NTSF Seafoods Joint Stock Co. v. United States Catfish Farmers of America v. United States, No. 20-00104 (Ct. Int'l Trade 2022)

Annotate this Case

Download PDF

Slip Op. 22-38 UNITED STATES COURT OF INTERNATIONAL TRADE Court No. 20-00104 Court No. 20-00105 NTSF SEAFOODS JOINT STOCK CO., CATFISH FARMERS OF AMERICA, et al., Plaintiff, Plaintiffs, v. v. UNITED STATES, UNITED STATES, Defendant, Defendant, and and CATFISH FARMERS OF AMERICA, et al., NTSF SEAFOODS JOINT STOCK CO., Defendant-Intervenor. Defendant-Intervenor. Before: M. Miller Baker, Judge OPINION [In Case 20-104, the court denies Plaintiff’s motion for judgment on the agency record and grants judgment for Defendant and Defendant-Intervenor. In Case 20-105, the court grants Plaintiff’s motion for judgment on the agency record in part, denies it in part, and remands to Commerce.] Dated: April 25, 2022 Jonathan M. Freed and Kenneth N. Hammer, Trade Pacific PLLC of Washington, DC, for NTSF Seafoods Joint Stock Co., plaintiff in Case 20-104 and defen- Ct. Nos. 20-00104, 20-00105 Page 2 dant-intervenor in Case 20-105. With them on the briefs was Robert G. Gosselink. Jonathan M. Zielinski, Cassidy Levy Kent (USA) LLP of Washington, DC, for Catfish Farmers of America et al., plaintiffs in Case 20-105 and defendant-intervenors in Case 20-104. With him on the briefs was James R. Cannon, Jr. Kara M. Westercamp, Trial Attorney, Commercial Litigation Branch, U.S. Department of Justice of Washington, DC, for Defendant. With her on the brief were Brian Boynton, Acting Assistant Attorney General; Jeanne E. Davidson, Director; and Patricia M. McCarthy, Assistant Director. Of counsel on the brief was Kirrin Hough, Attorney, Office of the Chief Counsel for Trade Enforcement and Compliance, U.S. Department of Commerce of Washington, DC. Baker, Judge: In litigation, as in war, the enemy of my enemy is usually my friend. But in these two sprawling cases that arise out of the Department of Commerce’s 15th administrative review of its antidumping order applicable to certain imported fish from Vietnam, the enemy of my enemy turns out to also be my enemy. In one case, the plaintiff—a Vietnamese fish producer and exporter—contends that Commerce was too harsh. The plaintiffs in the other case—domestic catfish producers—contend that the Department was not harsh enough. The government, caught in the middle, Ct. Nos. 20-00104, 20-00105 Page 3 finds itself defending a two-front war against both the Vietnamese producer and the domestic producers. In Case 20-104, where the Vietnamese producer claims that Commerce was too harsh, the court denies the producer’s motion for judgment on the agency record and instead enters judgment for the government and domestic producers. In Case 20-105, where domestic producers claim that the Department was not harsh enough, the court grants their motion for judgment on the agency record in part and denies it in part, and remands for further administrative proceedings. Factual and Procedural Background A 2003 antidumping order for frozen fish imported from Vietnam provides the backdrop to this litigation. See Notice of Antidumping Duty Order: Certain Frozen Fish Fillets from the Socialist Republic of Vietnam, 68 Fed. Reg. 47,909 (Dep’t Commerce Aug. 12, 2003). There, Commerce found that certain frozen fish from Vietnam were being sold in the U.S. at less than normal value and imposed duties to make up the difference. The order imposed specific rates for certain exporters and a “Vietnam-wide” rate for all others. See 68 Fed. Reg. at 47,909–10. In the intervening years, Ct. Nos. 20-00104, 20-00105 Page 4 that order has undergone multiple administrative reviews. 1 Both cases here present issues arising out of the 15th such review, which Commerce initiated in 2018 at the request of a domestic trade association, Catfish Farmers of America, and several of its constituent members (collectively, Catfish Farmers). See Initiation of Antidumping and Countervailing Duty Administrative Reviews, 83 Fed. Reg. 50,077, 50,080–81 (Dep’t Commerce Oct. 4, 2018). The period of review was August 1, 2017, to July 31, 2018. See id. at 50,080. A. Proceedings before Commerce 1. Preliminary determination Because the fish in question are produced in Vietnam, a country with a non-market economy, the statute requires Commerce to calculate the production costs—in the statutory vernacular, the “factors of production”—“based on the best available information” as to such costs “in a market economy country or countries considered to be appropriate by [the Department].” 19 U.S.C. § 1677b(c)(1); see also Hung Vuong, 483 F. Supp. 3d at 1339–41 (describing antidumping proceedings involving non-market economies). For a primer on antidumping orders and administrative reviews of those orders, see Hung Vuong Corp. v. United States, 483 F. Supp. 3d 1321, 1334–1341 (CIT 2020). 1 Ct. Nos. 20-00104, 20-00105 Page 5 The “market economy country or countries” referred to in the statute are known as “surrogate countries.” See, e.g., 19 C.F.R. § 351.408(c)(2) (“[The Department] normally will value all factors in a single surrogate country.”). To select surrogate country candidates, the statute directs Commerce to use, “to the extent possible,” market economy countries that “are—(A) at a level of economic development comparable to that of the nonmarket economy country, and (B) significant producers of comparable merchandise.” 19 U.S.C. § 1677b(c)(4). After initiating its review here, Commerce identified six potential surrogate countries it found to be at a comparable level of economic development to Vietnam based on 2017 gross national income data from the World Bank: Bolivia, Egypt, Honduras, Nicaragua, Nigeria, and India. Appx16539. Commerce further found that India was a significant producer of comparable merchandise because “[i]nformation on the record” so established and because no interested party had submitted any information about the other five potential surrogate countries identified by the Department. Appx16540. Finally, the Department determined that Indian factors of production data submitted by NTSF Seafoods Joint Stock Co. were superior to competing Indonesian data submitted by Catfish Farmers. Appx16540–41. As a result, Commerce selected India as the primary surrogate country for valuing the factors of production. Appx16542. Ct. Nos. 20-00104, 20-00105 Page 6 Commerce selected NTSF as the sole mandatory respondent and issued questionnaires to the company seeking information about its factors of production. Appx16542. NTSF responded by providing Commerce a database that the company said included factors of production data for it and its affiliated fish farming operation, NTSF Vinh Long (Vinh Long). Appx89843– 89844. Commerce preliminarily calculated a $0.00-per-kilogram dumping margin based on NTSF’s responses, subject to various adjustments. Appx16542, Appx16548. As relevant here, one of the adjustments to NTSF’s data involved deducting certain “movement expenses” from NTSF’s reported gross unit price per 19 U.S.C. § 1677a(c)(2)(A). 2. Verification After issuing its preliminary determination, the Department conducted verification of NTSF’s questionnaire responses in Vietnam. During verification, NTSF informed the Department that it had not reported Vinh Long’s farming factors in the database despite having previously said it had done so. Appx102468–102469, Appx1026. NTSF stated it discovered the error while preparing for verification and therefore sought to provide corrected data on the first day of verification, characterizing it as a “minor correction.” Case 20-104, ECF 40-1, at 16–17. Commerce refused to accept the corrected data because it “represented significant new factual information.” Ct. Nos. 20-00104, 20-00105 Page 7 Appx1026; Appx102454 n.1 (“[T]he verification team informed company officials that the correction would not be accepted as ‘minor.’ ”). 3. Final determination After verification, Commerce received briefing from the parties. The Department then issued its final determination, which assigned NTSF an antidumping rate of 15¢ per kilogram instead of the zero rate from the preliminary determination. Appx1002. As relevant here, the Department (1) declined Catfish Farmers’ request to reject all NTSF’s data and apply total facts otherwise available with an adverse inference in determining NTSF’s dumping margin, Appx1008, Appx1011–1015; (2) reaffirmed (over Catfish Farmers’ objection) its decision to use India rather than Indonesia as the relevant surrogate country for determining normal value, Appx1015–1025; (3) applied partial facts otherwise available with an adverse inference to calculate the portion of NTSF’s normal value based on farming factors of production relating to Vinh Long, Appx1025–1027; and (4) reaffirmed its calculation of international movement expenses. 2 Appx1024–1025. After Commerce issued its final determination, NTSF filed a “ministerial error allegation” asserting that the Department’s calculations were erroneous. Appx17350– 17353. Commerce rejected the ministerial error allegation as untimely, reasoning that although NTSF could have raised the issue in its case brief (and even earlier), it 2 Ct. Nos. 20-00104, 20-00105 Page 8 B. This litigation NTSF and Catfish Farmers brought these two cases challenging Commerce’s final determination. See Case 20-104, ECF 7 (NTSF complaint); Case 20-105, ECF 7 (Catfish Farmers complaint). Each then intervened in the other’s case to defend the final determination from the other’s challenge. The court consolidated these cases for briefing and argument. Case 20-104, ECF 25; Case 20-105, ECF 26. The plaintiffs then filed their pending Rule 56.2 motions for judgment on the agency record. Case 20-104, ECF 38 (confidential) and 40 (public); Case 20-105, ECF 31 (confidential) and 32 (public); see also USCIT R. 56.2. The government (Case 20-104, ECF 48 (public) and 49 (confidential); Case 20-105, ECF 42 (public) and 43 (confidential)) and the intervenors (Case 20-104, ECF 50 (public) and 51 (confidential); Case 20-105, ECF 44 (confidential) and 45 (public)) oppose. The court then heard oral argument. Jurisdiction and Standard of Review The court has subject-matter jurisdiction under 28 U.S.C. § 1581(c). In actions such as this brought under 19 U.S.C. § 1516a(a)(2)(A)(i)(II), “[t]he court shall hold unlawful instead waited until after the final determination. Appx17341. Ct. Nos. 20-00104, 20-00105 Page 9 any determination, finding, or conclusion found . . . to be unsupported by substantial evidence on the record, or otherwise not in accordance with law . . . .” 19 U.S.C. § 1516a(b)(1)(B)(i). As to evidentiary issues, the question is whether the administrative record, taken as a whole, permits Commerce’s conclusion, even if the court might have weighed the evidence differently: Substantial evidence has been defined as more than a mere scintilla, as such relevant evidence as a reasonable mind might accept as adequate to support a conclusion. To determine if substantial evidence exists, we review the record as a whole, including evidence that supports as well as evidence that fairly detracts from the substantiality of the evidence. Nippon Steel Corp. v. United States, 337 F.3d 1373, 1379 (Fed. Cir. 2003) (cleaned up). As to legal questions, the familiar framework of Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842–45 (1984), governs judicial review of Commerce’s interpretation of the antidumping statute. See United States v. Eurodif S.A., 555 U.S. 305, 316 (2009) (Commerce’s “interpretation governs in the absence of unambiguous statutory language to the contrary or unreasonable resolution of language that is ambiguous”). Ct. Nos. 20-00104, 20-00105 Page 10 Discussion I. NTSF’s case (No. 20-104) A. Commerce’s decision to apply partial adverse facts available At verification, NTSF informed Commerce that, contrary to the representations in the company’s questionnaire responses, it did not report the farming factors of production of its affiliate, Vinh Long, 3 and simultaneously tendered the missing information. Appx1026. The Department refused to accept the information, id., and applied partial facts otherwise available under 19 U.S.C. § 1677e(a) based on NTSF’s “withhold[ing]” of requested information, failure to provide information in the “form and manner requested,” and “significantly imped[ing]” Commerce’s investigation. Appx1026–1027; see also 19 U.S.C. § 1677e(a)(2)(A), (B), and (C). Commerce also decided to employ an adverse inference in applying such facts based on NTSF’s failure to cooperate to the best of its ability. Appx1026–1027; see also 19 U.S.C. § 1677e(b). NTSF now argues that Commerce never asked for the information that (1) the company at first said it supplied in its questionnaire responses, (2) at verification sheepishly admitted not supplying, and then As NTSF explains in its brief, Vinh Long farmed fish that in turn were later processed at NTSF’s production facility before being exported to the U.S. See ECF 40-1, at 11 n.3. 