Apex Frozen Foods Private Ltd. v. United States

208 F. Supp. 3d 1398, 2017 CIT 23, 39 I.T.R.D. (BNA) 1017, 2017 Ct. Intl. Trade LEXIS 23

• Court: United States Court of International Trade
• Decided: March 2, 2017
• Docket: Slip Op. 17-23; Court 15-00282
• Status: Published

Opinion

OPINION

Kelly, Judge:

This action comes before the court on Plaintiffs’ motion for judgment on the agency record pursuant to USCIT Rule 56.2. See Pis.’ Rule 56.2 Mot. J. Agency R., Apr. 19, 2016, ECF No. 36 (“Pis.’ 56.2 Mot.”). Plaintiffs, Apex Frozen Foods Private Limited, et al., 1 challenge various aspects of the Department of Commerce’s (“Department” or “Commerce”) application of its differential pricing analysis in its antidumping duty margin calculations in the final determination in the ninth administrative review of the antidumping duty order on certain frozen warmwater shrimp from India for the period of February 1, 2013 through January 31, 2014. See id.; see generally Certain Frozen Warmwater Shrimp From India, 80 Fed. Reg. 54,524 (Dep’t Commerce Sept. 10, 2015) (final results of antidumping duty administrative review; 2013-2014) (“Final Results”) and accompanying Issues and Decision Memorandum for the Final Results of the Anti-dumping Duty Administrative Review of Certain Frozen Warmwater Shrimp from India, A-533-840, (Sept. 2, 2015), ECF No. 22-3 (“Final Decision Memo”); Certain Frozen Warmwater Shrimp from India, 70 Fed. Reg. 5,147 (Dep’t of Commerce Feb. 1, 2005) (amended final determination of sales at less than fair value and antidump-ing duty order) (“Order”).

Plaintiffs challenge three aspects of Commerce’s differential pricing analysis and resultant application of the A-to-T methodology to all U.S. sales. See Mem. Support Pis.’ Rule 56.2 Mot. J. Agency R. Confidential Version 13-37, Apr. 18, 2016, ECF No. 35 (“Pis.’ Br.”). First, Plaintiffs challenge Commerce’s use of annual and quarterly weighted-average prices in the differential pricing analysis for administrative reviews generally and specifically as applied in this review. Pis.’ Br. 14-22. Second, within Commerce’s meaningful differences test, Plaintiffs challenge Commerce’s inclusion of all U.S. sales and Commerce’s decision to offset positive dumping margins with negative dumping margins in the average-to-average (“A-to-A”) comparison methodology but not in the average-to-transaction (“A-to-T”) comparison methodology. Pis.’ Br. 22-35. Third, Plaintiffs challenge Commerce’s application of the A-to-T methodology to all U.S. sales for both mandatory respondents in this review. Pis.’ Br. 35-36. For the reasons set forth below, the court sustains the final determination in all respects.

BACKGROUND

On February 1, 2005, Commerce issued the underlying antidumping duty (“ADD”) order covering certain frozen warmwater shrimp from India. See Order, 70 Fed. Reg. at 5,147. On April 30, 2014, Commerce initiated the ninth administrative review of the Order for the period of February 1, 2013 through January 31, 2014 for 211 respondent companies. See Certain Frozen Warmwater Shrimp from India and Thailand, 79 Fed. Reg. 24,398 (April 30, 2014) (initiation of antidumping and countervailing duty administrative reviews and request for revocation in part); Certain Frozen Warmwater Shrimp from India and Thailand, 79 Fed. Reg. 18,510 (April 2, 2014) (notice of initiation of ADD administrative review).

Commerce selected Devi Fisheries Limited (“Devi Fisheries”) and Falcon Marine Exports Limited (and its affiliate K.R. Enterprises) (“Falcon Marine”) as mandatory respondents in this administrative review. See Certain Frozen Warmwater Shrimp From India: Preliminary Results of Antidumping Duty Administrative Review; 2013-2014, 80 Fed. Reg. 12,147 (Dept, of Commerce March 6, 2015) (preliminary results of ADD administrative review) (“Prelim. Results”) and accompanying Decision Memorandum for the Preliminary Results of the 2013-2014 Administrative Review of the Antidumping Duty Order on Certain Frozen Warmwater Shrimp from India, A-533-840, at 2, PD 151, bar code 3262269-01 (Mar. 2, 2015) (“Prelim. Decision Memo”); see also 19 U.S.C. § 1677f-l(c)(2)(B); Final Results, 80 Fed. Reg. at 54,524. Commerce found a pattern of significant price differences among purchasers, regions, or time periods for both mandatory respondents’ U.S. sales of comparable merchandise during the period of review. Prelim. Decision Memo at 7. Commerce preliminarily determined that the weighted-average dumping margins for both mandatory respondents should be calculated by applying the A-to-T methodology to all U.S. sales because more than 66% of both mandatory respondents’ U.S. sales passed the Cohen’s d test. Id. Commerce also determined that its A-to-A methodology 2 could not account for these patterns of differentially priced sales because the dumping margin calculated using the A-to-A methodology was below the de min-imis dumping threshold and the dumping margin calculated using the A-to-T methodology was above the de minimis threshold for both mandatory respondents. Id. Commerce thus preliminarily applied the A-to-T methodology to all sales and determined that both respondents had U.S. sales at less than fair value during the period of review. See Prelim. Results, 80 Fed. Reg. at 12,147. Commerce preliminarily assigned Devi Fisheries a weighted-average dumping margin of 3.28% and Falcon Marine a weighted-average dumping margin of 2.63%. Id. at 12,148.

Commerce published the final results on September 10, 2015, in which Commerce continued to find that both mandatory respondents had sold subject merchandise at less than fair value during the period of review. See Final Results, 80 Fed. Reg. at 54,524; Final Decision Memo at 1. Commerce also continued to find that its A-to-A methodology could not account for the pattern of significant price differences found among comparable merchandise for both mandatory respondents and accord ingly applied its A-to-T methodology to all U.S. sales for both mandatory respondents to calculate the weighted-average dumping margins. 3 Final Decision Memo at 3.

On April 18, 2016, Plaintiffs moved for judgment on the agency record pursuant to USCIT Rule 56.2, see Pis.’ 56.2 Mot., arguing that Commerce’s findings are unsupported by substantial evidence and otherwise not in accordance with law. See Pis.’ Br. 13-37.

JURISDICTION AND STANDARD OF REVIEW

The court has jurisdiction pursuant to Section 516A of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2012) 4 and 28 U.S.C. § 1581(c) (2012), which grant the court authority to review actions contesting the final determination in an administrative review of an anti-dumping duty order. The court will uphold Commerce’s determination unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i).

