nan Ya shipped it from one of its factories (where it was produced) to another factory (a few kilometers away), combined the shipment with another export sale to the same customer (to fill up the container), and then trucked the combined load to the port. Officials were not, however, able to identify the other sale with which this sale was combined. Company officials did note that because they pay by the container for trucking of export sales, combining this small quantity with the other larger load to make a full container did not result in any additional freight costs.(55) Nan Ya has reported that it incurred no inland freight costs on this one sale. However, as indicated in the excerpt above, Nan Ya officials were not able at verification to support with documentation their contention that there were no additional inland freight costs associated with this shipment. We further note that there is one other sale in Nan Ya's U.S. sales database for which no inland freight cost has been reported. As in the case of the sale examined at verification, the quantity of this second sale is small. There is no evidence on the record substantiating Nan Ya's claim that it incurred no expense in trucking either of these shipments to the port of export. To the contrary, assuming the officials' explanation to be true, some portion of the expense incurred in trucking the combined shipment should be allocated to these observations with small quantities. Because this information is not on the record, pursuant to section 776(a) of the Act, the Department must use facts available. However, we find that the petitioners' suggested facts available trucking expense to be inappropriate because it is based on a specific trucking distance and it is not clear from the record evidence exactly how far these two sales were trucked to their respective ports of export. Therefore, we have applied as facts available to these two sales an average inland freight expense. This average has been calculated as the quantity- weighted average of all unit inland freight expenses reported for Nan Ya's U.S. sales. Comment 21: Inland Freight-Affiliated Transactions at Arm's Length Petitioners' Comments: The petitioners note that at verification, Nan Ya officials stated that the inland freight rates it pays to its affiliated trucking company are higher than those it pays to unaffiliated trucking companies. These higher than market freight costs, the petitioners argue, result in a larger deduction to the home market prices than would the arm's- length freight prices from unaffiliated trucking companies. The petitioners contend that, following the Department's normal practice,(56) Nan Ya should not be allowed to use transfer prices with an affiliated party if it has not established that such transfer prices are at arm's-length. According to the petitioners, because it is unclear which of Nan Ya's reported sales were transported by the affiliate, as partial facts available the Department should use the lowest non-zero unit inland freight for all of Nan Ya's home market sales. However, so that Nan Ya does not benefit from any non-arm's-length transaction, the Department should use the reported inland freight amount for the purposes of the cost test. Nan Ya's Comments: Nan Ya contends that the freight rate schedules clearly establish that the prices paid to Nan Ya's transportation affiliate are at arm's-length. Department's Position: In determining whether prices are at arm's length, we look to see whether affiliated parties treat each other the same as unaffiliated parties in setting their prices. Evidence indicating that prices are at arm's length can include: a comparison of the affiliated provider's price to the respondent with prices to unaffiliated companies; prices paid by the respondent to the affiliated provider with prices paid for similar services to unaffiliated providers; or, if appropriate, an analysis showing that the affiliated provider is meeting its costs. In Steel from Brazil and Flat Products from Korea, cited by the petitioners, the Department did reduce adjustments to home market prices for transportation and warehousing expenses provided by affiliated parties as adverse facts available when the respondents at issue were not able to demonstrate that the transfer prices were at arm's length or, alternatively, above cost. However, this is not the situation we are faced with here. Nan Ya was able to show at verification that the prices it paid its affiliated provider were at arm's length when compared to the prices it paid for similar services from an unaffiliated provider. Accordingly, we determine that the prices charged by Nan Ya's affiliated transportation provider are suitable for use in our analysis and have used the reported amounts in this final determination. Comment 22: Indirect Selling Expenses Petitioners' Comments: The petitioners contend that, according to information obtained at verification, it appears that Nan Ya's accounting codes used to distinguish expenses associated with Nan Ya's domestic sales office from Nan Ya's export sales office cannot be used as a basis to separate export indirect selling expenses from home market indirect selling expenses. The reason is that it appears that certain expenses (e.g., those stemming from computer usage) that are recorded under the accounting code for the domestic sales offices are more likely common expenses (i.e., indirect) attributable to both domestic and export sales. Moreover, the petitioners argue, the reported indirect selling expense ratio verified by the Department was considerably lower than the indirect selling expense ratio reflected in the preliminary determination. The petitioners argue that the selling expense information the Department examined at verification contained significant revisions to what Nan Ya had last reported in its supplemental responses and, thus, once again Nan Ya had substantially revised its data without providing any discussion or explanation of the changes to the Department, or any substantiation of need for revisions. Therefore, the petitioners urge the Department to use as adverse partial facts available these revised indirect selling expense amounts for the purposes of calculating the offset for U.S. commission, but to use the unrevised data in Nan Ya's most recent home market sales database when comparing home market net prices to Nan Ya's reported cost of production. Nan Ya's Comments: Nan Ya counters that the Department successfully verified its indirect selling expenses at the sales verification and concluded that there was no indication that any selling expenses had been improperly treated in the indirect selling expense calculation. Therefore, according to the respondent, the indirect selling expenses, that were reported prior to verification, have been fully substantiated. Department's Position: At verification, we fully verified Nan Ya's indirect selling expenses and found no indication that any selling expenses had been improperly treated in the indirect selling expense calculation generally, or that any selling expenses had been incorrectly accounted for when attributed to domestic sales, export sales, or both. As is clear in Nan Ya's Sales Verification Report, Nan Ya tracks in its accounting records numerous sub- categories of selling expenses related to its sales and transportation departments. Mindful of the time constraints that would not allow for a complete examination of all of these sub-accounts, at verification we selected and examined the nature and categorization of a representative sample of specific reported selling expense accounts, including those attributable to domestic sales as well as those attributable to export sales. After so doing, we concluded that, "with regard to indirect selling expenses, in tracing through the calculations derived from Nan Ya's accounting system, we noted no discrepancies between what was reported in the sales database and officials' statements, the supporting worksheets, and the accounting records."