3 Ct. Nos. 20-00104, 20-00105 Page 11 (3) tendered to the Department out of time. Cf. Marks v. Comm’r, 947 F.2d 983, 986 (D.C. Cir. 1991) (“It is quite apparent that the reason the Markses kept the Commissioner—and the government—unapprised of their whereabouts was because they were fugitives from criminal prosecution. To turn around and blame the Commissioner for not finding them runs afoul of this court’s developing ‘chutzpah’ doctrine.”). NTSF “may be guilty of chutzpah, but [the court] must consider the merits of its argument anyway.” Int’l Union of Operating Eng’rs v. Cnty. of Plumas, 559 F.3d 1041, 1044 (9th Cir. 2009). NTSF argues that it processes all its fish at a single facility, and that it reported a single company-wide per-kilogram amount of each farming factor required to process and pack one kilogram of fish. Case 20-104, ECF 40-1, at 19. NTSF further contends that, “[c]ritically, Commerce did not request that [factor-of-production] usage be reported separately at each facility where [factors of production] are consumed in the production of upstream inputs later used to produce [merchandise under consideration].” Id. at 19–20 (emphasis added). “In other words, NTSF reported per-unit farming [factors of production] consumption reflecting its consumption of each farming [factor of production] required to produce a [kilogram] of subject merchandise regardless of whether the whole fish material input was produced at NTSF’s own farms or at NTSF Vinh Long’s farms.” Id. at 20. NTSF argues that this is all Commerce directed NTSF to do—to report factor usage and output Ct. Nos. 20-00104, 20-00105 Page 12 “at each facility at which it produces subject merchandise (i.e., frozen fish fillets).” Id. (citing Appx1120). The government responds by quoting the questionnaire, which required respondents to: report factors information for all models or product types in the U.S. market sales listing submitted by you (or the exporter) in response to Section C of the questionnaire, including that portion of the production that was not destined for the United States. The reported amounts should reflect the factors of production used to produce one unit of the merchandise under consideration. Case 20-104, ECF 48, at 55 (emphasis removed) (quoting Appx6908). The government describes this language as requiring factors data for everything NTSF produces that includes some portion shipped to the U.S., and the government maintains that the requirement extends to factors of production for affiliated fish farms. Id. “Thus, Commerce had requested information pertaining to NTSF Vinh Long’s factors of production, but NTSF had failed to timely provide the information in the form or manner requested.” Id. (citing Appx1026–1027). Catfish Farmers, in turn, note that NTSF affirmatively represented to Commerce that the questionnaire responses included all factors of production for both NTSF and Vinh Long. Case 20-104, ECF 52, at 10 Ct. Nos. 20-00104, 20-00105 Page 13 (quoting Appx89844 (“The attached response contains the factors of production for NTSF and NTSF Vinh Long.”)). They refer to the disconnect between (1) NTSF’s argument here that Commerce never asked for Vinh Long’s farming factors of production, (2) NTSF’s initial statement to the Department that its reporting included exactly those same factors of production, and (3) NTSF’s offer at verification to produce the data for those same factors because in fact they had not been reported. Id. at 11 (“It is unreasonable for NTSF to now argue that Commerce never asked it for NTSF Vinh Long’s farming factors of production when NTSF first told Commerce that its reporting included NTSF Vinh Long’s factors of production, and then told Commerce during the underlying proceeding that it failed to report them and attempted to untimely add them to the record.”). NTSF admits that some portion of Vinh Long’s farmed fish was used in producing frozen fish fillets, including both frozen fish fillets sold to the U.S. and frozen fish fillets shipped elsewhere. Case 20-104, ECF 56, at 5–7. By admitting that (1) a portion of Vinh Long’s farmed fish was used to produce frozen fish fillets exported to the United States and (2) the factors of production data did not include the factors Vinh Long used in producing those fish, NTSF admits that there was a gap in the record. NTSF thus admits that its questionnaire responses inaccurately stated that they included factors data for both NTSF and Vinh Long. At a minimum, therefore, the administrative Ct. Nos. 20-00104, 20-00105 Page 14 record permitted Commerce to conclude that NTSF withheld information and significantly impeded the proceeding by submitting misleading questionnaire responses and not seeking to correct them until the first day of verification. The Department’s decision to apply facts otherwise available to fill the gap in the record related to Vinh Long’s data is therefore supported by substantial evidence. 4 B. Commerce’s manner of applying facts otherwise available NTSF argues that even if Commerce permissibly decided to apply facts otherwise available, the Department still erred in how it did so. NTSF essentially argues that the Department identified a narrow gap in the record relating to a fraction of a single control NTSF makes two other arguments that are easily dispatched. First, it argues that the Department is “foreclosed” from applying facts otherwise available “based on the notion that necessary information concerning NTSF’s farming [factors of production] consumption was not available on the record.” ECF 40-1, at 22 (citing 19 U.S.C. § 1677e(a)(1)). This argument is odd, because Commerce did not rely on § 1677e(a)(1) to apply facts otherwise available. Second, the company challenges Commerce’s use of an adverse inference, arguing that the statutory prerequisite of a permissible application of facts otherwise available is absent here. See ECF 40-1, at 22–23. Because the court rejects NTSF’s challenge to the Department’s application of facts otherwise available, the company’s challenge to Commerce’s use of an adverse inference necessarily fails. 4 Ct. Nos. 20-00104, 20-00105 Page 15 number (i.e., a single type of frozen fish fillet) sold to the United States, so any application of facts otherwise available and adverse inference should have been limited to that narrow gap in the record. NTSF objects that Commerce instead acted more broadly by applying facts otherwise available and the adverse inference as part of the normal value calculation for every control number NTSF sold to the United States. Case 20-104, ECF 40-1, at 25–26 (citing Appx102796); see also id. at 28 (“[T]he AFA that Commerce applied did not fill the gap that Commerce identified”). NTSF contends that Commerce based its application of facts otherwise available and the adverse inference on what percentage of the overall fish harvested at the NTSF and Vinh Long farms came from Vinh Long’s fish farms, rather than the percentage of Vinh Long’s fish used in producing fish sold to the United States. Id. at 29. NTSF contends that all its factors of production for five of the six control numbers it reported were based on data from NTSF itself and that Commerce verified that fact. 5 Case 20-104, ECF 40-1, at 29. NTSF argues that the Department therefore discarded verified data rather than limiting its use of facts available, and by extension its use of an adverse inference, to the percentage of the single control Commerce’s verification report, however, makes clear that “[t]his report does not draw conclusions as to whether the reported information was successfully verified . . . .” Appx16986–16987 (emphasis in original). 5 Ct. Nos. 20-00104, 20-00105 Page 16 number for which Vinh Long’s farmed fish were relevant. Id. at 29–30. The government responds that NTSF admitted it did not track actual quantities of each farming factor of production on a control number–specific basis, instead using a system that took “the farming factors of production numerator (including those provided by NTSF Vinh Long) divided by the harvested whole live fish denominator (including those same fish).” Case 20-104, ECF 48, at 60 (citing Appx1027). NTSF acknowledges it did not track quantities on a control number–specific basis, Case 20-104, ECF 40-1, at 12– 13, but also notes that Commerce’s instructions directed respondents unable to report factor-of-production consumption on an actual basis to explain how they derived their estimated consumption on a control number basis, id. at 10–11 (citing Appx1125). The final determination explained why the agency rejected this argument by NTSF. Commerce characterized NTSF’s position as being that “Vinh Long’s tolled production overlaps with few [control numbers] sold by NTSF to the United States” and then found the argument “unavailing” because “whole fish are the starting point for all of the [control numbers], not just a subset of those reported in the U.S. sales listing.” Appx1027 (emphasis added). Crucially, [t]his construct starts with the farming [factors of production] numerator (including [factors of production] for fish provided by NTSF Vinh Ct. Nos. 20-00104, 20-00105 Page 17 Long) divided by the harvested whole live fish denominator (including those same fish). Thus, we disagree with NTSF that these [factors of production] can necessarily be specifically associated with specific U.S. [control numbers] at this point. Id. Commerce further found that NTSF’s argument “has no bearing on how the farming [factors of production] are initially constructed, as the initial information (i.e., the [factor of production] numerator and the whole fish denominator) is not limited to a single final product.” Id. NTSF does not respond to these concerns. But Commerce clearly found that the gap in the record—the omission of Vinh Long’s factors of production consumption data—pervaded all NTSF’s control numbers because Vinh Long’s data were, or should have been, included in the framework for the calculations used for those control numbers. As the government notes, NTSF used Vinh Long’s data in constructing its farming factors of production calculations and only later “applied a [control number–]specific standard consumption and then a variance to report its reported farming factors of production.” Case 20-104, ECF 48, at 60. Because NTSF included Vinh Long’s data in the initial calculation on which all its control number data were based, Commerce’s application of facts otherwise available with an adverse inference to all NTSF’s Ct. Nos. 20-00104, 20-00105 Page 18 control numbers that were based on those data is supported by substantial evidence. NTSF further argues that Commerce’s selection of partial facts available with an adverse inference for fish feed and fingerlings was not supported by substantial evidence because the Department (1) chose the highest consumption rate among all fingerling size-specific factors and assigned that to all the fingerling factors and (2) assigned the highest consumption rate among the three different feed types NTSF reported using. Case 20-104, ECF 40-1, at 30–31. NTSF contends that this was erroneous because NTSF reported using 16 sizes of fingerlings (based on their size when they were placed into the fish-farming ponds) and three types of fish feed (based on the feed’s protein content). Thus, NTSF argues that Commerce’s approach “bears no rational relationship to how NTSF reported its [factors of production] and is contradicted by the record.” Id. The problem with this argument is that it was not NTSF’s data that were missing from the record and as to which Commerce applied facts otherwise available and an adverse inference. Rather, it was Vinh Long’s data that were missing and