DISCUSSION

I. Commerce’s Use of Annual and Quarterly Average Prices

Plaintiffs argue that Commerce’s use of annual and quarterly weighted-average prices in the Cohen’s d test during the differential pricing analysis for administrative reviews is contrary to law and is not supported by substantial evidence. Pis.’ Br. 14-22. Specifically, Plaintiffs first incorporate by reference the argument they unsuccessfully made before this court in previous litigation that the statute precludes Commerce’s methodology, id. at 15-16, n.3; see Apex Frozen Foods Private Ltd. v. United States, 40 CIT -, -, 144 F.Supp.3d 1308, 1324-1326 (Feb. 2, 2016) (“Apex II”), and additionally contend that the use of annual and quarterly weighted-averages is unreasonable and distortive because Commerce uses monthly weighted-averages to calculate dumping margins in reviews under the standard A-to-A methodology. Id. at 14-18. Defendant responds first that Plaintiffs have waived any argument that the statute precludes Commerce’s methodology by failing to fully develop the argument in the body of its moving brief. Def.’s Resp. Opp’n Pis.’ Rule 56.2 Mot. J. Agency R. 15, Aug. 5, 2016, ECF No. 43 (“Def.’s Resp.”). Defendant argues that the use of annual and quarterly weighted-average prices in the differential pricing analysis is a reasonable exercise of Commerce’s discretion. Id. at 14-23. For the reasons that follow, the court concludes that Commerce’s use of annual and quarterly weighted-averages in the Cohen’s d stage of the differential pricing analysis during administrative reviews is in accordance with law and supported by substantial evidence.

As a preliminary matter, the court exercises its discretion to consider Plaintiffs’ argument that the use of annual and quarterly weighted-averages conflicts with the statute, although Plaintiffs made this argument primarily in a footnote. See Pis.’ Br. 15-16, n.3. As a general rule, when an argument is raised primarily in a footnote and referenced perfunctorily in the- main body of an opening brief, the argument is not preserved. See SmithKline Beeeham Corp. v. Apotex Corp., 439 F.3d 1312, 1320 (Fed. Cir. 2006). However, the court retains discretion to consider improperly-raised arguments, Becton Dickinson and Co. v. C.R. Bard, Inc., 922 F.2d 792, 800 (Fed. Cir. 1990), which courts have specifically recognized to include the discretion to consider arguments raised only in footnotes. See, e.g., Smithkline Beeeham Corp., 439 F.3d at 1320, n.9; Otsuka Pharmaceutical Co., Ltd, v. Sandoz, Inc., 678 F.3d 1280, 1294 (Fed. Cir. 2012).

Plaintiffs stated above the line in their moving brief that they “do not concede the statute is silent” with regard to determining the existence of a pattern of significant price differences, Pis.’ Br. 15-16, and expanded upon this statement in a footnote:

Plaintiffs have repeatedly challenged Commerce’s assertion that the statute does not address this issue, or is otherwise silent on this point. Plaintiffs acknowledge that this Court recently ruled otherwise. Nevertheless, Plaintiffs recently appealed Apex II to the Federal Circuit. Plaintiffs continue to assert that the statutory language as well as the underlying purpose of the statute do not permit Commerce to use annual weighted-averages for this purpose. Plaintiffs specifically preserve that argument here and cla[i]m it is unreasonable and otherwise contrary to the statute (upon which Commerce relies for authority in [ADD] reviews) and contrary to congressional intent to use annual weighted-averages to determine the existence of [significant price differences] in all instances without regard to the facts and circumstances.

Id. at 16, n.3 (internal citations omitted). Plaintiffs aver that, although this argument has been rejected by a prior decision of this court, they are appealing that decision and wish to “preserve that argument” here as well. Id.; see Pis.’ Reply Br. Confidential Version 6, n.4, Oct. 12, 2016, ECF No. 54 (“Pis.’ Reply”).

The court exercises its discretion to review Plaintiffs’ argument because it was previously fully developed and squarely addressed, see Apex II, 40 CIT at-, 144 F.Supp.3d at 1324-1326, and, as Plaintiffs referenced this argument in the main body of its moving brief, the court finds that Defendant was on notice of the argument. See Pis.’ Br. 15-16. Accordingly, it is not unfair to allow Plaintiffs to preserve this argument for appeal. See Becton Dickinson and Co., 922 F.2d at 800 (noting that the practice to consider as waived an issue not properly raised in an opening brief is “not governed by a rigid rule but may as a matter of discretion not be adhered to where circumstances indicate that it would result in basically unfair procedure.”).

Nonetheless, Plaintiffs’ argument that the statute precludes Commerce’s use of annual and quarterly weighted-average prices in the Cohen’s d test during the differential pricing analysis for administrative reviews is unpersuasive. Commerce has broad authority under the statute to craft an appropriate methodology to discern whether there is in fact a pattern of prices that differ significantly. Apex II, 40 CIT at-, 144 F.Supp.3d at 1324-1326 (noting that “Congress has granted Commerce considerable discretion to construct a methodology to apply in a review,” and that Commerce is not required by statute to use individual rather than weighted-average export prices or to use monthly rather than annual and quarterly weighted-average export prices.)

Likewise unavailing is Plaintiffs’ argument that the use of annual and quarterly weighted-averages in the Cohen’s d test in the differential pricing analysis for administrative reviews is unreasonable generally and as applied in this review. See Pis.’ Br. 14-20. According to Plaintiffs, Commerce’s use of annual and quarterly, rather than monthly, weighted-average prices in the Cohen’s d test for administrative reviews is unreasonable because it conflicts with the purpose of the Cohen’s d test and leads to distorted results. Id Defendant responds that it is reasonable to use annual and quarterly weighted-average prices when comparing prices of the test and comparison groups in the Cohen’s d test because the purpose of the Cohen’s d test is to conduct an overview assessment of a respondent’s pricing behavior. Def.’s Resp. 15-19.

Commerce ordinarily uses the A-to-A methodology to calculate dumping margins. 19 U.S.C. § 1677f—1(d)(1)(A); 19 C.F.R. § 351.414(c)(i). 5 Where Commerce determines that there is “a pattern of export prices ... for comparable merchandise that differ significantly among purchasers, regions, or periods of time” and finds that its A-to-A methodology cannot account for those differences, Commerce calculates dumping margins using the alternate A-to-T methodology in investigations. See 19 U.S.C. § 1677f-l(d)(l)(B)(i)-(ii). Commerce has, through practice, adopted the same basis for applying its A-to-T methodology in administrative reviews as is permitted by statute for investigations. See JBF RAK LLC v. United States, 790 F.3d 1358, 1364 (Fed. Cir. 2015). 6

In the absence of statutory or regulatory guidance instructing Commerce how to determine whether a pattern of significant price differences exists, Commerce is afforded broad discretion to select a methodology to make that determination. See Fujitsu General Ltd., 88 F.3d 1034, 1039 (Fed. Cir. 1996); Torrington Co. v. United States, 68 F.3d 1347, 1351 (Fed. Cir. 1995); Smith-Corona Group v. United States, 713 F.2d 1568, 1571 (Fed. Cir. 1983), cert. denied, 465 U.S. 1022, 104 S.Ct. 1274, 79 L.Ed.2d 679 (1984). In situations involving complex and technical methodological choices, such as here, Commerce is given a wide level of discretion and the court need only address whether Commerce’s methodological choice is reasonable. See Motor Vehicle Mfrs. Ass’n of U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 48-49, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) (“[A]n agency must cogently explain why it has exercised its discretion in a given manner.”); Fujitsu Gen. Ltd., 88 F.3d at 1039 (granting Commerce significant deference in determinations “involv[ing] complex economic and accounting decisions of a technical nature”); Ceramica Regiomontana, S.A. v. United States, 10 CIT 399, 404-405, 636 F.Supp. 961, 966 (1986), aff'd, 810 F.2d 1137, 1139 (Fed. Cir. 1987).

