OPINION AND ORDER

POGUE, Chief Judge:

This case returns to court following remand, by Grobest & I-Mei Indus. (Vietnam) Co. v. United States, 36 CIT ___, 815 F.Supp.2d 1342 (2012) ("Grobest I"), of the final results of the fourth administrative review of the antidumping duty order on certain frozen warmwater shrimp from the Socialist Republic of Vietnam.2 Specifically, Grobest I remanded the Final Results for the Department of Commerce ("Commerce" or "the Department") to (1) provide further explanation or reconsider its policy of zeroing in administrative reviews but not in investigations consistent with recent case law from the Court of Appeals for the Federal Circuit; (2) reconsider the request of Plaintiff Grobest & I-Mei Industrial (Vietnam) Co., Ltd. ("Grobest") for individual review as a voluntary respondent consistent with 19 U.S.C. § 1677m(a); and (3) accept Amanda Foods' separate-rate certification and reconsider Amanda Foods' duty rate. Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1367-68. Upon remand, in the Final Results of Redetermination Pursuant to Remand, A-552-802, ARP 08-09 (Apr. 30, 2012), Remand R. Pub. Doc. 6 ("Remand Results"), Commerce (1) provided further explanation to support its zeroing policy; (2) declined to individually review Grobest as a voluntary respondent because such review would be unduly burdensome and inhibit the timely completion of the review; and (3) accepted Amanda Foods' separate-rate certificate and assigned it the separate rate of 3.92%. Plaintiffs challenge the first and second determinations in the Remand Results. For the reasons discussed below, the court affirms the Remand Results on the first and third determinations, but remands again on the second.

The court has jurisdiction pursuant to § 516A(a)(2)(B)(iii) of the Tariff Act of 1930,3 codified, as amended, at 19 U.S.C. § 1516a(a)(2)(B)(iii) (2006) and 28 U.S.C. § 1581(c) (2006).

STANDARD OF REVIEW

"The court will sustain the Department's determination upon remand if it complies with the court's remand order, is supported by substantial evidence on the record, and is otherwise in accordance with law." Jinan Yipin Corp. v. United States, 33 CIT ___, 637 F.Supp.2d 1183, 1185 (2009) (citing 19 U.S.C. § 1516a (b)(1)(B)(i)).

DISCUSSION4

I. Commerce's Policy of Zeroing in Administrative Reviews but Not in Investigations

A. Background

In Grobest I, Plaintiffs challenged Commerce's policy of employing zeroing in administrative reviews but not in investigations.5 The court remanded the Final Results to Commerce for reconsideration and redetermination consistent with the Court of Appeals' holdings in Dongbu Steel Co. v. United States, 635 F.3d 1363 (Fed.Cir.2011) and JTEKT Corp. v. United States, 642 F.3d 1378 (Fed.Cir.2011). Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1350. In Dongbu, the Court of Appeals held that "the government has not pointed to any basis in the statute for reading 19 U.S.C. § 1677(35) differently in administrative reviews than in investigations.... In the absence of sufficient reasons for interpreting the same statutory provision inconsistently, Commerce's action is arbitrary." Dongbu, 635 F.3d at 1372-73. In JTEKT, Commerce attempted to comply with Dongbu by pointing out that different methodologies are employed in investigations and reviews. The Court of Appeals rejected Commerce's explanation as insufficient, stating:

JTEKT, 642 F.3d at 1384. The issue now before the court is whether Commerce's further explanation, as provided in the Remand Results, is sufficient to satisfy the Court of Appeals' concerns in Dongbu and JTEKT.

B. Analysis

In the Remand Results, Commerce puts forward three arguments to support its use of zeroing in reviews but not in investigations. First, Commerce argues that the courts have previously affirmed the reasonableness of Commerce's current review and investigation methodologies. Second, Commerce argues that the change of methodology in investigations was a reasonable implementation of an adverse World Trade Organization ("WTO") decision. Finally, Commerce argues that its inconsistent interpretations reasonably account for inherent differences between investigations and reviews.

Commerce contends that its first and second arguments "sufficiently justify and explain why the Department reasonably interpreted section [1677(35)] differently in average-to-average comparisons in antidumping duty investigations relative to all other contexts." Remand Results at 11. However, contrary to Commerce's assertion, in Dongbu the Court of Appeals held both of these arguments insufficient to justify the inconsistent interpretations.