as to which the Department applied facts otherwise available and the adverse inference: “[W]e based NTSF Vinh Long’s farming factors on the highest farming [factors of production] on the record for each farming factor category . . . .” Appx1027 (emphasis added); see also Appx17322 Ct. Nos. 20-00104, 20-00105 Page 19 (analysis memorandum for NTSF) (“We applied partial adverse facts available . . . with respect to NTSF Vinh Long’s farming [factors of production].”). While NTSF contends that “[t]here is no basis in the record to assume that each NTSF Vinh Long farm pond cycle consumed every size of fingerling at the highest possible consumption rates,” Case 20-104, ECF 40-1, at 33, it has not cited any evidence in the record showing that Vinh Long’s pond cycles did not do so or, for that matter, what Vinh Long did consume. Instead, NTSF cites its own consumption. But it cites nothing to establish that its own consumption is relevant to Vinh Long’s consumption. The lack of evidence from NTSF is the problem here. As the Federal Circuit has noted, the purpose of applying an adverse inference is to “ensure[ ] an uncooperative party does not obtain a more favorable result by failing to cooperate than if it had cooperated fully.” Ad Hoc Shrimp Trade Action Comm. v. United States, 802 F.3d 1339, 1361 (Fed. Cir. 2015) (cleaned up) (citing Statement of Administrative Action accompanying the Uruguay Round Agreements Act, H.R. Rep. No. 103316, vol. 1, at 870 (1994), reprinted in 1994 U.S.C.C.A.N. 4040, 4199). When, as here, there is no evidence from which Commerce can determine what the outcome would have been had the respondent cooperated fully, the Department’s use of the highest farming factor values available on the record is supported by substantial evidence. Otherwise, there Ct. Nos. 20-00104, 20-00105 Page 20 would be no way to know whether using a lower value might lead to a more favorable result than the respondent would have obtained by cooperating fully. C. Calculation of international movement expenses NTSF’s final two issues appear to be variants on a single question—whether Commerce erred by (1) rejecting NTSF’s “ministerial error” allegation relating to the inclusion of expenses NTSF reported in a particular database field called “USOTHTRU” and (2) including those same expenses in its calculation. In other words, NTSF contends that Commerce should not have included the USOTHTRU figures in its calculation of NTSF’s export price regardless of whether it was permissible for Commerce to reject the ministerial error allegation. NTSF explains that this issue arose because it reported miscellaneous U.S. delivery and warehousing expenses via two different methods, intended as alternatives to each other. The company says that it did this because it was concerned that Commerce might object to its preferred method. NTSF’s preferred method, used in USOTHTRU, reported expenses “on a per-unit basis . . . for all sales by dividing the sum of all [period-of-review] other transportation expenses by the total net weight of merchandise shipped during the [period of review].” Case 20-104, ECF 40-1, at 34 (citing Appx89666, Appx89683–89697); id. at 35 (describing the method as “total [period-of-review] U.S. other Ct. Nos. 20-00104, 20-00105 Page 21 transportation expenses divided by total [period-of-review] net weight in pounds”). NTSF contends that its preferred method was the most accurate way to allocate these expenses, but explains it also reported the same expenses using an “invoice-specific” allocation method in a separate database field called “USOTHTR2U” out of concern that Commerce might reject the method used in USOTHTRU. Id. (describing the alternative method as “U.S. other transportation expenses in U.S. dollars reported on each invoice from NTSF’s logistics provider divided by the net quantity reported for the same transaction in pounds”). Commerce’s preliminary determination stated that the Department deducted movement expenses reported in three fields, including USOTHTRU but not USOTHTR2U, 6 but the calculations showed that Commerce included the data from both USOTHTRU and USOTHTR2U. Id. at 36 (citing Appx97357 (preliminary determination) and Appx97377 (calculations)). NTSF contends this error was immaterial because Commerce preliminarily assigned the company an antidumping margin of zero. Id. (citing Appx16593). NTSF’s post-verification case brief addressed the shipping calculations and asked Commerce to replace the equation used in the preliminary determination For present purposes, the other two fields are unimportant. 6 Ct. Nos. 20-00104, 20-00105 Page 22 (which, as noted above, included both fields) with a revised version. Whether NTSF intended it or not is unclear, but the revised equation the company proposed continued to include both fields. 7 The brief did not object to the preliminary determination having included both USOTHTRU and USOTHTR2U in its equation and did not flag the use of both fields as an error. The calculations accompanying Commerce’s final determination likewise continued to include both fields. 8 The Department provided the parties with calculation data supporting the final determination on The preliminary determination used the following equation: 7 INTLMOVEU = (BHSV × (GROSS_WEIGHT ÷ QTYU)) + INTNFRU_REV + INTNFRU + USDUTYU + USOTHTRU + USOTHTR2U + USBROKU Appx16559 (emphasis added). NTSF recommended that Commerce instead use the following: INTLMOVEU = (329 ÷ container wgt) + INTNFRU_REV + INTNFRU + USDUTYU + USOTHTRU + USOTHTR2U + USBROKU Appx16983 (emphasis added) (stating both equations shown above and urging Commerce to use the second one). 8 The final determination used the following equation: INTLMOVEU = (BHSV × G2N) + INTNFRU_REV + INTNFRU + USDUTYU + USOTHTRU + USOTHTR2U + USBROKU Appx17324 (emphasis added). Ct. Nos. 20-00104, 20-00105 Page 23 April 24, 2020. Appx102795 et seq.; see also Case 20-104, ECF 23-1, at 34 (index to administrative record listing six documents dated April 24, 2020, relating to the final determination). Five days later, NTSF filed a “ministerial error allegation” asserting that the use of both fields in the calculations was erroneous because the two were meant to be alternative ways of reporting the same data on different bases, and so Commerce should have used only one or the other. The company also argued that because the Department verified the data in USOTHTR2U but not the other field, it should have used only the verified data in its final calculation. Appx17350–17353. A few weeks later, Commerce rejected the “ministerial error allegation” as untimely, finding that NTSF should have raised the issue in its case brief because it was discoverable earlier in the proceeding. Appx17341. The government and Catfish Farmers emphasize that not only did NTSF fail to object to Commerce’s use of both fields in its preliminary calculations, but in fact the company urged the Department to continue to use both fields by proposing a revision to the equation that included both. 1. Commerce’s rejection of NTSF’s “ministerial error” allegation Commerce’s regulations require parties to use their case briefs to call the Department’s attention to issues they consider significant. After the Department issues its preliminary determination, the parties have 50 Ct. Nos. 20-00104, 20-00105 Page 24 days to file case briefs that “must present all arguments that continue in the submitter’s view to be relevant to the Secretary’s determination or final results, including any arguments presented before the date of publication of the preliminary determination or preliminary results.” 19 C.F.R. § 351.309(c)(2). Both Commerce and reviewing courts normally find an argument not presented in a party’s case brief to be waived unless the argument could not have been raised in the case brief. “Generally, the ‘prescribed remedy’ for a party in disagreement with Commerce’s Preliminary Results is to file a case brief, and that case brief must present all arguments that continue in the submitter’s view to be relevant to Commerce’s final determination or final results.” Pakfood Pub. Co. v. United States, 724 F. Supp. 2d 1327, 1350 (CIT 2010) (cleaned up and emphasis in original). The Department’s regulations do recognize that in some cases, a mistake might first appear in the final determination, when it would be too late for a party to address the issue via the (already-filed) case brief. The regulations therefore provide that when Commerce notifies a party to the proceeding of the calculations used in connection with “a final determination or the final results of a review,” the party may submit comments about a “ministerial error” 9 within five days. “[M]inisterial error means an error in addition, subtraction, or other arithmetic function, clerical error resulting from inaccurate copying, duplication, or the like, and any 9 Ct. Nos. 20-00104, 20-00105 Page 25 19 C.F.R. § 351.224(c)(1)–(2). The regulation also provides, however, that “[c]omments concerning ministerial errors made in the preliminary results of a review should be included in a party’s case brief.” Id. § 351.224(c)(1). 10 NTSF’s case brief did not address Commerce’s use of both fields in its preliminary determination—the first time NTSF raised the issue was via a ministerial error allegation following the final determination. Undeterred, NTSF argues that there are three reasons why its failure to raise the issue in its case brief is not dispositive. First, NTSF contends that to the extent there was an error in the preliminary determination, it was irrelevant and immaterial because NTSF preliminarily received a zero antidumping margin: “There was no benefit, in terms of administrative efficiency, to require NTSF to raise this argument in its case brief other similar type of unintentional error which the Secretary considers ministerial.” 19 C.F.R. § 351.224(f) (italics in original). The regulation prescribes a different procedure for addressing a narrower category of “significant ministerial errors” in preliminary decisions; as to those, a party is to submit separate comments within five days rather than waiting to address the matter in its case brief. 19 C.F.R. § 351.224(c)(1). That alternate procedure is not at issue here—no party contends there was a “significant ministerial error.” 10 Ct. Nos. 20-00104, 20-00105 Page 26 where it had no reason to anticipate that the calculation error would have impacted the zero margin calculated in the Preliminary Results.” Case 20-104, ECF 40-1, at 45 n.15. NTSF cites a case in which this court found that a respondent was “not required to predict that Commerce would accept other parties’ arguments and change its decision” between the preliminary and final determinations. Id. at 42 (citing Qingdao Taifa Grp. Co. v. United States, 637 F. Supp. 2d 1231, 1237 (CIT 2009)). The court is not persuaded. In Qingdao, the preliminary determination did not address the issue in question; instead, Commerce first did so in its final determination. 637 F. Supp. 2d at 1236. Thus, the Qingdao court rightly reasoned that the respondent did not have a fair opportunity to challenge the application of facts otherwise available with an adverse inference. Id. at 1237. Here, in contrast, Commerce’s preliminary determination used both fields—USOTHTRU and USOTHTR2U. By definition, that (preliminary) determination was subject to change in the final determination. If the Department increased NTSF’s antidumping margin (as it did in the final determination), then the use of both fields could well have become relevant—and, unlike in Qingdao, that was clear when the preliminary determination was issued. Therefore, NTSF did have a fair opportunity to argue that (1) Commerce should not depart from its preliminary Ct. Nos. 20-00104, 20-00105 Page 27 determination but (2) if it did so depart, it should correct the calculation error. Separately, NTSF argues that despite Commerce using both fields in both the preliminary and final determinations, the Department made a different error each time such that the error in the final determination was a “new” error that the company did not have the chance to address earlier. NTSF argues that because the text of the preliminary determination referred only to USOTHTRU, Commerce meant to include only that field in its calculation even though the calculation included USOTHTR2U as well. NTSF then claims that because at verification Commerce examined the expenses reported in USOTHTR2U and not USOTHTRU, it was “apparent” the Department intended to use only USOTHTR2U in the final determination. Case 20-104, ECF 40-1, at 44 & n.14. 