Commerce has chosen to determine whether a pattern of significant price differences exists by applying the differential pricing analysis. See Prelim. Decision Memo at 5-7; Final Decision Memo at 15-17. The Cohen’s d test is the first stage of the differential pricing analysis, and Commerce uses the Cohen’s d test to measure the degree of price disparity between two groups of sales. Prelim. Decision Memo at 6; Final Decision Memo at 20. The Cohen’s d test calculates the number of standard deviations by which the weighted-average net prices of U.S. sales for a particular purchaser, region, or time period (the “test group”) differ from the weighted-average net prices of all other U.S. sales of comparable merchandise (the “base group”). 7 See Prelim. Decision Memo at 6.

It is reasonably discernible that Commerce intends the Cohen’s d test, its first examination into whether a pattern of prices that differ significantly among purchasers, regions, or time periods exists, to be an overview assessment of pricing behavior over the course of the one-year period being reviewed. See Final Decision Memo at 19-20; Def.’s Resp. 17-18. Although Commerce’s explanation is less robust than the court would like, it is reasonably discernible that Commerce developed a methodology using weighted-averages because the “A-to-A comparison method also uses weighted-average CONNUM-specific 8 export prices.” Final Decision Memo 20. It is reasonably discernible that, because Commerce is searching for a pattern of prices over the course of the entire period of review, Commerce compares annual average prices for customers and regions and quarterly average prices for time periods. Given that Commerce seeks to identify a pattern of prices for customers, regions, and time periods for each CONNUM, the court cannot say that it is unreasonable to choose the least burdensome time frames for each comparison it must make. Commerce did not act unreasonably in choosing to use annual and quarterly weighted-averages to make this initial pattern assessment in the Cohen’s d test. The reasonableness of using annual or quarterly weighted-averages in this first stage of the differential pricing analysis is tied to the purpose of this first stage. The Cohen’s d test is designed to determine whether conditions exist that may serve to mask dumping, not to determine whether the masked dumping is sufficient to warrant applying its A-to-T methodology and to what extent to do so. Final Decision Memo at 19-20; Def.’s Resp. 17-18. Whether masked dumping is a factor that may warrant application of Commerce’s A-to-T methodology is addressed when examining whether the A-to-A method “can account for” such a condition. Final Decision Memo at 24; Def.’s Resp. 17-18. Thus, placed in context, it is not unreasonable for Commerce to first check for a pattern of price differences using what Defendant calls a “10,000 foot overview” assessment of the entire period of review, Oral Argument at 00:32:11-00:32:22, Jan. 9, 2017, ECF No. 68 (“Oral Argument”), before comparing monthly weighted-average prices in subsequent stages of the differential pricing analysis. See Prelim. Decision Memo at 6-7; 19 C.F.R. §§ 351.414(d)(3), (e). As annual and quarterly weighted-averages reasonably provide such an overview assessment, the choice to use annual and quarterly weighted-averages appears tailored to effectuate the statutory directive, as implicated by Commerce’s practice, and is therefore reasonable. See Final Decision Memo at 21; Cerámica Regiomontana, S.A., 636 F.Supp. at 966, affd, 810 F.2d at 1139 (affording deference to Commerce’s methodology so long as it reasonably effectuates the statu tory purpose and is supported by substantial evidence).

Plaintiffs’ argument against the use of annual and quarterly weighted-averages in the Cohen’s d test mischaracterizes the objective of the Cohen’s d test. Plaintiffs characterize the Cohen’s d test as an examination, for significant price differences, of the weighted-average prices that would ultimately be used to calculate dumping margins using A-to-A prior to actually calculating dumping margins with those averages. 9 See Pis.’ Br. 17-18; Oral Argument at 00:06:34-00:09:22 (quoting Final Decision Memo at 20). Plaintiffs contend that Commerce erred by “simply borrowing]” the differential pricing methodology developed for investigations and applying it to reviews without adjusting for the fact that, in investigations, Commerce typically uses annual and quarterly weighted-averages to calculate dumping margins using A-to-A and, in reviews, typically uses monthly weighted-averages to calculate dumping margins using A-to-T. Pis.’ Br. 16-18. Plaintiffs reason that, as monthly weighted-averages are used in the calculation of dumping margins for reviews, it would be unreasonable for Commerce to use annual and quarterly weighted-averages in the Cohen’s d test because, in their view, the Cohen’s d test should search for patterns within the same time frame over which the dumping margins will be calculated. Id. Their theory supposes that if masking as envisioned by Congress in 19 U.S.C. § 1677f—1(d)(1)(B) occurs, it will occur within that time frame, he., monthly in reviews. 10 See id.; 19 C.F.R. § 351.414(d)(3). Plaintiffs’ theory fails to confront the fact that 19 U.S.C. § 1677f-1(d)(1)(B)® itself authorizes Commerce to search for patterns of prices that differ significantly by purchasers, regions, or time periods. See 19 U.S.C. § 1677f-1(d)(1)(B)®; efi Final Decision Memo at 6. 11 Therefore, it is not unreasonable for Commerce to develop a methodology that reflects the fact that masking as contemplated by the statute may be occurring across time periods.

Further, the objective of the Cohen’s d test, according to Commerce, is to obtain an overview of prices for the entire period of review before conducting a more precise comparison of monthly weighted-average prices. See Prelim. Decision Memo at 6-7; Final Decision Memo at 20. Only in later stages does Commerce actually compare monthly weighted-average prices. Plaintiffs incorrectly infer that Commerce’s use of annual and quarterly weighted-average prices in Commerce’s Cohen’s d analysis is improperly imported from Commerce’s practice in investigations. See Pis.’ Br. 16-18; Oral Argument at 00:12:50-00:13:00, 00:53:25-00:53:39. But, as discussed, Commerce’s use of annual and quarterly weighted-average prices in its Cohen’s d analysis flows from the Cohen’s d test’s purpose. Commerce has reasonably grounded its determination to use annual and quarterly weighted-averages in conducting its Cohen’s d analysis, and monthly weighted-averages in comparing prices at later stages of its differential pricing analysis, by distinguishing the purposes of the various stages of its methodology. Plaintiffs offer no reason why annual and quarterly weighted-averages are not tailored to the purpose of obtaining an overview of prices for the entire period of review.