(57) The petitioners have seized on a handful--out of the many reported--of expense accounts, ones that the Department's verifiers did not happen to specifically examine, for the domestic sales department and allege that these "are likely common expenses for both domestic and export sales."(58) The petitioners support this allegation, in part, with their own translation from the original Chinese of information in a verification exhibit. The petitioners have not substantiated their assertion, however. The verifiers identified no instances where the expenses had been incorrectly categorized as domestic versus export. Moreover, that Nan Ya reported its indirect selling expenses correctly is further supported by the fact that Na Ya itself identified other areas where its accounting system did not normally track selling expenses separately. For example, to calculate accurate indirect selling expenses ratios, Nan Ya reclassified the common operating expenses of its "inventory moving departments" into domestic versus export using an allocation methodology based on respective sales.(59) With regard to the petitioners' argument that the indirect selling expense information examined by the Department was substantially revised over that last reported in Nan Ya's supplemental responses, we agree that the indirect selling expense ratios presented by Nan Ya at verification were different from those contained in its sales databases.(60) Nan Ya had explained to the Department that in preparing for the verification of these selling expenses, it had identified errors in how it initially had allocated the expenses of its moving departments in calculating its domestic and export selling expense ratios. Nan Ya also had made small revisions to its domestic and export fiber sales denominators used to calculate the ratios. As discussed above, we verified that the revised data was accurate. Therefore, for the final determination we have used Nan Ya's revised home market and export indirect selling expense ratios, as verified. Comment 23: Imputed Credit Expenses on Certain U.S. Sales Petitioners' Comments: The petitioners calculate a theoretical credit period based upon a general description of the process used to transport merchandise to the United States presented by Nan Ya officials at verification. Upon comparing this theoretical credit period to the U.S. sales listing, the petitioners note that several transactions have reported credit periods that are less than the theoretical period. The petitioners argue that reported credit periods that are less than the theoretical credit period are unrealistic and should, therefore, be rejected for the final determination. As facts available, for the transactions with reported credit periods less than the theoretical period, the petitioners propose that the highest per-unit credit expense found in Nan Ya's U.S. sales listing be used. Nan Ya's Comments: Nan Ya points out that the petitioners' argument rests on the assumption that there can be no exception to its general practice. While the petitioners complain about credit periods that are less than the theoretical amount, Na Ya states that they failed to substantiate their argument. Contrary to the petitioners' allegations, Na Ya asserts that the Department verified the accuracy of its reporting with respect to credit periods on U.S. sales. Department's Position: Verification showed that for some U.S. sales, Nan Ya would negotiate its letter of credit with a bank in advance of the date of realization specified on the letter of credit. As a result, Nan Ya was able to receive cash in advance of the date indicated by the terms of payment on the sale. When early payments were received in this manner, Nan Ya calculated imputed interest expenses up until the date of the negotiation of the letter of credit. From the date of negotiation through the date of realization stipulated on the letter of credit, Nan Ya incurred interest expenses payable to the bank which it reported as part of its bank charges. Between reporting either imputed credit expenses or interest expenses included as part of actual bank charges, documentation examined at verification showed that Nan Ya had properly accounted for any "carrying" expenses it incurred, actual or imputed, between the date of shipment and date of realization of the letter of credit for all U.S. sales examined. Therefore, we have made no adjustment to the credit period. Comment 24: Bank Charges Petitioners' Comments: The petitioners note that in the course of tracing a sale at the verification of Nan Ya, the Department identified a bank statement showing that Nan Ya had paid a bank processing fee for this transaction. However, in its sales database, Nan Ya had not reported any bank fees for that particular sale. Given that there are other sales observations in Nan Ya's U.S. sales database showing no bank charges, the petitioners contend that it is reasonable to use the unreported bank charge amount found during the verification as the minimum unit bank charge for those U.S. sales for which Nan Ya either no bank charges or bank charges less than the minimum amount. Nan Ya's Comments: Nan Ya responds by stating that it is the Department's practice to make correction for minor errors found during the course of verification. The respondent argues that the Department discovered only one error in reported bank charges for only one sales observation during the sales trace; no other errors in reported bank charges were found for other sales observations. Moreover, the respondent continues, the one error in the reported bank charges was negligible, and this has no significant impact on the Department's analysis. Therefore, Nan Ya argues, for these reasons there is no basis to apply as facts available a minimum bank charge to all U.S. sales. Department's Position: We agree with Nan Ya that there is no basis to apply a minimum bank charge to all U.S. sales. We only identified one sale at verification for which bank charges had been incorrectly reported, and the error was relatively small. In the course of the entire sales verification, in verifying the reported bank charges for other sales, we noted no indication that this error was either widespread or anything other than inadvertent. Therefore, for the final determination, we have corrected for this error by adding in the unreported bank charge for this one sale only. Comment 25: Commission and Marine Insurance Petitioners' Comments: The petitioners argue that the Department should correct what Nan Ya has reported in the marine insurance and commissions fields for one of Nan Ya's U.S. sales based on information the Department examined at verification while tracing through various documentation for this sale. Moreover, the Department should apply these same corrections to all other U.S. sales for which similar information was reported in the marine insurance and commissions fields. Nan Ya's comments: Nan Ya replies that, contrary to the petitioners' arguments, no evidence on the record shows that any commission or marine insurance expenses were incurred on this