11 As Catfish Farmers argue, however, to the extent that there was an error, the same error appeared in both the preliminary and final determinations: Commerce used both fields in its constructed export price NTSF acknowledges that the final determination is silent about which field Commerce intended to use but asserts that “it is obvious” the Department intended to use USOTHTR2U based on how it conducted verification. Id. at 44 n.14. The verification report states, however, that it “does not make findings or conclusions regarding how the facts obtained at verification will ultimately be treated in Commerce’s determinations.” Appx16987 (emphasis in original). 11 Ct. Nos. 20-00104, 20-00105 Page 28 calculation in both decisions. Case 20-104, ECF 52, at 18 (citing Appx97377 and Appx102819). NTSF’s argument about the Department’s “intent” is unconvincing because if NTSF’s case brief had addressed why it was important to avoid using both fields, Commerce might well have been prompted to avoid doing so in its final determination. To that end, the government and Catfish Farmers both emphasize that not only did NTSF not seek to correct the error in the preliminary determination, but in fact NTSF urged Commerce to repeat or retain the error by recommending it adopt an equation that included both fields. 12 As Catfish Farmers note, NTSF’s case brief suggested that the use of both fields was correct. Finally, NTSF essentially argues that raising the error in its case brief was optional. See Case 20-104, ECF 40-1, at 47 (“[T]he plain language of Commerce’s ministerial error regulation does not require a party to present any calculation error present in the See Case 20-104, ECF 48, at 64 (government) (“Notably, NTSF did not recommend removing ‘USOTHTR2U’ from the program, because it used that field in its proposed replacement formula. In other words, NTSF used ‘USOTHTR2U’ in its own proposed calculation.”) (emphasis in original) (citing Appx16983); Case 20-104, ECF 52, at 19 (Catfish Farmers) (“. . . NTSF included in its case brief both USOTHTRU and USOTHTR2U expenses fields in its proposed revision of Commerce’s preliminary INTLMOVEU calculation.”) (also citing Appx16983). 12 Ct. Nos. 20-00104, 20-00105 Page 29 Preliminary Results in its case brief.”); id. at 46 (citing 19 C.F.R. § 351.224(c)(1) with a parenthetical reading, “stating that a party ‘should’ raise a ministerial error made in the preliminary results in its case brief”). NTSF therefore claims that Commerce “lacked any legal authority to reject” NTSF’s allegation of error. The Federal Circuit, however, treats the regulations as mandatory: Commerce discloses any calculations made in the preliminary results to interested parties, and interested parties must point out any ministerial errors in their case briefs. . . . Commerce’s refusal to make a ministerial correction is not reversible error when the alleged mistake was discoverable during earlier proceedings but was not pointed out to Commerce during the time period specified by regulation. QVD Food Co. v. United States, 658 F.3d 1318, 1328 (Fed. Cir. 2011) (cleaned up and emphasis added); see also Stanley Works (Langfang) Fastening Sys. Co. v. United States, 964 F. Supp. 2d 1311, 1341 (CIT 2013) (same). 13 The regulation’s language that a party “may NTSF seeks to avoid this clear language by arguing that Stanley Works addressed a matter in which the preliminary and final determinations contained the same error, which NTSF contends is not the case here. Case 20-104, ECF 56, at 24–25. For the reasons explained above, NTSF’s 13 Ct. Nos. 20-00104, 20-00105 Page 30 submit comments” does not mean that compliance with the regulation and its time limits is optional—rather, it simply means that parties have a choice whether to submit ministerial error allegations at all. In a similar vein, NTSF argues, in effect, that the Department’s obligation to calculate NTSF’s margin as accurately as possible overrides all other considerations. See Case 20-104, ECF 56, at 23 (stating that because Commerce must set margins as accurately as possible, NTSF’s recommendation of an equation using both fields “does not excuse Commerce from its legal obligation to accurately calculate NTSF’s dumping margins”). As an administrative agency has discretion in setting and enforcing deadlines, courts apply an abuse-ofdiscretion standard in examining whether accuracy and fairness may require a departure from those deadlines. See, e.g., Dongtai Peak Honey Indus. Co. v. United States, 777 F.3d 1343, 1351 (Fed. Cir. 2015) (recognizing that absent constitutional constraints or “extremely compelling circumstances,” courts defer to agency judgment on development of administrative records and that Commerce must be permitted to enforce its deadlines). Here, the court finds Commerce did not abuse its discretion because (1) NTSF never asserted that using argument is unpersuasive, and the court therefore follows the Federal Circuit’s instruction. Ct. Nos. 20-00104, 20-00105 Page 31 both fields in the calculation was erroneous until after the proceeding was over and NTSF was unhappy with the result and (2) NTSF’s own case brief urged Commerce to use both fields in its calculation. Under those circumstances, it was not an abuse of discretion for the Department to reject a party’s attempt to change course only after the final determination was not to its liking. That NTSF’s case brief urged Commerce to use both fields is a “compelling circumstance” in favor of upholding the decision. Thus, because NTSF could have challenged the use of both fields in its case brief and did not do so—and instead repeated the use of both fields in its recommended equation—Commerce did not abuse its discretion in rejecting the company’s “ministerial error” allegation as untimely. 2. NTSF’s “substantial evidence” theory NTSF also argues that the administrative record “clearly demonstrates” that the two fields “reflect the same expenses reported on different bases” and that therefore Commerce’s decision to use both fields was not supported by substantial evidence. Case 20-104, ECF 40-1, at 50. For the reasons stated above, the court finds that the company could and should have addressed this issue in its case brief before the agency. Because NTSF failed to do so, it has not exhausted its administrative remedies and therefore has not preserved the issue for judicial review. Ct. Nos. 20-00104, 20-00105 Page 32 Even if the court were to agree with NTSF that the final determination contained a new error that could not have been addressed in the case brief, however, the court would find it significant that the company’s own case brief urged Commerce to use both fields in the final equation. That NTSF recommended the Department revise its equation proves that NTSF considered what fields should be included and did not consider the use of both fields problematic. While NTSF now contends that its inclusion of both fields in its brief was “inadvertent” and unimportant because the alleged error “had no impact on” the company’s preliminary margin, its own actions belie that argument: Because Commerce preliminarily assigned NTSF a zero margin, presumably nothing in the equation would have mattered going forward. Yet NTSF proposed other changes, so clearly NTSF believed changes were needed in case the final determination differed from the preliminary results. * * * For the foregoing reasons, the court denies NTSF’s motion for judgment on the agency record and grants judgment on the agency record in favor of the government and Catfish Farmers. See USCIT R. 56.2(b) (authorizing the court to enter judgment for a party opposing a motion for judgment on the agency record, “notwithstanding the absence of a cross-motion”). A separate judgment will enter. See USCIT R. 58(a). Ct. Nos. 20-00104, 20-00105 Page 33 II. Catfish Farmers’ case (No. 20-105) In its challenge to the antidumping rate Commerce assigned to NTSF’s imports, Catfish Farmers raise four main issues: (1) Commerce’s selection of India as the primary surrogate country, Case 20-105, ECF 32, at 15–41; (2) the accuracy of NTSF’s production inputs, id. at 42–49; (3) NTSF’s reporting of its factors of production, id. at 49–55; and (4) whether NTSF’s data overstated the moisture content in its fish. Id. at 55– 61. A. Primary surrogate country In selecting surrogate countries to value the factors of production for imports from non-market economies, the statute directs Commerce to use, “to the extent possible,” market economy countries that “are—(A) at a level of economic development comparable to that of the nonmarket economy country, and (B) significant producers of comparable merchandise.” 19 U.S.C. § 1677b(c)(4). Once the Department identifies such a country or countries, the statute charges it with valuating the factors of production “based on the best available information.” Id. § 1677b(c)(1). As this statutory scheme leaves many questions unanswered, Commerce has adopted a policy for surrogate country selection. Import Administration Policy Bulletin 04.1, Non-Market Economy Surrogate Ct. Nos. 20-00104, 20-00105 Page 34 Country Selection Process (Mar. 1, 2004) (“Bulletin”). 14 That policy, which neither Catfish Farmers nor NTSF challenges, has four steps. First, Commerce identifies a list of countries that are at “a comparable level of economic development” as the non-market economy country using “per capita gross national income” information from the World Bank. Id. at 2. Next, the Department determines whether any of those countries are producers of comparable merchandise. Id. Third, Commerce determines whether any of the countries that satisfy the first two criteria are “significant” producers of that comparable merchandise. Id. at 3. Finally, and most importantly, “the country with the best factors [of production] data is selected as the primary surrogate country.” Id. at 4. Critically, data quality is so important that it can justify selecting a country that did not otherwise survive steps one through three above: “After all, a country that perfectly meets the requirements of economic comparability and significant producer is not of much use as a primary surrogate if crucial factor price data from that country are inadequate or unavailable.” Id. at 4. Catfish Farmers challenge Commerce’s selection of India as the primary surrogate country for essentially two reasons. First, they contend India should have 14 http://enforcement.trade.gov/policy/bull04-1.html. Ct. Nos. 20-00104, 20-00105 Page 35 been disqualified as a surrogate country in the first instance because it was not a significant producer of comparable fish. Second, they argue that even if India were eligible, Commerce should have selected Indonesia instead because (a) the latter is at a comparable level of economic development to Vietnam and (b) the Indonesian factors of production data are superior to the corresponding Indian data. 1. India’s eligibility Before Commerce, Catfish Farmers contended that India is not a “significant producer[ ] of comparable merchandise,” 19 U.S.C. § 1677b(c)(4), because (1) most of its pangasius production does not consist of frozen fish fillets, (2) its production is of a different “grade” than the frozen fillets exported from Vietnam, and (3) Indonesia is a much larger exporter of frozen fish fillets than India. Appx1018 (Commerce summarizing Catfish Farmers’ arguments). As to the first argument, the Department emphasized that neither the statute nor the regulations define what constitutes “comparable merchandise.” Commerce therefore looked to its treatment of the issue in prior administrative reviews and found that it had examined “which countries are significant producers of frozen fish fillets,” not just which countries produce pangasius fillets. Appx1019. “Although ‘frozen fish fillets’ represent a broader category than in-scope pangasius frozen fish fillets, the category is nonetheless comparable because it allows for the selection of Ct. Nos. 20-00104, 20-00105 Page 36 surrogate financial ratios from producers of similar products with similar capital structures.” Id. The