Highlighting a passage from Commerce’s final determination, Plaintiffs note that Commerce indicated it “divided the weighted-average price used in the calculation of individual dumping margins into [the test group and the comparison group] in order to examine whether there is a pattern of prices that differ significantly among purchasers, regions or time periods.” Oral Argument at 00:48:47-00:49:07 (quoting Final Decision Memo at 20). Plaintiffs contend that Commerce’s use of the phrase “used in the calculation of individual dumping margins” indicates Commerce’s intention to use monthly weighted-averages in the differential pricing analysis because it uses monthly averages to calculate individual dumping margins. See id. Read in the context of the Final Decision Memo as a whole, Plaintiffs’ reading of Commerce’s language is not persuasive. Commerce does not indicate an intention to apply the same weighted-averages that would be used in the A-to-A margin calculation in this first stage of the analysis. After considering Plaintiffs’ suggestion to use monthly weighted-averages in the Cohen’s d portion of its methodology, Commerce affirmatively chose to continue to use annual and quarterly weighted-averages, see Final Decision Memo at 21, despite the fact that Commerce uses monthly weighted-averages in reviews to calculate dumping margins. See 19 C.F.R. §§ 351.414(d)(3), (e). As already discussed, the court considers that determination reasonable because it effectuates the purpose of the Cohen’s d test and Commerce high lights a reasonable basis for using different averaging at different stages of its methodology. See Final Decision Memo at 19-20. Plaintiffs point to no reason why the use of annual and quarterly averages in the Cohen’s d test is inconsistent with the purpose of the Cohen’s d test to identify patterns of significant price differences over the overall annual period of review. See id.; Def.’s Resp. 17-18.

Plaintiffs also claim that Commerce’s use of annual and quarterly weighted-averages in the Cohen’s d test in this review is distortive as applied, such that Commerce’s determination was unsupported by substantial evidence. 12 Pis.’ Br. 18-20. Plaintiffs have not, demonstrated that Commerce’s use of annual and quarterly weighted-averages is distortive.. Plaintiffs present calculations which show that the application of annual and quarterly weighted-averages results in a different outcome for both mandatory respondents than would result from the application of monthly weighted-averages. 13 Id.; Oral Argument at 00:17:28-00:18:22. However, a difference does not necessarily reflect a distortion. Whether a practice is distortive is inextricably intertwined with whether that practice is reasonable. An unreasonable practice may or may not be distortive but, logically, a reasonable practice cannot be distortive. A reasonable practice that leads to different results is simply different.

II. The Meaningful Differences Test

Plaintiffs argue that Commerce acted unreasonably and unlawfully in applying the meaningful differences test to determine whether the standard A-to-A methodology could account for the price differences uncovered by the differential pricing analysis. Pis.’ Br. 22-35. The court first discusses the reasonableness of Commerce’s inclusion of all sales in its assessment of whether the A-to-A methodology can account for price differences (i.e., the meaningful differences test). 14 Second, the court addresses the reasonableness of Commerce’s decision not to offset positively dumped sales with negatively dumped sales in the meaningful differences test. For the reasons that follow, Commerce’s application of the meaningful differences test is reasonable.

A. Commerce’s Inclusion of All U.S. Sales in the Meaningful Differences Test

Plaintiffs challenge Commerce’s comparison of an A-to-A margin calculation for all of the mandatory respondents’ sales to an A-to-T calculation for all of those same sales. Pis.’ Br. 24-31. Plaintiffs contend that it was unreasonable for Commerce to apply the meaningful differences test to U.S. sales not exhibiting significant price differences. 15 See id.

According to Plaintiffs, the language of 19 U.S.C. § 1677f-l(d)(l)(B)(ii), which applies directly to investigations and which Commerce has adopted by practice in reviews, suggests that only sales exhibiting significant price differences should be analyzed to determine whether A-to-A accounts for those price differences, as “ ‘such differences’ ties to and directly relates to the export sales that demonstrated significant price differences.” Pis.’ Br. 24; see Pis.’ Reply 10-12. The statutory language of 19 U.S.C. § 1677f-l(d)(l)(B)(ii) does not suggest that only sales exhibiting significant price differences should be included in the meaningful differences test. See 19 U.S.C. § 1677f-l(d)(l)(B)(ii). The statute upon which the practice is based provides that Commerce may determine if “there is a pattern of export prices ... that differ significantly among purchasers, regions, or periods of time, and [ ... Commerce] explains why such differences cannot be taken into account using [the standard A-to-A methodology].” See 19 U.S.C. § 1677f-l(d)(l)(B)(i)-(ii). It is unclear from the language of the statute what the impact of such differences may be. Plaintiffs assume that the only impact is the significant price differences themselves. But, as Commerce has explained, the pattern identified may include higher or lower priced sales, Final Decision Memo at 24, and the impact of such differences may not lie only in the sales forming the pattern but may also lie outside the pattern. Id. at 24-25. Therefore, the language of the statute does not implicitly or explicitly preclude application of the meaningful differences test to all sales. See id.

Plaintiffs further argue that it is unreasonable to base the meaningful differences test on a comparison of an A-to-A margin calculation for all of a respondent’s U.S. sales to an A-to-T margin calculation for all of those same sales, as “[t]he whole point of applying the alternative [A-to-T] remedy is to unmask dumping on [sales exhibiting significant prices differences]. Thus, the question whether Commerce’s normal [A-to-A] methodology can ‘account’ for ‘such differences’ relates only to those sales demonstrating ‘such differences’ and not to all sales.” Pis.’ Br. 24.

Commerce uses the meaningful differences test to determine whether the standard A-to-A methodology can account for the pattern of significant price differences found in the differential pricing analysis, prior to using the alternate A-to-T methodology to calculate the overall dumping margin. See Final Decision Memo at 23-25; Prelim. Decision Memo at 6-7; see also 19 U.S.C. § 1677f-l(d)(l)(B)(i)-(ii). In the meaningful differences test, Commerce compares the respondent’s overall weighted-average dumping margin calculated for all U.S. sales using the alternative methodology 16 to the respondent’s overall weighted-average dumping margin for all U.S. sales calculated using the A-to-A methodology. See Final Decision Memo at 24; Prelim. Decision Memo at 6-7. Commerce uses the meaningful differences test to assess whether the margins that would be calculated using each methodology differ meaningfully. Prelim. Decision Memo at 7. The difference between the margins is considered meaningful if: (1) there is at least a 25 percent relative change in the dumping margins calculated by the two method ologies where both margins are above the de minimis threshold (0.5%), or (2) the dumping margin is below the de minimis threshold for one method and above the de minimis threshold for the other method. Id. at 7; see Final Decision Memo at 24. Commerce considers a meaningful difference to demonstrate that the standard A-to-A method cannot account for the significant price differences found, such that the use of the alternate methodology is justified to calculate an accurate margin. Final Decision Memo at 24; Prelim. Decision Memo at 7. In calculating the margins, Commerce offsets sales with negative dumping margins when applying the A-to-A methodology and does not offset sales with negative margins when applying the A-to-T methodology. See Final Decision Memo at 26-28.