sale. Nan Ya contends that this is further evidenced by the fact that it has never paid commissions on sales to the reported customer of this sale. Department's Position: We agree with the petitioners that, according to the sales confirmation and export order memorandum included in exhibit 9(e) of Nan Ya's Sales Verification Report, it appears as though Nan Ya paid a commission on this sale, though no commission was reported in the U.S. sales database. We also note that the bill of lading for this same sale specifies that two separate companies should be notified upon entry into the United States of that shipment: the company that Nan Ya has listed as the customer for this sale and a second company. This supports the conclusion that the company that Nan Ya has listed as the customer for this sale was, in fact, acting as a agent and was paid a commission for its services, whereas the second company was the ultimate customer. This scenario would be consistent with the seller-agent-customer arrangement between Nan Ya and these same two companies reported by Nan Ya for other U.S. sales. We further agree with the petitioners that, according to the sales confirmation and export order memorandum included in exhibit 9(e) of Nan Ya's Sales Verification Report, it appears as though Nan Ya paid marine insurance on transporting this shipment, though no marine insurance was reported in the U.S. sales database. We also agree with the petitioners that this also brings into question the accuracy of the reporting of marine insurance for other sales observations, where the reported sales terms indicate that marine insurance is included in the price, but where no marine insurance was reported. In its rebuttal comments, Nan Ya did not offer any alternative explanation for the information in the exhibit indicating that a commission and marine insurance was paid. In light of the record evidence, therefore, we conclude Nan Ya has made an error in reporting the commission expense for this sale. We identified no information at verification indicating that Nan Ya had inaccurately reported commissions on other sales. Accordingly, we are applying as facts available a commission to this sale only in the amount of the stated percentage in the verification exhibit documents. We further conclude, for the reasons cited above, that Nan Ya has made an error in reporting of marine insurance expense for this sale. Because this information is not on the record, pursuant to section 776(a) of the Act, the Department must use facts available. Accordingly, we are applying as facts available an average marine insurance rate to all seven of the U.S. sales for which the sales terms indicate that marine insurance was paid but where no marine insurance expenses were reported. This facts available marine insurance rate has been calculated, based on all Nan Ya's U.S. sales for which marine insurance as reported, as the average percentage of the reported marine insurance to the reported gross unit price, weighted by sales quantity. Comment 26: U.S. Short-Term Interest Rate Nan Ya's Comments: Nan Ya explains that the Department found at verification that it had made a simple arithmetic error in calculating the aggregate interest on one small group of loans. Since this error was inadvertent and had only a minor difference in the calculation of the overall U.S. interest rate, Nan Ya requests that the Department use the verified U.S. short-term interest rate. According to Nan Ya, it is the Department's practice to correct errors found at verification when the errors are isolated and do not affect the integrity of the data.(61) Petitioners' Comments: The petitioners did not comment on this issue. Department's Position: We agree with Nan Ya and have used the U.S. short-term interest found at verification to calculate U.S. credit expenses. Comment 27: Home Market Short-Term Interest Rate Petitioners' Comments: The petitioners argue that the Department should implement the changes to the reported inventory carrying costs and imputed credit expenses reported by Nan Ya as part of its corrections submitted at the commencement of verification on January 13, 2000. Nan Ya's Comments: Nan Ya did not comment on this issue. Department's Position: We agree with the petitioners and have done the same. C. Cost of Production/Constructed Value Issues Comment 28: Recovery of Inputs Nan Ya's Comments: Nan Ya argues that the Department should not adjust the reported costs for its claimed credit for EG which is recovered from its PSF production lines. According to Nan Ya, the regenerated PSF line uses EG as a catalyst agent in the production process and as a result EG is recovered and reused. Nan Ya argues that the Department verified that Nan Ya correctly reported its credit for recovered EG and, therefore, no adjustment is warranted. Petitioners' Comments: The petitioners did not comment on this issue. Department's Position: We have made no adjustment to the reported credit for recovered EG. Based on our review of production documents at verification, we confirmed that excess EG was recovered from the regenerated PSF line. Additionally, we found that the recovered EG was reused in the PSF production process. Therefore, it is appropriate that the cost of regenerated PSF products includes a credit for the recovered EG. Comment 29: Exchange Gains Nan Ya's Comments: Nan Ya argues that the Department should include net exchange gains from bonds and loans in the financial expense calculation even though these items were left out of the submitted costs. Additionally, Nan Ya believes that the Department should include net exchange gains from accounts payable in the G&A expense calculation. According to Nan Ya, the Department normally corrects inadvertent errors found during verification, see Notice of Final Determination of Sales at Less Than Fair Value: Static Random Access Memory Semiconductors From Taiwan, 63 FR 8909, 8929 (February 23, 1998). Petitioners' Comments: The petitioners did not comment on this issue. Department's Position: We have adjusted the financial expense and G&A calculation to include net exchange gains. At verification, we found that Nan Ya excluded these amounts from the calculation of the financial expense ratio and the G&A expense ratio. It is the Department's longstanding practice to include foreign exchange gains and losses on financial assets and liabilities in our cost of production and constructed value calculations, provided that the gains and losses are related to the company's manufacturing operations. See, e.g., Final Determination of Sales Less Than Fair Value: Polyethylene Tenephthalate Film, Sheet, and Strip From the Republic of Korea, 56 FR 16305, 16313 (April 22, 1991), and Final Determination of Sales at Less Than Fair Value: Certain Cut-to-Length Carbon-Quality Steel Plate Products from Indonesia, 64 FR 73164, 73174 (December 29, 1999). We have included the foreign exchange gains and losses attributable to Nan Ya's debt in the financial expense calculation. In addition, we have included foreign exchange gains and losses generated by accounts payable in the G&A expense calculation. Comment 30: Minor Verification Corrections Nan Ya's Comments: Nan Ya argues that the Department should include in the final determination the corrections presented at the beginning of verification because the corrections were submitted in a timely fashion and were verified by the Department. Petitioners' Comments: The petitioners argue that the Department should reject Nan Ya's request because the revisions were extensive and