court understands Commerce to have reasoned that because the frozen fish fillet production process is similar regardless of the type of fish, using the broader category of “frozen fish fillets” generally allows for more data sources. Catfish Farmers’ briefing does not address this point, focusing instead only on whether India exports frozen pangasius fillets. As to the second argument, the Department characterized Catfish Farmers’ argument as “attempt[ing] to introduce additional criteria into the mix.” Id. (referring to Catfish Farmers’ arguments about “white and flaky” pangasius fillets and their “reliance on which countries are net exporters”). Commerce rejected this argument because the relevant antidumping order “does not distinguish among grades of subject merchandise.” Id. Although the order does discuss various forms of subject merchandise, grades are not among the considerations mentioned. Id. Commerce also correctly noted that Policy Bulletin 04.1 does not require that a surrogate country be a net exporter and instead only requires that the country could be a net exporter. Appx1019–1020; see Bulletin, above, at 3 (“In another case, there may not be adequate data available from major producing countries. In such a case, ‘significant producer’ could mean a country that is a net exporter, even though the Ct. Nos. 20-00104, 20-00105 Page 37 selected surrogate country may not be one of the world’s top producers.”). Rather than responding to the Department’s reasoning, Catfish Farmers insist that only one grade of frozen pangasius fillets can constitute “comparable” merchandise and that it is therefore irrelevant whether Commerce considered grade, color, and texture “in the buildup of normal values.” Case 20-105, ECF 32, at 23–24 (arguing that the Department failed to address whether the Indian fish fillets are sufficiently comparable). Catfish Farmers contend that Indonesia “is a better surrogate for Vietnam” than India because “only Indonesia is a significant producer of an equivalent quality frozen fish fillet.” Id. at 24 (emphasis added). Catfish Farmers fail to address Commerce’s points that the antidumping order does not distinguish between grades of subject merchandise, that Policy Bulletin 04.1 does not require that the surrogate country be a net exporter, and that the use of the broader category of “frozen fish fillets” rather than “frozen pangasius fillets” allows for a more useful selection of data from producers like pangasius producers in all ways other than the particular fish. Contrary to Catfish Farmers’ argument, the relevant question here is not whether Indonesia may be a more significant producer of comparable merchandise. The question is instead whether India is a significant producer. Cf. Bulletin, above, at 2 (“The statute does Ct. Nos. 20-00104, 20-00105 Page 38 not require that the Department use a surrogate country that is . . . the most significant producer of comparable merchandise”) (emphasis in original). Therefore, Commerce’s finding that India is a “significant producer of comparable merchandise” is supported by substantial evidence. 2. Indonesia versus India a. Indonesia’s economic development level Catfish Farmers observe that the Department selected Indonesia as the primary surrogate country in the last eight reviews, including four in which Commerce found Indonesia to be at a comparable or the same level of economic development as Vietnam. See Case 20-105, ECF 32, at 18–19. Yet in this review, when Commerce compiled its list of six countries at levels of economic development comparable to Vietnam using per capita gross national income information from the World Bank, Indonesia was not on the list, even though Catfish Farmers argued for its inclusion. Catfish Farmers also observe that the Department failed to discuss “Indonesia’s per-capita [gross national income]” and whether Indonesia qualifies for the list under the same standard. See id. at 18. Commerce’s preliminary determination did not discuss any country’s economic comparability beyond noting the six countries identified on its “potential surrogate country” list. Appx16539–16540. Its final Ct. Nos. 20-00104, 20-00105 Page 39 determination did note that Catfish Farmers had argued that Indonesia was at a comparable level of economic development “even though it was not on the Surrogate Country List” (with no explanation of how mere inclusion on that list could somehow be determinative), but Commerce’s response to that argument was limited to finding that “[t]he band of countries that Commerce selected in this review, in absolute terms, is a reasonable range of countries given the entire worldwide range of GNIs.” Appx1018. Commerce then simply found that “Indonesia is not at the same level of economic development as Vietnam,” id. (emphasis added), without acknowledging that the finding potentially misapplies the statutory standard that Commerce use a surrogate country that is “at a level of economic development comparable to that of the nonmarket economy country,” 19 U.S.C. § 1677b(c)(4) (emphasis added). The government ignores Catfish Farmers’ argument and instead contends that substantial evidence supported Commerce’s determination that India was at a level of economic development comparable to Vietnam for purposes of the statute. See Case 20-105, ECF 42, at 17–20. But that finding is irrelevant to whether, as Catfish Farmers argued before the Department, Indonesia is also economically comparable to Vietnam, and it leaves open the possibility that Commerce may have applied too strict a standard in rejecting Indonesia for not being at “the same level of economic development.” As a result, the court will Ct. Nos. 20-00104, 20-00105 Page 40 remand for Commerce to explain whether Indonesia is comparable using the same World Bank gross national income data used to identify India and the five other countries. b. Indonesian versus Indian data Commerce’s final determination did not dispute that Indonesia is a significant producer of fish comparable to those produced by Vietnam. And as discussed above, the Department failed to address whether Indonesia is also at a level of economic development comparable to Vietnam. Commerce acknowledged, however, that under its Policy Bulletin 04.1, where more than one country satisfies the “economically comparable” and “significant producer” criteria, the Department is to select the country “with the best factors data” as the surrogate country. Appx1021. “Commerce must weigh the available information with respect to each input value and make a product-specific and case-specific decision as to what constitutes the best available [surrogate value] for each input.” Appx1022. Accordingly—perhaps anticipating that its discussion of Indonesia’s economic comparability to Vietnam would not withstand scrutiny—the Department went ahead and compared the Indian data to the Indonesian data. See Appx1021–1025. In so doing, it concluded that “India offers the best available [surrogate value] Ct. Nos. 20-00104, 20-00105 Page 41 information.” Appx1021. 15 Catfish Farmers challenge this determination on several grounds. i. Circular reasoning Catfish Farmers correctly argue that Commerce impermissibly employed circular reasoning. See Case 20-105, ECF 32, at 26. Specifically, the Department found the Indian data were superior in part because “the Indonesian information is not from the primary surrogate country which we have selected in this case, India.” Appx1022 (quoting Appx16541) (citations omitted by Commerce). This means that Commerce—absurdly—found the Indian data superior because they were from India. That alone warrants remand for the Department to reconsider its determination. ii. “Broad market averages” Catfish Farmers argue that the Indian data cannot be considered the “best available information” because they do not represent a “broad market average.” See generally Case 20-105, ECF 32, at 26–36. (a). In general The central thrust of Catfish Farmers’ argument is that while pangasius in India are farmed in over 300 As noted below, however, Commerce did not find Indian data superior in every respect. 15 Ct. Nos. 20-00104, 20-00105 Page 42 villages, the Fishing Chimes 16 study on which Commerce relied “only involved ‘54 farmers from 46 villages’ within only two of 13 districts of a single state, Andhra Pradesh.” Case 20-105, ECF 32, at 27–28 (citing Appx13785–13791). Catfish Farmers argue that Commerce’s finding that the data were acceptable because Andhra Pradesh is the largest pangasius-producing region in India “accounting for approximately 80 percent of pangasius production during the” period of review “is not true.” Id. at 28 (quoting Appx1022– 1023). As a general matter, a “broad market average” need not include an entire country if the relevant production comes from a smaller area. Catfish Farmers, however, are correct that other evidence in the administrative record casts doubt on Commerce’s reasoning. Evidence shows that Andhra Pradesh produced about 80 percent of India’s output in 2017, but it also shows that state’s share fell to about 60 percent in 2018 because farmers switched to producing shrimp and because farmers in four other states (Bihar, Tripura, Uttar Pradesh, and West Bengal) began farming more Fishing Chimes is “a monthly journal devoted to the development of fisheries and aquaculture.” Appx13750. The journal is produced by an Indian publisher. Appx13754. The parties’ discussion of “the Fishing Chimes study” refers to an article appearing in the July 2019 issue. See Appx13785–13791 (Rao, Avanigadda, and Godumala, Current Status of Pangasius (Pangasianodon hypophthalmus) Farming in India, Fishing Chimes, July 2019, at 34–40). 16 Ct. Nos. 20-00104, 20-00105 Page 43 pangasius. Appx15534. The period of review here was August 1, 2017, to July 31, 2018. It is unclear how much the data reported on Appx15534 overlap with the period of review because that page merely says “in 2018.” Nor does the source say how much fish other states produced. Catfish Farmers further object that the Fishing Chimes study on which Commerce relied reflects only two of 13 districts within Andhra Pradesh. Fishing Chimes states that pangasius is produced in 15 states but that because information is “scarce,” Appx13785, the authors surveyed 54 farmers in two districts within Andhra Pradesh. Yet Fishing Chimes also states that between 2010 and 2012 pangasius farming “rapidly expanded to other states, viz., Telangana, Tamil Nadu, Kerala, Maharashtra, Goa, Himachal Pradesh[,] and Rajasthan.” Id. Thus, there is evidence in the record that pangasius is farmed in at least 11 Indian states. Id. Fishing Chimes states that the total is 15. Id. Fishing Chimes also estimates that “pangasius is being farmed currently in more than 300 villages in West Godavari and Krishna districts surrounding Kolleru lake area.” Appx13786. 