Here, Commerce’s ratio test revealed that more than 66% of both mandatory respondents’ U.S. sales passed the Cohen’s d test and Commerce accordingly determined it was appropriate to apply the A-to-T methodology to all U.S. sales for both respondents. Final Decision Memo at 25. When comparing the margins calculated by the two methodologies in the meaningful differences test, Commerce included all sales in the overall margin calculations. See id. at 24-25. In accordance with its practice, Commerce offset sales with negative dumping margins for the A-to-A methodology but did not offset sales with negative dumping margins in calculating its A-to-T margin for purposes of the meaningful difference test. See id. at 28. For both mandatory respondents, the A-to-A methodology yielded an overall dumping margin below the de minimis threshold and the A-to-T methodology yielded an overall dumping margin above the de minimis threshold. Prelim. Decision Memo at 7; see Final Decision Memo at 24-25. Commerce therefore determined that a meaningful difference existed for both mandatory respondents, demonstrating that A-to-A could not account for the pattern of significant price differences for either mandatory respondent, and applied A-to-T to all U.S. sales for both mandatory respondents. Prelim. Decision Memo at 7; see Final Decision Memo at 3, 24-25, 30.

According to Commerce, sales not exhibiting patterns of significant price differences are relevant to assessing whether the A-to-A method can accurately measure a respondent’s dumping margin because “the purpose of considering an alternative comparison method is to examine whether the A-to-A method is appropriate to measure each respondent’s amount of dumping, some of which may be hidden because of masked dumping.” Final Decision Memo at 24. Commerce further explained:

The existence of both dumped and non-dumped sales is necessary to have the potential for masked dumping, and one must consider both low-priced and high-priced sales when determining whether a pattern of prices that differ significantly exists and whether masking is occurring. When the Department looks for a pattern of prices that differ significantly, a pattern can involve prices that are lower than the comparison price or higher than a comparison price. Lower, higher, or both are all possibilities for establishing a pattern consistent with section 777A(d) (1) (B) (i) of the Act.

Id. at 24. Commerce includes all sales in the meaningful differences test because masked dumping requires the presence of higher and lower priced sales. Commerce considers that those which are not differentially priced may still nonetheless be dumped, which can be ascertained by applying A-to-T to all sales. Def.’s Resp. 26 (noting that the SAA “explains that so-called ‘targeted dumping’ may be occur ring -with the sales that constitute a pattern of prices that differ significantly, with the possibility that so-called ‘targeted’ or masked dumping may be occurring elsewhere, too.”). This rationale is reasonable given the objective of the meaningful differences test to compare the effect of the two calculation methodologies prior to applying either as a remedy.

By arguing that the meaningful differences test should be applied only to sales exhibiting differential pricing rather than to all sales, Plaintiffs mistake the objective of the test. Plaintiffs claim that “[t]he whole point of applying the alternative [A-to-T] remedy is to unmask dumping on targeted sales.” Pis.’ Br. 24 (emphasis in original). But the purpose of the meaningful differences test is not to uncover masked dumping, but rather to assess whether the standard A-to-A methodology would yield a dumping margin that accurately reflects the differences uncovered. Final Decision Memo at 24-25. Commerce makes this assessment by comparing the overall dumping margin that would be calculated using the standard A-to-A methodology to the overall dumping margin that would be calculated using the alternate A-to-T methodology. Id. at 24. The overall dumping margin (he., the remedy) will be calculated for all sales, without regard to whether sales are dumped or differentially priced. Therefore, Commerce considers all sales, rather than a subset of the sales, in its analysis to determine whether A-to-A would capture the masked dumping activity and yield an accurate overall margin for all sales. Yangzhou Bestpak Gifts & Crafts Co., Ltd. v. United States, 716 F.3d 1370, 1379 (Fed. Cir. 2013) (“An overriding purpose of Commerce’s administration of antidumping laws is to calculate dumping margins as accurately as possible.”). 17

Plaintiffs also argue that applying the meaningful differences test to all U.S. sales was not supported by substantial evidence under the facts of this case. Pis.’ Br. 27-31. Specifically, Plaintiffs present the results of an alternate meaningful differences test, in which they disaggregated the mandatory respondents’ sales into differentially priced and non-differentially priced sales and conducted the meaningful differences test only on the differentially priced sales with positive dumping margins. Id. This alternate test results in de minimis dumping margins for both mandatory respondents using both A-to-A and A-to-T. 18 Id. Plaintiffs argue that, because this alternate meaningful differences test would lead to the determination that A-to-A in fact does account for the significant price differences found, Commerce’s inclusion of all sales in the meaningful differ- enees test led to the erroneous conclusion that A-to-A could not account for the significant price differences. 19 Id.; Pis.’ Reply 15-17.

Plaintiffs have simply presented an alternate methodology that leads to different results. As discussed above, Commerce reasonably compared an A-to-A margin calculation for all of each respondent’s U.S. sales to an A-to-T margin calculation for all of those same sales to assess whether A-to-A would reflect each respondent’s overall pricing behavior, because “[t]he purpose of considering an alternative comparison method is to examine whether the A-to-A method is appropriate to measure each respondent’s amount of dumping, some of which may be hidden because of masked dumping.” Final Decision Memo at 24. The inclusion of all sales to examine whether the overall dumping margin calculated using A-to-A accurately reflects the differential pricing uncovered is reasonable, and Plaintiffs have not demonstrated that including all sales led to results that are unsupported by substantial evidence.

B. Commerce’s Decision to Offset Negative Dumping Margins Under A-to-A But Not A-to-T in the Meaningful Differences Test

Plaintiffs also claim that it is unlawful, unreasonable, and arbitrary for Commerce to offset positive dumping margins with negative dumping margins under A-to-A but not under A-to-T in the meaningful differences test. Pis.’ Br. 31-35. Plaintiffs contend that, by offsetting margins for only one method in the comparison, Commerce creates distinctions between the margins not otherwise present, Pis.’ Br. 32; Pis.’ Reply 20-21, “leading] to the erroneous conclusion that A-to-A failed to ‘account’ for any dumping on ‘such differences’ for both [respondent] companies.” Pis.’ Br. 32. Plaintiffs argue that Commerce must therefore either offset or not offset margins consistently across both methods in order to accurately determine whether A-to-A can account for significant price differences found. 20 Pis.’ Reply 21; Oral Argument at 01:05:38-01:05:55.

Plaintiffs’ argument is unavailing. Commerce applies the two methodologies in the meaningful differences test as they would be applied when calculating the overall dumping margins; that is, A-to-A with offsets and A-to-T without offsets. Final Decision Memo at 27-28. Commerce applies A-to-A with offsets when calculating the overall dumping margins because the A-to-A “methodology relies on averaging groups of sales, and so inherently allows higher-priced sales to offset lower-priced ones.” Id. at 27. In contrast, the A-to-T calculation is a tool provided to Commerce specifically to uncover dumping in instances in which certain transactions are dumped and certain transactions are not dumped. Id. Allowing the non-dumped sales to offset the dumped sales in this methodology would defeat the purpose of this alternative methodology because it would allow for low-priced sales to offset high-priced sales, which would mask the dumping that the A-to-T methodology is specifically designed to uncover. See id. at 27-28. The use of offsets in A-to-A and not in A-to-T is thus reasonable given the distinct objectives of the two methodologies.