were not merely "minor" or "isolated" errors. The petitioners argue that the revisions to Nan Ya's database immediately prior to verification were substantial and extensive. According to Petitioners, the errors identified by Nan Ya at the beginning of the cost verification were so broad that the cost of manufacture for the majority of the control numbers changed. The petitioners argue that Nan Ya made unexplained major, radical revisions to its sales and cost database on multiple occasions during the course of the proceeding. The petitioners conclude that these changes, as well as the additional errors found at verification, effect the overall integrity of Nan Ya's data and justifies the application of total facts available. Department's Position: We do not believe that facts available is warranted in this case. The corrections presented by Nan Ya on the first day of verification were minor and were of the type typically identified by the respondents during preparation for verification. Nan Ya changed to a new computer system during August 1998. The old files were kept on CD ROM and it was not possible for Nan Ya to download the data. Therefore, in preparing its responses Nan Ya manually input a significant amount of data. In doing so, Nan Ya made input errors which it identified while preparing for verification. We found that the revised cost presented at the beginning of verification corrected these errors and were based on the records maintained by the company in the ordinary course of business. Additionally, we found that the corrections were minor because they did not change the reporting methodology, and simply corroborated, supported, and clarified information already on the record. Therefore, we have included the corrections for purposes of the final determination. Comment 31: Product-Specific Costs Petitioners' Comments: The petitioners argue that the Department should revise Nan Ya's reported product costs because Nan Ya did not report any cost differences for silicon-coated products. They state that silicon-coated PSF requires additional materials and processing which are very expensive. While the information the Department obtained at verification is new information, the petitioners claim, it clearly shows that there should be a cost difference for silicon-coated PSF. The petitioners assert that Nan Ya failed to provide documentation to support the amounts shown on the verification document. Therefore, the petitioners argue that the Department should rely on facts otherwise available with respect to all sales of silicon-coated PSF. The petitioners argue that as facts otherwise available for silicon-coated PSF, the Department should use the highest cost of manufacturing reported in Nan Ya's cost database as the cost of manufacturing for silicon products. Alternately, the petitioners state that the Department could use the cost of manufacturing of antibacterial finish products for the silicon-coated products. The petitioners suggest that the antibacterial cost differences would be an appropriate surrogate because this type of finish incurs additional processing and material as do the silicon-coated products. The petitioners argue that the information obtained at verification is insufficient for the Department to make a complete adjustment to account for the cost that Nan Ya incurred to produce silicon-coated PSF. Moreover, they assert that the Department should not use the silicon cost differential of FETL as facts available because this would allow respondents who are represented by the same counsel to select what data it will report. Nan Ya's Comments: Nan Ya disagrees with the petitioners' claim that silicon- coating costs are expensive and require additional procedures. Nan Ya argues that the production process for silicon-coated PSF does not differ from that of non-silicon-coated PSF. Nan Ya states that all PSF goes through the same process and the silicon coated products do not incur any additional labor or overhead expenses. Nan Ya alleges that the silicon-coating costs are so small that it does not create any significant cost differences as shown on the document supplied to the Department at verification. Nan Ya states that its cost accounting system does not differentiate the cost of silicon coating and non-silicon coating, and therefore, the Department should accept its reported cost without adjustment because its financial statements are in accordance with the generally accepted accounting practices of Taiwan. Nan Ya argues that if the Department decides to adjust the silicon coated PSF, then the Department should use either the data from verification or FETL's reported silicon-coating costs. Department's Position: Section 776(a)(2) of the Act provides that if an interested party or any other person--(A) withholds information that has been requested by the administering authority; (B) fails to provide such information by the deadlines for the submission of the information or in the form and manner requested, subject to subsections (c)(1) and (e) of section 782; (C) significantly impedes a proceeding under this title; or (D) provides such information but the information cannot be verified as provided in section 782(i), the administering authority shall, subject to section 782(d), use the facts otherwise available in reaching the applicable determination under this title. In addition, section 776(b) of the Act provides that, if the Department finds that an interested party "has failed to cooperate by not acting to the best of its ability to comply with a request for information," the Department may use information that is adverse to the interests of the party as the facts otherwise available. Adverse inferences are appropriate "to ensure that the party does not obtain a more favorable result by failing to cooperate than if it had cooperated fully." See Statement of Administrative Action (SAA) accompanying the Uruguay Round Agreements Act (URAA), H.R. Doc. No. 103-316 at 870 (1994). This provision is intended to induce parties to cooperate in antidumping proceedings. As illustrated below, the record evidence in this case demonstrates that Nan Ya withheld crucial information necessary to calculate an accurate dumping margin and thus failed to cooperate to the best of its ability under section 776(b) of the Act. We have, therefore, determined that the application of adverse facts available is warranted. We agree with the petitioners that Nan Ya failed to provide product-specific costs for silicon-coated products. Silicon coating is a matching characteristic identified in sections B and C of the Department's questionnaire. Thus, it should be taken into account in determining product-specific costs. In the Department's view, the requirement of product-specific sales and cost data is one of the most basic and significant requirements in performing the antidumping analysis and margin calculation. See Certain Cold-Rolled and Corrosion-Resistant Carbon Steel Flat Products from Korea: Final Results of Antidumping Duty Administrative Reviews, 63 FR 13170, 13201 (March 18, 1998). The specific physical characteristics which make up a control number are those physical characteristics determined to be the most significant in differentiating between products. These are the physical characteristics that define a unique product for sales comparison purposes. Nan Ya spread the costs of silicon coating over all products rather than only to those products which were actually silicon coated. Absent product-specific cost of production information, the Department lacks