The Court of Appeals made clear in Dongbu that the question before it — and therefore currently before this court — was novel: "Although we have considered Commerce's zeroing policy in administrative reviews on numerous occasions ... we agree with [plaintiff] that this court has never addressed the reasonableness of Commerce's interpretation of 19 U.S.C. § 1677(35) with respect to administrative reviews now that Commerce is no longer using a consistent interpretation." Dongbu, 635 F.3d at 1371 (citations omitted); see also Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1355-56. While the Court of Appeals has repeatedly affirmed the use of zeroing in administrative reviews, see, e.g., Timken Co. v. United States, 354 F.3d 1334, 1341-45 (Fed.Cir.2004), and approved Commerce's decision to cease zeroing in investigations, see U.S. Steel Corp. v. United States, 621 F.3d 1351, 1360-63 (Fed.Cir.2010), it has not addressed the question of inconsistent interpretations in prior cases. Therefore, that the Court of Appeals has "upheld the reasonableness of Commerce's changed methodology does not necessarily lead to the conclusion that Commerce's use of zeroing in administrative reviews remains reasonable." Dongbu, 635 F.3d at 1372 (responding to the claim that U.S. Steel Corp. stands for the proposition that the Court of Appeals has endorsed Commerce's divergent interpretations). Commerce cannot now rely on prior endorsements of its methodology when those cases did not address the relevant question before the court.

Nor does Commerce's proper implementation of an adverse WTO ruling resolve the question. Rather, recognizing that the Court of Appeals has upheld the new investigation methodology as a reasonable implementation of an adverse WTO ruling, U.S. Steel, 621 F.3d at 1360-63, Commerce must now show that its decision "compl[ies] with domestic law including reasonably interpreting statutes." Dongbu, 635 F.3d at 1372. A proper and reasonable implementation of an adverse WTO ruling may be contrary to law if it leads to an unreasonably inconsistent interpretation of statutory language. Id. ("[T]he government's decision to implement an adverse WTO report standing alone does not provide sufficient justification for the inconsistent statutory interpretations."). While implementation of an adverse WTO ruling may be a partial justification, it is not sufficient to resolve the matter. Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1357-58.

Therefore, the court turns to Commerce's third argument, that the Department's interpretations of 19 U.S.C. § 1677(35) account for inherent differences between investigations and reviews. Commerce supports this argument in two ways. First, Commerce provides a detailed explanation of the methodological differences between the average-to-average comparison in investigations and the average-to-transaction approach taken in administrative reviews. Second, Commerce explains why the goal of the investigation differs from the goal of the administrative review. Plaintiffs argue that this reasoning does not satisfy the Court of Appeals' concerns in Dongbu and JTEKT because it is not tied to the language of the statute. Comments on Final Results of Redetermination Pursuant to Court Remand at 7-9, ECF No. 116 ("Pls.' Comments").

When considering whether Commerce's interpretation is reasonable, the court looks first, but not exclusively, to the language of the statute. See Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 843 n. 9, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984); see also Wheatland Tube Co. v. United States, 495 F.3d 1355, 1361 (Fed.Cir.2007) ("To determine whether Commerce's interpretation ... is reasonable, we may look to `the express terms of the provisions at issue, the objectives of those provisions, and the objectives of the antidumping scheme as a whole.'" (quoting NSK Ltd. v. United States, 26 CIT 650, 654, 217 F.Supp.2d 1291, 1297 (2002))). But, as the court pointed out in Union Steel, this issue cannot be decided solely on the basis of the statutory language found at § 1677(35)(A). Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1356-57. The statutory language (the word "exceeds" in particular) has been repeatedly held ambiguous; furthermore, it has been held to reasonably accommodate an interpretation consistent with both zeroing and offsetting methodologies. Compare Timken, 354 F.3d at 1341-45, with U.S. Steel, 621 F.3d at 1360-63. What is at issue is not simply whether Commerce takes two different interpretations of § 1677(35)(A) in reviews and investigations; what is at issue is whether the statute can accommodate two different means to achieve two different ends. Stated differently: is it reasonable under the statutory scheme as a whole for Commerce to employ a methodology incorporating zeroing in reviews but not in investigations?

To begin answering this question, we note that the statute distinguishes between "the amount by which normal value exceeds the export price or constructed export price," i.e., the dumping margin, and the manner in which this margin becomes the basis for the antidumping duty. Section 1677(35)(A) defines a dumping margin as "the amount by which the normal value exceeds the export price or constructed export price of the subject merchandise." Pursuant to 19 U.S.C. § 1673, if Commerce finds sales of subject merchandise at less than fair value and the International Trade Commission finds injury, then "there shall be imposed upon such merchandise an antidumping duty ... in an amount equal to the amount by which the normal value exceeds the export price (or constructed export price) for the merchandise." Though a "dumping margin" and an "antidumping duty" are both measures of how much the normal value exceeds the export or constructed export price, § 1673 does not conflate the antidumping duty with the dumping margin. Rather, § 1673 contemplates a calculation methodology for arriving at the antidumping duty when it makes that duty "an amount equal to the amount by which normal value exceeds export price" (emphasis added).