Plaintiffs do not contest Commerce’s authority to not allow offsets by negative sales when applying A-to-T at the remedy stage to calculate the dumping margin, Pis.’ Br. 34, and limit their challenge to Commerce’s not offsetting by negative sales in the A-to-T margin calculation during the meaningful differences test. Id. However, the purposes behind not allowing offsets in both stages is the same. The meaningful differences test is a tool to assess the A-to-A and A-to-T methodologies’ margins prior to application to determine whether A-to-T is a necessary remedy, and it would be illogical to make this assessment using a different version of the methodology than would ultimately be applied as the remedy.

III. Commerce’s Application of A-to-T to All U.S. Sales for Both Mandatory Respondents

Plaintiffs challenge Commerce’s application of the A-to-T methodology to all U.S. sales, including sales that did not pass the Cohen’s d test, when calculating the overall dumping margin for both mandatory respondents. Pis.’ Br. 35-36; Pis.’ Reply Br. 21-25. Plaintiffs challenge both the statutory basis for and the reasonableness of Commerce’s application of A-to-T to all sales. Pis.’ Br. 35-36; Pis.’ Reply 21-25. Both challenges are unpersuasive.

A. The Statutory Basis for A-to-T

Plaintiffs argue that the application of A-to-T to all U.S. sales is contrary to law, because 19 U.S.C. § 1677f-l(d)(l)(B) creates a “specific statutory requirement that Commerce only use the alternative [A-toT] remedy to ‘account’ for the significant ‘differences’ in export prices.” Pis.’ Reply 22; see Pis.’ Br. 36; 19 U.S.C. §§ 1677f-l(d)(l)(B)(i)-(ii). Plaintiffs argue that the “differences” referenced in the statute are the significant price differences in the sales that passed the Cohen’s d test, such that Commerce should have applied A-toT only to the sales that passed the Cohen’s d test because the statutory exception for A-to-T is only applicable to the sales within the pattern of differential pricing. Pis.’ Reply 22; see 19 U.S.C. §§ 1677f-l(d)(l)(B)(i), (ii). Defendant responds that the statute does not limit its authority to apply A-to-T to all sales. Def.’s Resp. 29-31.

The statute upon which Commerce has modeled its practice in reviews provides that Commerce may use the alternate A-to-T methodology to calcúlate a respondent’s dumping duty margin if Commerce finds that there is “a pattern of export prices ... for comparable merchandise that differ significantly among purchasers, regions, or periods of time, and [... Commerce] explains why such differences cannot be taken into account using [A-to-A].” 19 U.S.C. §§ 1677f-1 (d)(1)(B)(i)-(ii), 1677f—1(d)(2). Commerce explained that, since “[n]either the Act nor the SAA provides any guidance in determining how to apply the A-to-T method once the requirements of [19 U.S.C. § 1677f-l(d)(l)(B)(i) and (ii)] have been satisfied,” the Department “created a framework to determine how the A-to-T method may be considered as an alternative to the standard A-to-A method based on the extent of the pattern of prices that differ significantly as identified by the Cohen’s d test.” Final Decision Memo at 30. Commerce employs the ratio test to determine when, and to what degree, A-to-T should be used. In the ratio test, Commerce calculates the percentage of each respondent’s U.S. sales that passed the Cohen’s d test. See id. at 30; Prelim. Decision Memo at 6. Commerce uses that percentage to implement A-to-T in proportion to the degree of significant price differences uncovered, based on a series of thresholds. 21 Final Decision Memo at 30; Prelim. Decision Memo at 6. There is no limiting language in the statute upon which Commerce based its practice in reviews restricting the application of A-to-T only to those sales that passed the Cohen’s d test, where Commerce finds a pattern of significant price differences and determines the alternate methodology is warranted. See 19 U.S.C. § 1677f-l(d)(l)(B)(i)-(ii).

B. The Reasonableness of Commerce’s Methodology

Plaintiffs also argue that it was unreasonable for Commerce to apply A-to-T to all U.S. sales in this review. Pis.’ Br. 35-36; Pis.’ Reply 23-25. Plaintiffs contend that the application of A-to-T to all sales was “beyond the remedy necessary to unmask dumping on targeted sales presumably masked by [A-to-A]” because any “masked dumping” occurred only within the subset of sales that passed the Cohen’s d test. Pis.’ Br. 36; Pis.’ Reply 23-24. Defendant responds that applying A-to-T to all sales was reasonable here, as more than 66% of each respondent’s sales passed the Cohen’s d test and given Commerce’s interest in ensuring a remedy proportionate to the degree of masked dumping found. Def.’s Resp. 29-31. Plaintiffs have not demonstrated that Commerce’s application of A-to-T to all sales where more than 66% of a respondent’s sales pass the Cohen’s d test is unreasonable.

Commerce has determined that, in cases where more than 66% of a respondent’s sales pass the Cohen’s d test, significant price differences are pervasive to a degree that indicates that masked dumping could be occurring outside of the sales that passed the Cohen’s d test, rendering it prudent to check each individual transaction for dumping. See Final Decision Memo at 25 (explaining, by quoting Apex II, that “Commerce applied [A-to-T] across the board to reveal dumping hidden by sales that were neither targeted nor dumped” (internal quotation omitted)); see Def.’s Resp. 13-14, 29. In such instances in which differential pricing is so widespread, the Department reasonably considers it necessary to look individually at each transaction for additional dumping. Commerce has explained that the ratio test is calibrated to ensure that the alternate methodology is applied as a remedy proportionate to the targeted dumping activity the remedy aims to correct. Final Decision Memo at 30. As A-to-T allows for a more precise margin calculation, it is not unreasonable for Commerce to identify levels of behavior and then apply the A-to-T remedy proportionately, thus ensur ing an accurate calculation in instances of such a widespread pattern of differential pricing.

CONCLUSION

For the foregoing reasons, the U.S. Department of Commerce’s final determination in the ninth administrative review of the antidumping duty order covering certain frozen warmwater shrimp from India is sustained. Judgment will enter accordingly.