the necessary information to calculate a difference-in-merchandise adjustment to account for differences in physical characteristics when comparing sales of similar merchandise. In addition, without this information, we cannot perform an accurate sales-below-cost test, nor can we determine accurate constructed values for use as normal value, as required. In numerous cases, a respondent's normal cost accounting system does not differentiate nor provide product-specific costs to the level of detail required by the Department. Our consistent practice, however, is to have respondents start with the costs established in their normal cost accounting system and then further allocate the costs to specific products based upon a reasonable method available to them. See section 773(f)(1)(A) of the Act. See also Carbon Steel Flat Products from Korea, 63 FR at 13201; and Plate from Indonesia, 64 FR at 73174. While Nan Ya's financial and cost accounting records may not contain the information requested on separate product costs, the company could have developed a reasonable allocation methodology to allocate costs to products on a product-specific basis using the company's normal cost accounting records as a starting point to calculate product specific costs. This is precisely what Nan Ya did for antibacterial finished products. That is, while Nan Ya's normal cost accounting records do not track differences in antibacterial coating cost, Nan Ya did identify and report cost differences for antibacterial finish. On page 3 of its September 3, 1999, response, Nan Ya claimed that "other finishing material costs are so small that it does not create differences in production costs." At verification, we attempted to obtain documentation to support Nan Ya's claim that the cost of silicon coating is minor. Rather than provide factual information such as invoices for purchases of the silicon coating material and quantities of this material consumed in making PSF, Nan Ya simply provided a copy of a worksheet which it claimed shows the cost of fiber with and without silicon coating. The company, however, was unable to demonstrate that the information on the worksheet related to the PSF under investigation. In fact, the values on this worksheet are drastically different from the submitted cost information for PSF. We asked company officials to provide documentation to support the figures on the worksheet and were told that they were unable to do so. We also asked company officials whether there were any other documents available which could be used to identify the quantity and value of silicone applied to PSF production and were told that they did not have any further documentation. As a result, Nan Ya failed to support its claim that silicon coating is so minor that it should not be included in product- specific costs used in the final determination. Because Nan Ya failed to provide us with any verified data to support its claim that the cost differences for silicon coating are minor and it did not provide product-specific costs to account for differences in silicon coating, we find that Nan Ya did not act to the best of its ability in accordance with section 776(b) of the Act . We do not consider it an unreasonable request to have Nan Ya provide documentation, such as invoices, showing how much it paid for the silicon additive used to coat the PSF, and showing how much it consumed. As a result of Nan Ya's failure to comply with our request for information, we have relied on adverse facts available to value the cost of silicon coating in accordance with section 776(b) of the Act. For the final determination, as adverse facts available, we increased Nan Ya's reported COM for silicon-coated products, which as noted above only includes a diluted amount of silicon coating costs, using data from the other respondent in this case. Moreover, because Nan Ya's failure to provide silicon cost data precluded us from calculating accurate DIFMER adjustments, we did not allow a DIFMER adjustment when a home market silicon coated product was matched to a non- silicon coated product. Comment 32: General and Administrative Cost Petitioners' Comments: The petitioners argue that the Department should include Nan Ya's other operating costs in G&A expenses. The petitioners claim Nan Ya improperly included this amount in the denominator of the G&A rate calculation and failed to include the other operating expenses anywhere in the reported costs. The petitioners state that at verification the Department confirmed that the other operating expenses included maintenance expenses and expenses related to operating revenues. Nan Ya's Comments: Nan Ya did not respond to this issue. Department's Position: Nan Ya's other operating costs were not included in the reported costs (i.e., neither COM nor G&A). We agree that the other operating costs comprised mainly maintenance expenses and should be included in the G&A expense ratio because they relate to the production of the company as a whole. For the final determination we have excluded the other operating costs from the denominator of the G&A rate calculation and instead included the expenses in the numerator of the G&A expenses ratio calculation. Comment 33: Long-term Interest Income Petitioners' Comments: The petitioners argue that the Department should exclude the long-term interest income from the calculation of the financial expense ratio. Nan Ya's Comments: Nan Ya did not respond to this issue. Department's Position: Nan Ya included long-term interest income as an offset to interest expense in the calculation of the financial expense ratio. In calculating cost of production and CV, it is the Department's practice to allow a respondent to offset (i.e., reduce) financial expenses with short-term interest income earned from the company's working capital, see Notice of Final Results of Antidumping Duty Administrative Review: Silicon Metal from Brazil, 64 FR 6305, 6314 (February 9, 1999). In calculating a company's cost of financing, we recognize that, in order to maintain its operations and business activities, a company must maintain a working capital reserve to meet its daily cash requirements (e.g., payroll, suppliers, etc.). The Department further recognizes that companies normally maintain this working capital reserve in interest-bearing accounts. The Department therefore allows a company to offset its financial expense with the short-term interest income earned on these working capital accounts. The Department does not, however, allow a company to offset its financial expense with income earned from investing activities (e.g., long-term interest income, capital gains, dividend income) because such activities are not related to the current operations of the company. We note that the CIT has upheld the Department's approach to calculating the financial expense offset with only short-term interest income. See Gulf States Tube Division of Quanex Corp. v. United States, 981 F. Supp. 630, 651 (CIT 1997), and NTN Bearing Corp. v. United States, 905 F. Supp.1083, 1097 (CIT 1995) (quoting Timken Co. v. United States, 852 F. Supp. 1040, 1048 (CIT 1994)), in which the CIT held that, to qualify for an offset, interest income must be related to the "ordinary operations of the company." Therefore, for the final determination, we excluded long-term interest income in the calculation of the financial expense ratio. Comment 34: Packing Expenses Petitioners' Comments: The petitioners argue that the Department should apply the G&A and interest ratios to the sum of the total cost