17 But in a somewhat The cited page includes a screenshot from what appears to be Google Maps. After examining a map of India, the court takes judicial notice that the area shown on the map falls entirely within Andhra Pradesh and that the fishing 17 Ct. Nos. 20-00104, 20-00105 Page 44 self-contradictory statement, the analysis then examines those two districts and states that “[o]ut of the 300 villages that the study covered, 46 of them are in these two districts.” Id. This raises a serious question: If most of Andhra Pradesh’s fish producers are not in those two districts, how can a study that relies on data from only those two districts represent a broad market average, absent data (which no party has cited) showing that those districts produced far more fish than anywhere else? Fishing Chimes also contains a table showing the amount of production, and the number of hectares devoted to pangasius, in each Indian state as of January 2019. While the data in the table are not relevant here because they fall outside the period of review, the text on that page states that Andhra Pradesh’s contribution to Indian aquaculture was reduced from 80 percent in 2017 to 58 percent in 2018, though it also notes that data about pangasius farming in other states before 2018 is extremely limited. Appx13788. It appears, therefore, that that the 80 percent figure could be an overstatement. While the government is correct statutory requirement that a “broad must reflect a particular percentage the government is also correct that that there is no market average” of a market, and Andhra Pradesh villages mentioned on page Appx13786 are all within that one state. Ct. Nos. 20-00104, 20-00105 Page 45 remained a significant producer of pangasius even after its share of the market decreased, Case 20-105, ECF 42, at 29, Commerce placed considerable emphasis on the 80-percent figure and did not opine on what market share the Department considers “significant.” Therefore, because (1) Commerce emphasized that Andhra Pradesh accounted for “approximately 80 percent” of production during the period of review, (2) evidence in the administrative record contradicts that finding, and (3) Commerce failed to address that contradictory evidence, the court finds that Commerce’s findings about a “broad market average” are not supported by substantial evidence. (b). Fingerlings For similar reasons, the court agrees with Catfish Farmers that Commerce’s finding that the Fishing Chimes fingerlings data represent a “broad market average” is not supported by substantial evidence. Commerce noted that “Fishing Chimes indicates that its fingerling data were from commercial nurseries in two locales, which collectively supplied fingerlings to 94.5 percent and 96 percent of the pangasius farms in 2017 and 2018, respectively.” Appx1023. However, the paragraph from which Commerce drew that data also refers to “interviews with the farmers and stakeholders” about where they obtained their fingerlings, Appx13789, thus suggesting that the supply data referred to the specific subset of farms Fishing Chimes surveyed—i.e., 54 farms in 46 villages in two districts within Andhra Pradesh. For that reason, Commerce’s Ct. Nos. 20-00104, 20-00105 Page 46 reliance on the Fishing Chimes fingerling data suffers from the same problem as its overall reliance on that study. (c). Fish feed Commerce found that the Fishing Chimes analysis “containing the Indian fish feed data shows pricing collected by researchers from most of the pangasius feed producers in India, and that this information is corroborated by invoices from another major Indian fish feed supplier.” Appx1023 (footnote omitted). Catfish Farmers argue that this finding is not supported by the record because Fishing Chimes lists “[s]ome of the feed brands that the farmers used” and then mentions “other non-branded feeds,” Appx13790, which Catfish Farmers argue implies “that there are branded feeds which were not reported,” Case 20-105, ECF 32, at 32. Catfish Farmers do not try to explain why it matters whether the specific brand was reported if the producers reported what they paid for fish feed. Fishing Chimes states that “[p]rices of the feeds were collected from the farmers through a questionnaire and then their averages were calculated.” Appx13790. Catfish Farmers also object that Commerce’s finding of corroboration is unsupported by substantial evidence because the supplier whose data Commerce cited, Godrej Agrovet, “is one of the suppliers named in the Fishing Chimes study. Invoices from one of the suppliers included in the study do not provide an independent basis for evaluating the quality of that study.” Ct. Nos. 20-00104, 20-00105 Page 47 Case 20-105, ECF 32, at 32. The list of fish feed brands in Fishing Chimes mentions “Godrej.” Appx13790. The court need not try to ascertain whether that refers to Godrej Agrovet because both the government and NTSF fail to respond to Catfish Farmers’ argument on this point in their briefing—the government simply parrots what Commerce said with no further analysis, Case 20-105, ECF 42, at 31, and NTSF generally joins in Commerce’s analysis, Case 20-105, ECF 45, at 2. The court will therefore remand so that Commerce can address the corroboration issue. (d). Whole fish Commerce’s sole finding about the Indian data as to the whole fish input was as follows: [W]e find that the Indian data for this input are in fact a broad market average, for the reasons discussed above. Finally, as NTSF observes, the Fishing Chimes data are corroborated by three other sources. Taken together, these factors indicate that the Fishing Chimes data are themselves representative of a broad market average. Appx1024 (footnote omitted). “For the reasons discussed above” apparently refers to Commerce’s general discussion of whether the Fishing Chimes study overall represents a broad market average. Therefore, the court must conclude that finding is not supported by substantial evidence. It is also unclear how “corroboration” has anything to do with whether data Ct. Nos. 20-00104, 20-00105 Page 48 represent a broad market average, as something may be “corroborated” yet still represent an exceptionally narrow part of the overall market. Because this court is remanding the determination that forms the only stated basis for Commerce’s finding on this issue, the court must remand this finding as well. iii. Wage rates/labor data Catfish Farmers object to Commerce’s valuation of NTSF’s labor inputs because the Department “relied on a 2006 Indian labor rate from the International Labor Organization’s ILOSTAT statistics database.” Case 20-105, ECF 32, at 36 (citing Appx16570, Appx16575, Appx102929). Policy Bulletin 04.1 places significance on whether data are “contemporaneous” with the period of review (here, August 1, 2017, to July 31, 2018), and Commerce cited that same consideration in its preliminary determination. See Appx16540. But the administrative record here proves the Department ignored contemporaneity in valuing the labor factor of production: “We valued labor using 2006 Indian ILOSTAT data. . . . Because this value is not contemporaneous, we inflated it.” Appx16570 (emphasis added). Nothing in the record explains why the Department found data from 2006—eleven years before the period of review— to be more acceptable than 2017–2018 data. In its briefing before this court, the government’s only argument on the issue is its irrelevant contention that Commerce chose the Indian data because India was Ct. Nos. 20-00104, 20-00105 Page 49 the primary surrogate country. Case 20-105, ECF 42, at 35 (citing Appx1022). The final determination is also silent about labor aside from an unsupported statement that “India provides the best data and information with which to value [factors of production], such as direct materials, labor, energy, and financial ratios.” Appx1025. The preliminary determination contained a similar statement and concluded that “the data from India . . . are specific to the main inputs, are tax- and duty-exclusive, represent a broad market average, and are contemporaneous and useable.” Appx16540. That sentence is not merely “unsupported” by substantial evidence—as to the labor data, the record flat-out contradicts it. Because Policy Bulletin 04.1 requires Commerce to consider contemporaneity as part of its analysis, the Department cannot simply select data from 11 years before the period of review without explaining the reason for disregarding the temporal disconnect, and it certainly cannot make false statements about contemporaneity as it did here. Commerce’s unexplained decision to use 2006 Indian data to value labor inputs is thus not supported by substantial evidence and the court will therefore remand as to that issue as well. iv. Byproducts Catfish Farmers note that Commerce elected to rely on Indonesian data to value fish byproducts Ct. Nos. 20-00104, 20-00105 Page 50 because all but one of the Indian values were lacking and the Indonesian values were superior. Case 20-105, ECF 32, at 37. Catfish Farmers obviously do not object to Commerce’s use of Indonesian data, but they contend that it “highlight[s] the importance of a proper selection [of] a primary surrogate that allows for the valuation of most or all factors of production using the same surrogate country.” Id. at 39. The relevant statutory language, however, specifically allows for the use of multiple countries if Commerce deems it appropriate. See 19 U.S.C. § 1677b(c)(1) (“[T]he valuation of the factors of production shall be based on the best available information regarding the values of such factors in a market economy country or countries considered to be appropriate by [Commerce].”) (emphasis added); id. § 1677b(c)(4) (“[Commerce], in valuing factors of production under paragraph (1), shall utilize, to the extent possible, the prices or costs of factors of production in one or more market economy countries . . . .”) (emphasis added). While it is well-established that Commerce has a “regulatory preference” for using one primary surrogate country, that preference must yield to the statutory mandate: [B]ecause the statute requires Commerce to compare the chosen data set with other data sets on the record and thereby determine what is the best available information, the regulatory preference cannot suffice as adequate reasoning if it Ct. Nos. 20-00104, 20-00105 Page 51 is the only factor that Commerce considers. The preference for using data from a single country might support a choice between data sets that, upon a fair comparison, are otherwise seen to be fairly equal . . . . Peer Bearing Co.–Changshan v. United States, 752 F. Supp. 2d 1353, 1373 (CIT 2011) (cleaned up and emphasis added). As a result, Commerce’s use of factors of production data from both Indonesia and India was not in and of itself impermissible, as Catfish Farms suggest. v. Financial statements/financial ratios Catfish Farmers argue that “Commerce relied on a single[ ] Indian financial statement, rather than two[ ] viable Indonesian financial statements, stating simply that India is the primary surrogate country and providing no explanation for disregarding its preference for multiple financial statements.” Case 20-105, ECF 32, at 40. The government responds that Commerce stated that it departed from its two-statement preference because it “identified problems with both the quality (i.e., non-contemporaneous data) and the completeness of” one of the Indian financial statements and therefore reasonably decided to use only the contemporaneous Indian financial data. Case 20-105, ECF 42, at 38. In response to Catfish Farmers’ contention that Ct. Nos. 20-00104, 20-00105 Page 52 Commerce could have satisfied its regulatory preference for using multiple financial statements by using the two Indonesian statements, the government noted that Commerce also has a regulatory preference for valuing all factors of production in a single country. Id. Thus, here there were two conflicting regulatory preferences—the preference for using multiple financial statements and the preference for a single surrogate country. It is not this court’s role to balance those preferences. Commerce explained why it considered one Indian company’s financial statement reliable and why it found the Indonesian statements inadequate, and it then chose to give priority to the single-country preference over the two-statement preference. While Commerce explained its use of the tiebreaker incorrectly, 18 it is clear enough what Commerce meant. * * * In sum, while Commerce’s determinations that India is economically comparable to Vietnam and is a significant producer of comparable merchandise are supported by substantial evidence, the court must Rather than referring to the “primary surrogate country” for whose selection the financial ratio data were a part, Commerce should have said, “We have relied on [the Indian] financial statements to compute the surrogate financial ratios in this review in view of our regulatory preference to value all factors of production in a single surrogate country where possible.” 