1 . Apex Frozen Foods Private Limited, et al. includes the following parties: Apex Frozen Foods Private Limited, Amulya Seafoods, An-anda Group (Ananda Aqua Applications, An-anda Aqua Exports (P) Limited, and Ananda Foods), Asvini Fisheries Private Limited, Avanti Feeds Limited, Choice Canning Company, Choice Trading Corporation Private Limited, Devi Fisheries Group (Devi Fisheries Limited, Satya Seafoods Private Limited, and Usha Seafoods), Liberty Group (Devi Marine Food Exports Private Ltd., Kader Exports Private Limited, Kader Investment and Trading Company Private Limited, Liberty Frozen Foods Pvt. Ltd., Liberty Oil Mills Ltd., Premier Marine Products Private Limited, and Universal Cold Storage Private Limited), Falcon Marine Exports Limited and its affiliate K.R. Enterprises, Jagadeesh Marine Exports, Jaya-lakshmi Sea Foods Private Limited, Nekkanti Sea Foods Limited, Nila Sea Foods Pvt. Ltd., Sagar Grandhi Exports Private Limited, Sai Marine Exports Pvt. Ltd., Sai Sea Foods, Sandhya Marines Limited, Sprint Exports Pvt. Ltd., Suryamitra Exim Pvt. Ltd., and Wellcome Fisheries Limited. See Am. Summons, Attach. A, Oct. 16, 2015, ECF No. 9. These parties are referred collectively herein as "Plaintiffs.”

2 . The A-to-A methodology is the methodology that Commerce ordinarily uses to calculate antidumping duty margins. See 19 C.F.R. § 351.414(c)(i). The A-to-A methodology compares the weighted-average of the normal value pricé in the home market to the weighted-average export price in the U.S. market. See 19 U.S.C. § 1677f-l (d)(1)(A).

3 . Consistent with the preliminary determination, in the final determination Commerce calculated a weighted-average dumping margin of 3.28% for Devi Fisheries and 2.63% for Falcon Marine, from which Commerce assigned a rate of 2.96% to the other exporters and producers covered by the review. Final Results, 80 Fed. Reg. at 54,525.

4 . Further citations to the Tariff Act of 1930, as amended, are to the relevant provisions of Title 19 of the U.S. Code, 2012 edition.

5 . Further citations to Title 19 of the Code of Federal Regulations are to the 2014 edition.

6 . The Court of Appeals for the Federal Circuit has held that it is reasonable for Commerce to apply the alternate A-to-T method in administrative reviews by practice as it does in investigations by statute. See JBF RAK LLC v. United States, 790 F.3d 1358, 1364 (Fed. Cir. 2015).

7 . As Commerce explained, purchasers are identified by “the reported consolidated customer codes. Regions are defined using the reported destination zip code and are grouped into regions based upon standard definitions published by the U.S. Census Bureau. Time periods are defined by the quarter within the [period of review] being examined based upon the reported date of sale.” Prelim. Decision Memo at 6.

8 . "CONNUM” is short for "control number,” and is a product code consisting of a series of numbers reflecting characteristics of a product in the order of their importance used by Commerce to refer to particular merchandise. See Final Decision Memo at 17; Def.’s Resp. 11, n.2.

9 .According to Plaintiffs, the purpose of Commerce’s use of annual weighted-averages in its differential pricing methodology in investigations is to "screen” those averages for significant price differences prior to their use in the A-to-A methodology. Pis.’ Br. 17-18. Plaintiffs argue that, accordingly, "Commerce should follow the ‘screening’ logic of its investigation practice in reviews, but modify it to fit the monthly A-to-A comparison used in reviews®..] by using monthly averages as part of its [significant price differences] testing as the screen for the monthly averages used in the [A-to-A] comparison.” Id. at 17. Defendant objects to Plaintiffs’ " ‘screening logic’ argument,” contending that Plaintiffs did not exhaust the argument below and that, in any event, "Commerce has no such ["screening logic”] practice.” Def.’s Resp. 19. Plaintiffs respond that exhaustion is not applicable, as this point is “a summary of Commerce’s policy” rather than an argument in its own right, Oral Argument at 00:20:59-00:21:12, but argue in the alternative that this argument was properly raised before Commerce. Id. at 00:21:18-00:21:42. Although Plaintiffs concede that they did not use the same "screening” terminology below, Plaintiffs emphasize that they "addressed the same basic concept, talked about Commerce’s policy for identifying sales, [... and] pointed out the interrelationship between those two steps." Id. The court concludes that Commerce was on notice of this argument. See Case Brief on Behalf of Respondents at 22-25, CD 156, bar code 3270673-01 (Apr. 15, 2015).

10 . Plaintiffs base this argument on section 19 C.F.R. § 351.414(d)(3), which provides that annual weighted-averages are normally to be used to calculate dumping margins for investigations while monthly weighted-averages are normally to be used to calculate dumping margins for reviews. Pis.’ Br. 16-17. Plaintiffs contend this regulation demonstrates that monthly averages should also be used in the differential pricing analysis for reviews. Id.

11 . Responding to Plaintiffs’ argument that the limited use of monthly averaging in reviews sufficiently combats masked dumping, Commerce explained its rationale in light of the objectives of the statute upon which its practice is based:

The Department’s differential pricing analysis identified a pattern of prices that differ significantly not only by time period, but also by region and purchaser. Thus, even if the use of monthly averaging periods ad-' dressed potential implicit masking of dumping by time within period-wide averaging groups (albeit insufficiently as there may still be implicit masking of dumping within the monthly averaging groups), it does nothing to unmask dumping by purchaser or region, as provided for under section 777A(d)(l)(B) of the Act. As a result, we find that the mere usage of monthly weighted-average normal-values and U.S. prices is insufficient to unmask the full amount of the respondents' dumping in either an investigation or an administrative review as provided for by the statute.

Final Decision Memo at 6.

12 . Defendant-Intervenor argues that the court should disregard Plaintiffs’ calculations presented to demonstrate distortion because Plaintiffs did not provide this "evidence challenge” at the administrative level. Def.-Inter-venor Ad Hoc Shrimp Trade Action Committee’s Resp. Pis.' R. 56.2. Mot. J. Agency R. 18-19, Aug. 5, 2016, ECF No. 42 (“Def-Interve-nor Resp.”); Oral Argument at 00:46:15-00:46:44. Defendant-Intervenor concedes that Plaintiffs did suggest in their administrative brief that Commerce should use monthly averages because "a larger averaging period tends to amplify distortions,” Def.-Intervenor Resp. 18, but contends that Commerce was nonetheless not on notice of Plaintiffs’ substantial evidence argument regarding distortion because the evidentiary support for the distortion argument was not presented at the administrative level. Id. at 18-19 (citing Case Brief on Behalf of Respondents at 25, CD 156, bar code 3270673-01 (Apr. 15, 2015) ("Apex Admin. Br.”).

Although Plaintiffs did not provide the actual calculations to support their argument regarding distortion at the agency level, Plaintiffs did raise to Commerce their objection to the use of annual averages as distortive. Apex Admin. Br. at 17-20, 24-25. Plaintiffs' failure to provide actual calculations to support its argument that the use of annual averages is distortive did not deprive the agency of the opportunity to review the contested procedures and decisions and correct any errors prior to judicial review and intervention. See Parisi v. Davidson, 405 U.S. 34, 37, 92 S.Ct. 815, 31 L.Ed.2d 17 (1972). Commerce was aware of the thrust of Plaintiffs’ objection and had an opportunity to correct errors even without being provided with examples of particular distortions. Therefore, Plaintiffs' argument is not barred by the exhaustion doctrine.