of manufacture and packing expenses. The petitioners state that at verification the Department confirmed that Nan Ya included packing expenses in the denominator (cost of goods sold) used to calculate the G&A and interest expense ratios. The petitioners further state that Nan Ya's reported COM did not include packing. The petitioners claim that because the G&A and interest ratios are applied to the reported COM, including packing expenses in the denominator of the G&A and interest ratios would result in understated product-specific G&A and interest expenses. The petitioners argue that although it is the Department's preferred method of making an adjustment to remove packing expenses from the cost of goods sold denominator, often the company-wide packing expenses are not available, as is the case in this investigation. Nan Ya's Comments: Nan Ya did not respond to this issue. Department's Position: We applied the G&A and interest ratios on the same basis on which they were calculated. We confirmed that Nan Ya included packing expenses in the denominator (cost of goods sold) used to calculate the G&A and interest expense ratios and that the COM to which the ratios were applied excluded packing. In this case, as in The Final Determination of Sales at Less Than Fair Value: Stainless Steel Sheet and Strip in Coils From Japan, 64 FR 30574, 3059 (June 8, 1999), the company-wide packing expenses are not available. Thus, for the final determination, we have applied the G&A and interest expense ratios to the total cost of manufacturing plus average packing expense of each control number to calculate G&A and interest expense. Comment 35: Unreported Costs Petitioners' Comments: The petitioners argue that the Department should account for the costs that Nan Ya omitted from its cost calculation. According to the petitioners, Nan Ya did not include the costs described as "cost differences not counted for in FG-in cost," which consists of costs that were not assigned to any products at the polyester fiber division. The petitioners argue that these cost differences were due to process costs and mistakes not assigned to material and overhead expenses properly and therefore, the Department should increase Nan Ya's reported cost of manufacture to account for those cost differences. In addition, the petitioners argue that the Department should make an adjustment to account for the understatement of Nan Ya's reported cost of manufacture compared to the cost recorded in Nan Ya's financial records. Nan Ya's Comments: Nan Ya agues that the Department should reject Petitioner's argument to adjust Nan Ya's costs based on "cost differences not accounted for in FG-in cost" and the difference in the reported cost versus the costs reported in the financial records because they are unsupported by evidence on the record. Nan Ya claims that the Department verified the total reported cost of manufacture excluding packing labor and materials costs, whereas the cost recorded in Nan Ya's financial records include packing labor and materials. Nan Ya therefore argues that its reported cost of production was actually overstated. Department's Position: The "cost differences not accounted for in FG-in costs" represent costs that were incurred at the polyester fiber division but not allocated to any products. Because these costs relate to production of PSF, for the final determination, we increased Nan Ya's reported cost of manufacture to include these unallocated costs. We find that no adjustment is necessary to account for the claimed difference between Nan Ya's reported COM and its COM recorded in the financial records. Nan Ya's total reported cost of manufacturing excludes packing labor and materials, whereas the cost recorded in Nan Ya's financial records include these costs. Taking into account the packing cost difference, the reported total cost of manufacturing are not understated. Therefore, we made no adjustment. Comment 36: Revised Yields Petitioners' Comments: The petitioners argue that the Department should adjust Nan Ya's reported costs for production line 1B based on quantities used in Nan Ya's revised yield rate calculations. According to the petitioners, Nan Ya revised its yield calculations for production line 1B, cost center 5126, immediately prior to the Department's verification. The petitioners state that Nan Ya did not recalculate the per-unit costs based on the revised quantities used to calculate the revised yields. According to the petitioners, the quantity of material usage directly affects the unit material cost calculated for the products from the cost center. The petitioners therefore argue that the Department should adjust Nan Ya's reported cost of manufacture for the product control numbers affected. Nan Ya's Comments: Nan Ya states that the petitioner's arguments were contrary to the evidence on the record and thus should be rejected. Nan Ya argues that the petitioners misunderstood Nan Ya's calculation for cost center 5126. Nan Ya notes that the per-unit costs at this cost center are based on the total input value, which is directly quoted from the cost of manufacturing chart prepared by Nan Ya's ordinary cost system, and its total production quantity. Thus, Nan Ya argues that the error in the total material input quantity is irrelevant to the calculated per-unit cost at this cost center. Department's Position: During verification we reviewed the revised yields presented by Nan Ya on the first day of verification (see Nan Ya's Cost Verification Report at 26). While the minor correction affects the yield rate calculation for this cost center, the calculated yield rate was not used to determine the per-unit raw material and other costs at this cost center. Rather, the per-unit cost for this cost center was determined by dividing the total cost of inputs by the output quantity. Accordingly, we made no adjustment. Comment 37: Positive Yields Petitioners' Comments: The petitioners argue that the Department should adjust Nan Ya's reported cost for products made from the cost centers for which Nan Ya reported positive yields. They state that positive yields are impossible. Nan Ya's Comments: Nan Ya states that the Department should reject the Petitioner's argument that its reported costs be adjusted by the weighted- average yield or yield loss at the packing stage. Nan Ya argues that the Department verified that it correctly reported input and output quantities and that the Department recognized there were small positive yields at some of the cost centers at the final production stage due to the retention of oil and water. Department's Position: At verification we asked company personnel how the output weight for certain cost centers could exceed the weight of input raw materials. The company's production manager explained that the production output at the final stage of the production process can exceed the input quantity of materials due to the absorption of oil and water by the materials being processed (see Nan Ya's Cost Verification Report at 27). We note that this issue only pertains to certain production lines and only for the last production stage. Additionally, we noted that the cost of the oil and water was included in the reported cost. Nan Ya demonstrated that the other earlier stages of production which include the polycondensation and spinning stages, experienced yield losses which were included in the reported costs. We obtained a worksheet which shows the output production quantity and material input quantity at the last stage of each production line and traced the