18 Ct. Nos. 20-00104, 20-00105 Page 53 remand for the Department to reconsider (1) whether Indonesia is also economically comparable to Vietnam, and (2) whether, as discussed above, the Indian factors of production data are the best available information as compared to the competing Indonesian data submitted by Catfish Farmers. B. NTSF’s production inputs NTSF reported that its production inputs—whole live fish, fish feed, and other inputs used to produce the finished frozen fish fillets—weighed less than the finished outputs NTSF obtained from those inputs. Appx94001. Commerce issued a supplemental questionnaire asking NTSF for a breakdown of the total whole fish input and the various outputs. Appx12068. NTSF provided one. Appx94001. Commerce conducted verification and weighed the output products immediately after each stage of the production process, as they were generated, and found that the sum of the weight of the outputs, up to the stage following trimming, was greater than the weight of the whole live fish at the beginning of the process. Appx1012. It also examined whether the process of washing fish and then freezing them resulted in further weight gain and noted that it was impossible to fully replicate NTSF’s production process due to time constraints—normally, NTSF’s fish fillets sit in washing bins for several hours before being placed in the freezing machines. Id. Ct. Nos. 20-00104, 20-00105 Page 54 Similarly, the Department weighed the byproducts generated during verification but noted it would have been too disruptive to the production process to replicate NTSF’s entire procedure. Appx1012–1013. It found that even using the abbreviated procedures employed at verification, the byproducts gained some water weight, and therefore concluded it was reasonable to assume that the byproducts “would have similarly gained additional water weight had these by-products followed NTSF’s normal procedures.” Appx1013. Catfish Farmers object that Commerce’s testing at verification did not match up with NTSF’s reported data because the amount of water weight gain was less than NTSF had reported in its questionnaire responses. The Department nevertheless concluded that the discrepancy “could reasonably be accounted for” by (1) the fact that the fillets Commerce tested at verification “did not sit in chilled water for several hours, as is typical,” and (2) the fact that Commerce’s tests did not exactly match the way NTSF normally weighs byproducts, “i.e., after being transported by water to the by-products staging area.” Id. “Finally, in past verifications, we have seen the water output weights exceed the input weight by very similar amounts as NTSF’s experience here. As a consequence, we find that NTSF’s experience and yield reporting is consistent with our prior findings.” Id. Catfish Farmers now argue, citing Commerce’s decision in the 14th administrative review of the same Ct. Nos. 20-00104, 20-00105 Page 55 antidumping order at issue here, that “it is a ‘mathematical impossibility’ for the output value to be ‘much higher than the input’ value,” although Catfish Farmers concede that the discrepancy here is smaller. Case 20-105, ECF 32, at 43 (citing Certain Frozen Fish Fillets from the Socialist Republic of Vietnam: Issues and Decision Memorandum for the Final Results of the Fourteenth Antidumping Duty Administrative Review: 2016–2017, at 33). 19 The government and NTSF respond that Commerce reasonably found that fish fillets and byproducts gain water weight even if they are not directly submerged in water and that it was therefore reasonable to find that the weight gain seen here was explainable. NTSF notes that “the input in this review is a live animal and each stage of processing results in increased surface areas for water to cling and absorb on the various parts of the fish,” Case 20-105, ECF 45, at 8, and cites data from the verification report it contends “proves that the output can exceed the input,” id. (using data from Appx102459). After Catfish Farmers filed their opening brief here, the court remanded the finding they cite. Hung Vuong, 483 F. Supp. 3d at 1365. The administrative record in that case included the respondent’s explanation for the discrepancy and Commerce had failed to address that explanation. Id. at 1366. Commerce dropped the issue altogether after the remand. See Hung Vuong Corp. v. United States, Ct. No. 19-00055, Slip Op. 21-142, at 5–6, 13–14, 2021 WL 4772962, at **2, 5 (CIT Oct. 12, 2021). 19 Ct. Nos. 20-00104, 20-00105 Page 56 The government, for its part, argues that the important points are that (1) Commerce confirmed that the outputs weighed somewhat more than the whole live fish such that the difference could be explained, especially when the Department stated that in prior verifications it had seen output weights exceed input weight by amounts similar to those reported by NTSF; and (2) Commerce emphasized that the difference could be explained by the verification testing procedures not perfectly replicating NTSF’s normal production procedures. Case 20-105, ECF 42, at 40–42. The court concludes that substantial evidence in the administrative record supports Commerce’s conclusion. As noted above, the question is not whether the court would have reached the same conclusion on the same record or even whether the court agrees with the agency’s conclusion—the question is whether the agency has adequately addressed the record, including any evidence that detracts from its conclusion. Here, unlike in the 14th administrative review cited by Catfish Farmers, the Department did so. Commerce stated that it weighed the outputs immediately after every stage of the process and found that some amount of overall weight gain was possible. The Department also specifically confronted the fact that its verification testing produced different results from NTSF’s regular production methods, noting that while time constraints precluded further testing, it was “reasonable to assume” that use of NTSF’s normal procedures would have produced similar results, so the departure Ct. Nos. 20-00104, 20-00105 Page 57 from NTSF’s procedures “could reasonably” account for any difference. Appx1013–1014. Finally, Commerce found NTSF’s results consistent with past verification results. Appx1014. Accordingly, the Department cited substantial evidence in the administrative record to support its conclusion on the “input versus output weight” issue and sufficiently addressed the evidence detracting from its conclusion, so the court sustains its findings on that issue. This further means the court must reject Catfish Farmers’ argument that Commerce should have disregarded all NTSF’s data and applied total facts available with an adverse inference. C. Reported factors of production Catfish Farmers object to the way NTSF reported its factors of production for three reasons. First, they contend NTSF improperly included frozen fish fillets not exported to the United States. Case 20-105, ECF 32, at 49–51. Second, they claim NTSF improperly failed to report its data on a control number–specific basis. Id. at 51–53. Third, they contend NTSF understated the volume of whole fish required to produce the frozen fillets for all types of fillets. Id. at 53–55. 1. Inclusion chandise of non-U.S.-bound mer- As to whether it was correct to include non-U.S.bound merchandise, Catfish Farmers argue that the Ct. Nos. 20-00104, 20-00105 Page 58 respondent must report only U.S.-bound merchandise and that the data be reported on a control number– specific basis. Case 20-105, ECF 32, at 49–51. NTSF, however, responds that it followed Commerce’s instructions and that the final determination so recognizes. Case 20-105, ECF 45, at 10–11. The court agrees with NTSF. Commerce’s Section D questionnaire instructions read as follows: Unless otherwise instructed by the Department, you should report factors information for all models or product types in the U.S. market sales listing submitted by you (or the exporter), in response to Section C of the questionnaire, including that portion of the production that was not destined for the United States. Appx6908 (first emphasis in original, second added). The Department found that NTSF complied with the instructions: “Commerce instructs respondents to report [control number–]specific [factors of production] regardless of the ultimate destination of the finished product. NTSF has met this reporting requirement.” Appx1014. The instructions clearly required NTSF to include non-U.S.-bound merchandise. Catfish Farmers may believe it was inappropriate for Commerce to require that information as part of the reporting, but they do not object to the instructions, so the court will not second-guess the Department. Ct. Nos. 20-00104, 20-00105 Page 59 2. Use of “standard usage rates” As to how to report factors of production, the instructions read as follows: If you are not reporting factors of production using actual quantities consumed to produce the merchandise under review on a [control number–]specific basis, please provide a detailed explanation of all efforts undertaken to report the actual quantity of each [factor of production] consumed to produce the merchandise under review on a [control number–]specific basis. Additionally, please provide a detailed explanation of how you derived your estimated [factor-of-production] consumption for merchandise under review on a [control number–]specific basis and explain why the methodology you selected is the best way to accurately demonstrate an accurate consumption amount. For the most significant material input, for electricity, and for labor, please reconcile with worksheets the estimated consumption of these [factors of production] for a specific [control number] to your cost-of-production ledger or the equivalent production ledger. Appx6910 (“defined terms” omitted). NTSF’s response stated: NTSF has relied on its standard usage rates maintained in its normal course of business to Ct. Nos. 20-00104, 20-00105 Page 60 report [control number–]specific factors of production. Relying on the NTSF’s standard usage rates and applied variances fulfills the Department’s requirements for [control number–]specific product information maintained by NTSF in its normal course of business. Exhibit D-3 contains two worksheets demonstrating how NTSF calculated the usage rates for all of the farming inputs and whole fish for the largest [control number] produced for the United States. Appx6910–6911 (boldface removed). Elsewhere, NTSF explained how it calculated consumption for particular factors of production. See Appx6919–6920. Commerce concluded that NTSF’s reporting was acceptable and noted that as to the “weight band of the fillet,” Catfish Farmers were repeating an argument the Department had rejected in the previous review: “[T]he petitioners have not established with any compelling record evidence that the consumption rate would vary between products that are identical with respect to all characteristics except for the size of the fillet.” Appx1014. The Department agreed with Catfish Farmers’ general principle that a larger raw fish fillet requires more inputs but noted that the converse is also true—a smaller raw fillet requires fewer inputs, “and Commerce has yet to see any compelling information establishing that there is any meaningful difference regarding the per-unit consumption of inputs with regard to Ct. Nos. 20-00104, 20-00105 Page 61 raw fillet size on an unsoaked basis.” Id. (emphasis in original). In other words, if a larger fillet requires more inputs and a smaller fillet requires fewer inputs, the average is somewhere in between. “Moreover, this comparison becomes blurred as Commerce requires parties to submit [factors of production] and sales on a soaked basis, not an un-soaked basis,” because the levels of soaking could be considerably different and could mean that two fillets with the same pre-soaking weight could have very different weights after soaking. “In sum, to say that two soaked fillets necessarily consume equal amounts of factors would be misleading and unsupported by the record.” Id. Commerce also found that NTSF complied with instructions about how to report the form of its fish fillets because the control number instructions did not distinguish between the size of the fillet or how much it was trimmed or processed—“a shank fillet is a shank fillet no matter how much trimming/processing it may undergo.” Id. The administrative record reveals that the portion of the control number signifying the product form simply referred to the type of fillet—regular, shank, and so on—and that product size was reported via a different part of the control number. See Appx94288. Once again, Commerce found that NTSF followed instructions in reporting its data. The court will not second-guess that finding. Catfish Farmers’ complaint Ct. Nos. 20-00104, 20-00105 Page 62 appears to be that the Department should have issued different or stricter instructions. See Case 20-105, ECF 32, at 54–55 (“[H]ad Commerce followed its own precedent, and the teaching in [Mukand, Ltd. v. United States, 767 F.3d 1300, 1307 (Fed. Cir. 2014)], it would have been able to collect input data specific to the subject merchandise.”). 