13 . Plaintiffs argue that the use of annual and quarterly averages results in a higher percentage of sales passing the Cohen’s d test for both mandatory respondents than monthly averages would, and that this higher passing sales rate in turn significantly affects both mandatory respondents’ overall dumping margins. Pis.’ Br. 18-20. Specifically, Plaintiffs argue that using monthly weighted-averages in the differential pricing analysis would decrease Devi Fisheries' passing sales from the 73.78 percent calculated using annual averages to [[]] percent and would decrease Falcon Marine’s passing sales from the 83.79 percent calculated using annual averages to [[]] percent. Id at 19. Plaintiffs argue that, for both mandatory respondents, the decrease in passing sales is significant because it lowers the passing sales rate below the 66 percent threshold under which Commerce applies A-to-T only to passing sales and over which Commerce applies A-to-T to all sales. Id.; see Final Decision Memo at 30. Plaintiffs further argue that this change in methodology would lead to a lower overall antidumping margin calculation for both mandatory respondents. Pis.’ Br. 19. Plaintiffs calculate that Devi Fisheries' overall antidumping rate would decrease from 3.28 percent when A-toT is applied to all sales to 2.47 percent when A-to-T is applied only to sales passing the Cohen's d test, and Falcon Marine’s overall antidumping rate would similarly decrease from 2.63 percent to P.89 percent. Id.

14 . In the meaningful differences test, Commerce compares the overall weighted-average dumping margin for all of a respondent’s U.S. sales using the A-to-A methodology to the overall weighted-average dumping margin for all of a respondent’s U.S. sales using the A-to-T methodology. See Final Decision Memo at 23-25; Prelim. Decision Memo at 6-7. Plaintiffs argue that Commerce's meaningful differences test is flawed, as "Commerce should have confined the comparison of [A-to-A] and [A-to-T] margins only to the masking of positive dumping on targeted sales to comply with the statutory requirement whether [A-to-A] can account for ‘such differences.’ ” Pis.’ Br. 29. Accordingly, Plaintiffs would instead have Commerce disaggregate the overall dumping margins calculated using A-to-A and A-to-T into, for each methodology, a margin for sales that passed the Cohen's d test and a margin for sales that did not pass the Cohen’s d test, and then disaggregate each of those margins further into sales with positive dumping margins and sales with negative dumping margins. Id. at 26-31. Plaintiffs suggest Commerce then conduct its meaningful differences test by comparing only the A-to-A and A-to-T margins for the subset of sales that passed the Cohen’s d test and have a positive dumping margin. Id. at 27. According to Plaintiffs, only this subset of sales is relevant to determining whether the A-to-A methodology can account for the significant price differences found, and including all other sales is unreasonable. Id.

15 . Plaintiffs also refer to the use of all sales in the meaningful differences test as unlawful, Pis.’ Br. 24, though Plaintiffs acknowledge that 19 U.S.C. § 1677f-l(b)(ii) applies only to investigations. Id. at 23. To the extent that Plaintiffs argue that the use of all sales in the meaningful differences test is unreasonable, nothing in the statute precludes Commerce from using all sales to assess whether the A-to-A methodology can account for the pattern of significant price differences found, and, further, it is reasonable for Commerce to make this determination using all sales because doing so allows Commerce to compare the results of the A-to-A and A-to-T margin calculations, as they would be applied, prior to applying the remedy.

16 . In the meaningful differences test, Commerce applies either the A-to-T methodology to all U.S. sales or to only those sales found to pass the Cohen’s d test, based on the results of the ratio test. Where the ratio test reveals that more than 33% but less than 66%■ of a respondent’s U.S. sales passed the Cohen’s d test, Commerce applies A-to-T to the sales that passed the Cohen’s d test and A-to-A to the sales that did not pass the Cohen’s d test. Final Decision Memo at 30; Prelim. Decision Memo at 7. Where the ratio test reveals that 66% or more of a respondent's U.S. sales passed the Cohen’s d test, Commerce applies A-to-T to all U.S. sales, including those not passing the Cohen’s d test. Id. Where the ratio test reveals that less than 33% of a respondent’s U.S. sales passed the Cohen’s d test, Commerce finds that the application of A-toT is not justified and applies A-to-A to all U.S. sales. See id.

17 . Plaintiffs additionally argue that ''Commerce’s rationale for using ‘all sales’ is flawed because it justified its actions[ ]based on the use of ‘higher’ and ‘lower’ priced sales, and not based on the wrongful use of [sales that passed the Cohen’s d test and sales that did not pass the Cohen’s d test].” Pis.’ Reply 12-13 (citing Final Decision Memo at 25). According to Plaintiffs, Commerce has "still not explained why it includes non-targeted sales for [the meaningful differences test].” Id at 13. However, Commerce noted specifically that it “disagree[s] with the respondents that we incorrectly included both targeted and non-targeted sales in the analysis of meaningful differences,” Final Decision Memo at 25, reiterating prior CIT decisions’ analyses regarding the use of "nonUargeted” sales and emphasizing that the statute does not specify that only "targeted” sales are to be included in the A-to-T methodology. Id.

18 . Plaintiffs argue that their data shows that, when comparing only differentially priced sales, A-to-T unmasked an additional [[]] percent of total export value for Devi Fisheries than A-to-A unmasked, Pis.’ Br. 28, and A-to-T unmasked an additional [[]] percent of total export value for Falcon Marine than A-to-A unmasked. Id. at 32.

19 . Specifically, Plaintiffs disaggregated the margins under A-to-A and A-to-T into margins for sales that exhibited patterns of significant price differences and those that did not, and then disaggregated the margins for those two subsets again into margins for sales that were dumped and not dumped. Pis.' Br. 27. Doing so isolated the targeted sales which, according to Plaintiffs, reveals that "[n]on-targeted sales did not mask any dumping on the targeted sales." Id. at 27.

20 . Plaintiffs support their argument with data demonstrating that, if the dumping margins for both methods were calculated without offsets (and using only sales that passed the Cohen’s d test), the A-to-A and A-to-T margin calculations would differ only insignificantly. Pis.’ Br. 32-35. Plaintiffs’ position is that, as the margins generated using A-to-A with offsets differed significantly from the margins generated using A-to-T without offsets, "nearly the entire difference in the margins that Commerce found ‘meaningful’ for both the respondents is attributable to the offsetting of negative margins against positive margins.” Pis.’ Br. 32.

21 . As previously stated, where less than 33% of a respondent’s U.S. sales pass the Cohen's d test, Commerce determines the significant price differences are not sufficiently pervasive to trigger application of A-to-T so A-to-A is applied to all U.S. sales, including any passing sales. Final Decision Memo at 30. Where more than 33% but less than 66% of a respondent’s U.S. sales pass the Cohen’s d test, Commerce applies A-to-T to passing sales and A-to-A to non-passing sales. Id. Finally, where 66% or more of a respondent’s U.S. sales pass the Cohen’s d test, Commerce applies A-to-T to all of the respondent’s U.S. sales, including any sales that did not pass the Cohen’s d test. Id.