quantities to production statistics reports. As we noted no errors or inconsistencies and since there is no evidence on the record to support the petitioners' claim that the reported yields are incorrect, we do not consider an adjustment warranted. Comment 38: Scrap Credit Petitioners' Comments: The petitioners contend that the Department should disallow Nan Ya's claimed scrap credit for the fiber production stage. The petitioners state that for some of the cost centers, the scrap credit exceeded the scrap generated based on Nan Ya's production statistics reports and therefore, the scrap credit is overstated. Nan Ya's Comments: Nan Ya maintains that the petitioner's contention that Nan Ya's fiber scrap credit should be disallowed is unsubstantiated and therefore should be rejected. Nan Ya argues that the Department verified that there is only a small difference between the amount of scrap produced and the credit claimed in certain cost centers, and that the scrap in these cost centers represents a small percentage of the total cost of manufacture of the products produced at those lines. Nan Ya concludes that because the petitioners failed to substantiate their argument, and because the difference is negligible the Department should reject the petitioners' argument. Department's Position: At verification, we compared the reported fiber scrap credit to the amount of scrap generated as reported on the company's production statistics reports. We found that Nan Ya's fiber scrap credit exceeded the amount of scrap actually produced. Therefore, we have reduced the scrap credit by the amount of the overstatement. Comment 39: Inputs from Affiliates Petitioners' Comments: The petitioners argue that the Department should adjust the cost of production for PTA received from its affiliate, Formosa Chemical & Fiber Corporation (FCFC), before applying the major input rule. The petitioners argue that there were many errors in FCFC's reported cost of production for PTA including: 1) FCFC produced PTA at two plants during the POI but only reported the production quantity and costs for one of the plants; 2) Nan Ya did not include the overhead costs in its cost calculations for PTA costs; and 3) Nan Ya did not include "other expenses" in the SG&A ratio and did not substantiate that these expenses should be excluded. The petitioners argue that the Department should use the highest cost of production on the record due to Nan Ya's failure to substantiate costs for the major input, but at a minimum it should adjust FCFC's reported cost of production to account for the SG&A expenses that Nan Ya omitted from its cost calculation. Nan Ya's Comments: Nan Ya argues that the Department should reject the petitioners' argument on FCFC's cost of production for PTA because it is not supported by evidence on the record. Nan Ya claims that it was correct in not including the costs for the start-up period of FCFC's new plant from October through December 1998 in the cost of production of PTA because the new plant was not fully operational. Nan Ya also argues that the Department verified that the per-unit COM of the PTA at the new plant was lower than the per-unit COM reported. Nan Ya asserts that during October through December 1998 FCFC's new plant satisfies the requirements for a start-up adjustment in accordance with section 773(f)(1)(C)(ii) of the Act(62). Nan Ya argues that the record evidence demonstrates that the plant in question is a new plant, and that the plant just started production but had not yet reached commercial production levels. Therefore, Nan Ya contends, it properly excluded FCFC's new plant cost incurred during the period October through December 1998. Furthermore, Nan Ya argues that the Department should not adjust Nan Ya's PTA cost due to additional overhead costs and additional G&A expenses as the petitioners suggest. Nan Ya argues that these costs will only increase the PTA cost by a small percentage. Nan Ya argues that if these adjustments were made to FCFC's revised cost of production submitted the first day of verification, the total adjustment would not exceed the amount of the over-adjustment that Nan Ya made to its PTA costs. Department's Position: Nan Ya failed to include the production quantity and production related costs for one of FCFC's plants that produced PTA. We do not believe that a start-up adjustment is justified for this plant. Nan Ya failed to request the adjustment in a timely manner; moreover, it did not provide any of the required support for the adjustment. Nan Ya also failed to include certain overhead costs in its cost calculations for PTA and "other expenses" in its SG&A ratio calculation. On the first day of verification, Nan Ya provided a revised cost of production of PTA. Nan Ya, however, did not change its cost database to reflect the changes in the PTA cost that were identified at verification. For the final determination, we computed the net effect of these changes and each of the adjustments noted above, and have adjusted the cost of production of PTA accordingly. We revised the reported cost of manufacture of PSF, in accordance with sections 773(b)(3) and 773(f)(3) of the Act, to include the corrected cost of PTA because it was higher than either the market price or the transfer price. 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 _________ Richard W. Moreland Acting Assistant Secretary for Import Administration March 22, 2000 -------------------------------------------------------------------------------- footnotes 1. Notice of Preliminary Determination of Sales at Not Less Than Fair Value and Preliminary Negative Critical Circumstances Determination: Certain Polyester Staple Fiber from Taiwan, 64 FR 60771 (November 8, 1999) (Preliminary Determination). 2. Arteva Specialties S.a.r.l.,d/b/a KoSa; Wellman, Inc; and Intercontinental Polymers, Inc. 3. These, and additional problems identified by the petitioners, are addressed in detail in subsequent comments below. 4. As examples of the Department's use of best information available, the petitioners cite to Circular Welded Non-Alloy Steel Pipe from Brazil, 57 FR 42940 (September 17, 1992) and Certain Circular Welded Non-Alloy Steel Pipe from Mexico; Final Results of Review, 65 FR 6136 (February 8, 2000). 5. Section 782(e) of the Act. 6. The petitioners' Case Brief at 2. 7. Id. at 9. 8. According to section 351.301(b)(1) of the Department's regulations, in an investigation a submission of factual information is normally due no later than seven days prior to the start of verification. Verifications in this proceeding commenced on January 10, 2000. 9. Section 731 of the Act. 10. See e.g., Shakeproof Assembly Components Division of Illinois Toolworkers Inc. v. United States, slip op. 99-70, 59 F. Supp.2d 1354, 1358 (CIT 1999), (quoting Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed. Cir. 1990); and Lasko Metal Products Inc. v. United States, 43 F.3d 1442, 1446 (Fed. Cir. 1994); Union Camp Corp. v. United States, 1999 WL 342438 (CIT 1999); Olympia Indus. v. United States, 7 F. Supp.2d 997, 1000-01 (CIT 1998)). 11. Section 351.301(c) of the Department's regulations. 12. The petitioners also asked the Department to solicit new or revised information in additional submissions on the following dates: July 14, 1999; July 30, 1999; August 5, 1999; and August 17, 1999. 13. Department Letter to William Clinton at White & Case (September 14, 1999). 14. Letter to the Secretary of Commerce from White & Case (September 20, 1999). 15. Memorandum to the Case File from the Case Team (October 29, 1999). 