20 As with the inclusion of non-U.S.-bound merchandise, Catfish Farmers’ complaint is with how Commerce structured the review and directed the respondents to provide their data, but Catfish Farmers have not challenged any of that before this court. An agency has discretion to frame its instructions as it deems appropriate, and the court will not penalize NTSF for complying with those instructions simply because Catfish Farmers contend they were flawed. 3. Ratio of whole live fish to fillets Catfish Farmers contend that NTSF consistently understated the volume of whole live fish required to produce its fillets, arguing that NTSF’s reported ratio of live fish to fillets was consistently lower than the ratios reported by studies in the administrative record. Catfish Farmers point to a 2005 report by the Norwegian Institute of Fisheries and Aquaculture Research—known as Fiskeriforskning—that examined the slaughtering process for farmed pangasius in Mukand involved a respondent that failed to cooperate. Here, however, Commerce found that NTSF followed instructions in reporting its data. 20 Ct. Nos. 20-00104, 20-00105 Page 63 Vietnam. The report found that when farming fish in ponds, 3.2 kg of whole fish is needed to yield 1 kg of fillet (i.e., the average yield is 31 percent). Appx2560. Catfish Farmers also contend that the Fiskeriforskning report’s conclusion was “corroborated” by other data in the administrative record, citing a 2007 “discussion paper” prepared by an employee of Can Tho University in Vietnam and a 2017 affidavit from an Indonesian professor specializing in pangasius fish production and processing in Indonesia. See Case 20-105, ECF 32, at 52. The Can Tho paper stated, “On average, fillets account for 30–40% of the weight of a whole fish. More specifically, 3.2 kilograms of live Tra or 3.9 kilograms of Basa are required to produce one kilogram of fillets.” 21 Appx7668. The Indonesian professor stated that the yields would vary slightly depending on the fillet produced but would range from a ratio of 2.1 to 2.4 kg of whole live fish per kilogram of “untrimmed regular fillets” to a ratio of 3.1 to 3.4 kg of whole live fish per kilogram of “trimmed shank fillets.” Appx95581. He also stated, “The percent yields provided above will not vary significantly, even for the most efficient pangasius fillet processors. The yields provided above will also not vary for pangasius fish of all species grown in any country.” Appx95581–95582. The government’s briefing brushes past the reports Catfish Farmers cited and states that Commerce 21 “Tra” and “basa” are types of pangasius. Ct. Nos. 20-00104, 20-00105 Page 64 found NTSF’s reporting to be accurate. Case 20-105, ECF 42, at 49–50. NTSF, in turn, argues that the Fiskeriforskning report “is very old” due to its 2005 date and that it “does not appear to have been an indepth study of fish processing yields for particular products or specifications,” such that “this fifteenyear-old general survey does not offer useful information to benchmark NTSF’s [factors-of-production] consumption rates for particular specifications of fillets.” Case 20-105, ECF 45, at 11. 22 The court can find no indication that Commerce engaged with the reports Catfish Farmers offered. Instead, the extent of any reference to the reports in the final determination is a single sentence reading, “We note that both parties point to reports on the record and proffered their estimations of what the appropriate yield should be.” Appx1012. This sentence appears as part of the overall discussion of whether the output weight could exceed the input weight, and the Department followed the sentence by concluding as follows: However, given the above analysis, we find that the small difference between the outputs and inputs could reasonably be accounted for by: (1) the fact that Commerce’s post-trimmed fillets did not sit in chilled water for several hours, as is typical; and (2) the fact that Commerce’s yield NTSF does not address the Can Tho University paper or the Indonesian professor’s affidavit. 22 Ct. Nos. 20-00104, 20-00105 Page 65 test did not weigh the by-products in the way NTSF normally weighs them, i.e., after being transported by water to the by-products staging area. Appx1013. This discussion is not responsive to Catfish Farmers’ assertions about whether NTSF’s reported yields were accurate, and it fails to address the reports themselves. 23 Because Commerce did not discuss the reports Catfish Farmers offered, its decision about the accuracy of NTSF’s reporting is not supported by substantial evidence. “Not addressing the conflicting evidence on the record fails the substantial evidence test because it does not consider record evidence contrary to Commerce’s determination.” New Am. Keg v. United States, Catfish Farmers also contend that NTSF allocated part of the whole fish input to byproducts for which NTSF later took an offset, reducing its normal value. Catfish Farmers argue, citing case law from this court, that “[i]f the normal value of the finished frozen fillets is reduced to the extent of any revenue from the sale of fish heads and bones, it is improper to also subtract the weight of those fish heads and bones from the input assigned from the fillets.” Case 20-105, ECF 32, at 54. The court understands the concern to be with double-counting: If a producer takes an offset for the revenue made from selling the byproducts, then the producer should have to account for the cost of producing the byproducts in the first instance. See Case 20-105, ECF 55, at 22. It does not appear that Commerce, the government, or NTSF have addressed Catfish Farmers’ argument on this point. 23 Ct. Nos. 20-00104, 20-00105 Page 66 Ct. No. 20-00008, Slip Op. 21-30, at 35, 2021 WL 1206153, at *13 (CIT Mar. 23, 2021) (cleaned up) (quoting Camau Frozen Seafood Processing Imp. Exp. Corp. v. United States, 929 F. Supp. 2d 1352, 1356 (CIT 2013)). The court must remand for Commerce to address the reports, and on remand the Department should also address the issue noted in footnote 23, above. D. Moisture content Catfish Farmers contend that NTSF systematically overstated the amount of water in its finished frozen fish fillets and understated the amount of actual fish and that Commerce erred by accepting NTSF’s figures. Catfish Farmers explain that “[t]he impact of soaking is substantial” because as more water is added to the fillet, less fish is needed to produce the same net weight. Thus, they argue, “the weight of soaking water per finished frozen fish fillet has a major impact on usage rate for whole fish used to produce that frozen fillet.” Case 20-105, ECF 32, at 55. Catfish Farmers argue that NTSF’s product labels, Case 20-105, ECF 31, at 56; studies and other documentation in the administrative record, Case 20-105, ECF 32, at 56 (citing Appx91078–91091, Appx91108–91113, Appx91114– 91123, and Appx92604–92652); and the results of verification in this review, id. at 56–58, all show moisture content that differed from what NTSF reported in its questionnaire responses. Ct. Nos. 20-00104, 20-00105 Page 67 Commerce briefly addressed the moisture content issue in its final determination and stated that two of its three moisture tests “were not done in a manner that fully conforms to NTSF’s actual production experience” for two reasons. Appx1013. First, the fillets were “patted dry more than is typical in the regular production process.” Id. Second, in one of the tests the fillet did not sit in chilled water for several hours before freezing. Id. The Department noted that time constraints on verification prevented re-testing and concluded, “Given these facts, we do not find that the moisture test results are fully representative of NTSF’s actual experience, and thus, do not necessarily undermine the reliability of NTSF’s reporting with respect to moisture.” Id. NTSF argues that it would have been unreasonable for Commerce to have disregarded NTSF’s data based on a single test the Department acknowledged was procedurally problematic and further notes that third-party inspector reports in the administrative record confirmed NTSF’s reported data. Case 20-105, ECF 45, at 12–13 (citing Appx17028). Catfish Farmers raise three points in reply. First, they contend Commerce’s arguments are “not persuasive” and contradict the record evidence Catfish Farmers introduced that the Department did not address. Second, they assert that it is unreasonable to disregard departures from NTSF’s standard procedures during verification because “NTSF itself conducted the trials at verification and had every incentive and opportunity to duplicate the reported results.” Third, Ct. Nos. 20-00104, 20-00105 Page 68 they note that “this Court recently affirmed Commerce’s reliance on a total adverse facts available rate in a prior review of this proceeding because the respondent failed to report its net weight (moisture content) physical characteristic on an equal basis to allow for accurate matching by [control numbers].” Case 20-105, ECF 54, at 24–26 (citing, for the final argument, Hung Vuong, 483 F. Supp. 3d at 1357–64). As to the final point, Hung Vuong has no relevance because here the Department found that NTSF did follow the relevant control number instructions. Appx1014. In Hung Vuong, it was essentially undisputed that the respondent failed to do so. See 483 F. Supp. 3d at 1358–61. As to verification procedures, Commerce stated that time constraints prevented following NTSF’s actual production processes in full. The court will not disturb that finding because the agency has discretion to determine verification procedures, including on an ad hoc basis. “This Court has previously acknowledged Commerce’s discretion in setting the length of verifications, in recognition of the time constraints imposed by statute for the completion of the review as well as limits on the agency’s resources.” Fujian Mach. & Equip. Imp. & Exp. Corp. v. United States, 178 F. Supp. 2d 1305, 1318 (CIT 2001) (citing, inter alia, Micron Tech., Inc. v. United States, 117 F.3d 1386, 1396 (Fed. Cir. 1997) (affording Commerce “the latitude to derive verification procedures ad hoc”)). “The Ct. Nos. 20-00104, 20-00105 Page 69 Court defers to the agency’s sensibility as to the depth of the inquiry needed.” FAG Kugelfisher Georg Schafer AG v. United States, 131 F. Supp. 2d 104, 106 (CIT 2001), rev’d on other grounds sub nom. FAG Italia S.p.A. v. United States, 402 F.3d 1356 (Fed. Cir. 2005) (mem.). Catfish Farmers’ contentions about the record evidence, however, are valid. Commerce’s final determination does not address any of the record evidence introduced by either Catfish Farmers or NTSF in support of their contentions. “To determine if substantial evidence exists, we review the record as a whole, including evidence that supports as well as evidence that fairly detracts from the substantiality of the evidence.” Nippon Steel, 337 F.3d at 1379. “Not addressing the conflicting evidence on the record fails the substantial evidence test because it does not consider record evidence contrary to Commerce’s determination.” New Am. Keg, Slip Op. 21-30, at 35, 2021 WL 1206153, at *13 (cleaned up). In this case, the Department failed to address both the record evidence contrary to its decision and the record evidence potentially supportive of its decision. Thus, the court must remand for Commerce to address the parties’ evidence on the “moisture content” issue. * * * For the reasons set forth above, the court grants Catfish Farmers’ motion for judgment on the agency Ct. Nos. 20-00104, 20-00105 Page 70 record in part and denies it in part and remands for further proceedings consistent with this opinion. Conclusion In Case 20-104, the court denies NTSF’s motion for judgment on the agency record and enters judgment on the agency record in favor of the government and Catfish Farmers. See USCIT R. 56.2(b). A separate judgment will issue. See USCIT R. 58(a). In Case 20-105, the court grants Catfish Farmers’ motion for judgment on the agency record in part and denies it in part and remands for further proceedings consistent with this opinion. A separate remand order will issue. Dated: April 25, 2022 New York, NY /s/ M. Miller Baker Judge

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.