16. Id. at 5. 17. FETL's Sales Verification Report at 6. 18. The petitioners' Case Brief at 15. 19. Letter to the Secretary of Commerce from Collier, Shannon, Rill & Scott (September 23, 1999) at page 3. 20. See section 782(d) of the Act. 21. A separate, detailed analysis of the problems found at verification in Nan Ya's reporting of its sale date for several of its U.S. sales is contained in the Memorandum to Richard Moreland from Case Team; Errors in Nan Ya's Reported Dates of Sale (March 22, 2000) (Dates of Sale Memorandum). 22. Nan Ya's Sales Verification Report at 13. 23. FETL Cost Verification Report, at 4-7. 24. FETL Sales Verification Report, at 1. 25. See e.g., Sulfanilic Acid from the People's Republic of China, 63 FR 63834, 63837 (November 17, 1998); Brake Rotors From the People's Republic of China: Postponement of Final Results of Antidumping Duty New Shipper Administrative Review, 64 FR 9972, 9977 (March 1, 1999); Certain Corrosion- Resistant Carbon Steel Flat Products and Certain Cut-to-Length Carbon Steel Plate From Canada: Final Results of Antidumping Duty Administrative Reviews and Determination To Revoke in Part, 62 FR 2173, 2177 (January 13, 1999). 26. See Notice of Final Determination of Sales at Less Than Fair Value: Certain Cut-to-Length Carbon-Quality Steel Plate Products from Indonesia, 64 FR 73164, 73172, (December 29, 1999) (Comment 4). 27. See FETL's Cost Verification Report at Memorandum to Neal M. Halper through Theresa L. Caherty from Gina K. Lee (February 10, 2000) at 13, and FETL's Cost Verification Exhibit B6 at 4. 28. Although it may appear that the blended category pertains to blends of finished fibers, a review of the interested parties' submissions makes clear that blended refers blends of input types. See, for example, Letter to Secretary of Commerce from Collier, Shannon, Rill & Scott (May 25, 1999). 29. FETL's Sales Verification Report at 12. 30. FETL's explanation of its coding methodology also sheds light on the statements of FETL's counsel, at the March 10, 2000 oral hearing, to the effect that, in at least one sense, most of FETL's products could be viewed as blended because they contain some amount of non-virgin inputs. 31. As is indicated throughout FETL's written submissions, and evidenced at verification, FETL often uses terms such as recycled, regenerated, waste, blended, and condux interchangeably. This explains in part FETL's description in its earlier questionnaire responses of its regenerated production despite the fact that it only manufactures virgin and blended merchandise. 32. At verification, we understood the statement that "FETL has no 100 percent virgin lines" to mean that there are no lines that use 100 virgin raw materials. Consistent with officials' other statements in the verification report, even in the production of PSF reported as virgin FETL uses some small amount (less than 0.1 percent) of waste inputs. 33. According to section 351.301(b)(1) of the Department's regulations, in an investigation, a submission of factual information is normally due no later than seven days prior to the start of verification. Verifications in this proceeding commenced on January 10, 2000. 34. See e.g., sales verification exhibits 6(a) through 6(d). 35. This is evidenced in the fact that FETL's internal production codes do not identify the fiber type of each product.(36) 36. See FETL's January 4, 2000 supplemental response. 37. See e.g., FETL's August 27 supplemental response at D-3 and the Sales Verification Report at verification exhibit 6(b). 38. In other words, the same or similar products can be produced using more than one chip recipe. 39. As seen in Exhibit 6 of FETL's August 27 response, the PET chips are grouped according to other physical properties in addition to waste content. 40. See e.g., the Department's August 24, 1999 letter to interested inviting comments on the Department's proposed matching criteria. 41. The most detailed comments on antibacterial and flame-retardant properties were provided by a respondent in the concurrent investigation of PSF from Korea. See Letter to the Secretary of Commerce, from Dorsey & Whitney (May 26, 1999) (at 4) where the Korean respondent urges the Department adopt a separate category of matching criteria that breaks out "special treatment" versus "no treatment." 42. For instance, at the tour of one petitioner's plant, plant officials stated that the four most important physical characteristics were finish, cross- section, denier, and cut-length (see Memorandum to the File from Suresh Maniam and Alysia Wilson; Plant Tour (May 11 , 1999). We further note that the matching criteria for the category "Finish" adopted by the Department reflects those criteria proposed by the petitioners (see Letter to the Secretary of Commerce from Collier, Shannon, Rill & Scott (May 27, 1999) at 8. 43. It was the Department's understanding at the time that antibacterial properties in PSF were obtained only through finish treatments. 44. FETL's Sales Verification Report at 20. 45. Sales Verification Report at 27. 46. See 19 CFR 351.403(c). 47. The following home market sales observations reported by FETL were pre- selected: 1499, 30, 1156, 1225, 1252. Likewise, the following U.S. sales were pre-selected: 455, 341, 342, 365, 359. 48. See e.g., Notice of Final Determination of Sales at Less Than Fair Value: Stainless Steel Round Wire from Canada, 64 FR 17324, 17335 (April 9, 1999). 49. 64 FR 17324, 17335. 50. See section 773(f)(1)(A) of the Act. 51. See, e.g., Notice of Final Determination of Sales at Less Than Fair Value; Certain Hot-Rolled Flat-Rolled Carbon-Quality Steel Products From Brazil, 64 FR 38756 (July 19, 1999) (Comment 44); Notice of Final Determination of Sales at Less Than Fair Value: Emulsion Styrene-Butadiene Rubber From the Republic of Korea ("ESBR"), 64 FR 14865, 14871 (March 29, 1999) (Comment 7). 52. See Notice of Final Results of Antidumping Duty Administrative Review: Certain Pasta From Italy, 65 FR 7349, 7359 (February 14, 2000). 53. See, e.g., Round Wire (64 FR at 17340). 54. Slip op. 99-118 (CIT Oct. 29, 1999). 55. Nan Ya's Sales Verification Report at 22. 56. See Notice of Final Determination of Sales at Less Thank Fair Value; Certain Hot-Rolled Flat-Rolled Carbon-Quality Steel Products From Brazil, (Steel from Brazil) 64 FR 38756, 38781 (July 19, 1999), and Certain Cold-Rolled and corrosion-Resistant Carbon Steel Flat Products From Korea; Final Results of Antidumping Duty Administrative Reviews, (Flat Products from Korea) 64 FR 12927 (March 16, 1999. 57. Nan Ya's Sales Verification Report at 30. 58. The Petitioners' Case Brief at 52. 59. See Nan Ya' Sales Verification Report at 29. 60. As a result of the revisions to Nan Ya's indirect selling expense data, as presented at verification, Nan Ya's domestic indirect selling expense ratio decreased by about 37 percent, where as the export indirect selling expense ratio increased by a factor of roughly 4.5. 61. See Certain Corrosion-Resistant Carbon Steel Flat Products and Certain Cut- to-Length Carbon Steel Plate From Canada, 62 FR 2173, 2177 (January 13, 1999), and Brake Rotors From the People's Republic of China, 64 FR 9972, 9977 (March 1, 1999). 62. Section 773(f)(1)(C)(ii) of the Act states that the Department will make an adjustment for startup operations only where: (1) a producer is using new production facilities or producing a new product that requires substantial additional investment, and (2) production levels are limited by technical factors associated with the initial phase of commercial production.