However, the process by which a dumping margin becomes an antidumping duty is not provided by statute. See U.S. Steel, 621 F.3d at 1360. What the statute does provide is the basis for an antidumping duty. In some instances, the statute simply requires that the antidumping duty be based on the amount by which normal value exceeds export price. See 19 U.S.C. §§ 1673, 1673e(c)(3) ("[Commerce] shall publish notice in the Federal Register of the results of its determination of normal value and export price (or the constructed export price), and that determination shall be the basis for the assessment of antidumping duties ...."), 1675(a)(2)(C) ("The determination under this paragraph[6] shall be the basis for the assessment of ... antidumping duties ... and for deposits of estimated duties.").7 In other instances, the statute requires that an estimated antidumping duty be based on the weighted average dumping margin, which, in turn, is based on the dumping margin, see § 1677(35), also known as the amount by which normal value exceeds export price. See 19 U.S.C. § 1673b(d)(1)(B) (requiring security in an amount based on the weighted average dumping margin calculated for the preliminary determination); 19 U.S.C. § 1673d(c)(1)(B)(ii) (requiring security in an amount based on the weighted average dumping margin calculated for the final determination). Thus, pursuant to statute, Commerce must calculate the amount by which the normal value exceeds the export price, i.e., the dumping margin, and this margin becomes the basis for the antidumping duty.

Prescribing that the dumping margin be the basis of the antidumping duty does not tell Commerce how to translate a dumping margin into an antidumping duty. Commerce has reasonably interpreted the various provisions to involve a three-step calculation: (1) calculation of a dumping margin; (2) calculation of a weighted average dumping margin; and (3) application of the weighted average dumping margin to sales or entries to determine the cash deposit or antidumping duty.8 See Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1349-50, app. Therefore, the § 1677(35)(A) dumping margin serves as the basis for a § 1677(35)(B) weighted average dumping margin, and the weighted average dumping margin serves as the basis for the antidumping duty or estimated antidumping duty.9

Where, then, does zeroing come into the calculation methodology? As the Court of Appeals has held, the statute neither requires nor forbids it at any step in the process. U.S. Steel, 621 F.3d at 1361 ("[Section] 1677(35)(A) does not unambiguously preclude — or require — Commerce to use zeroing methodology."). Commerce explains in the Remand Results:

Remand Results at 12-14.

Pursuant to both methodologies, Commerce calculates the § 1677(35)(A) dumping margin by subtracting the export price from normal value for each averaging group. Once a dumping margin has been established, Commerce aggregates these dumping margins to determine a weighted average dumping margin. In an investigation, Commerce aggregates all of the dumping margins to determine "overall pricing behavior." Remand Results at 14. In a review, Commerce zeros negative margins prior to aggregation to arrive at a more accurate margin and to uncover masked dumping. Id. at 7, 14. Thus, "Commerce is not reading `exceeds' to mean two things. It is reading it as basically irrelevant to the calculation methodology, whether it expresses its view in that manner or not." Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1357.

Rather than reading "exceeds" in § 1677(35)(A) inconsistently, Commerce adopts two different methodologies for translating the dumping margin into an antidumping duty. In investigations, Commerce adopts a methodology intended to capture overall pricing behavior for the purpose of determining who should and should not fall within the purview of the antidumping duty order; for this purpose, Commerce allows negative margins to offset positive margins. In administrative reviews, however, Commerce has determined that a methodology that establishes the antidumping duty with greater accuracy is warranted both because the importer must actually pay the resulting antidumping duty and because it serves to uncover masked dumping10; for this purpose, Commerce zeroes negative dumping margins.

Commerce has offered a reasonable basis for treating investigations and reviews differently. See Nat'l Org. of Veterans' Advocates, Inc. v. Sec'y of Veterans Affairs, 260 F.3d 1365, 1379-80 (Fed.Cir. 2001) (requiring agency to provide reasonable explanation for treating the same language in two statutory provisions differently). This reasoning is based in part on the antidumping statute's objective of determining margins as accurately as possible, Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir.1990), in order to remedy the effect of unfair trading practices, Chaparral Steel Co. v. United States, 901 F.2d 1097, 1103-04 (Fed.Cir. 1990). Such is the basis behind seeking accuracy through zeroing in administrative reviews. Commerce's reasoning is also based on its decision to implement an adverse WTO decision regarding investigations, which was properly done consistent with Commerce's authority and has been upheld as reasonable by the Court of Appeals. While implementation of an adverse WTO determination is not a sufficient basis for an inconsistent interpretation of a statute, as noted above, it is a reasonable basis for the Department to reconsider the purpose, and therefore methodology, of an antidumping investigation, i.e., a reasonable basis for departing from its prior position.

Thus, the court agrees with the recent decision in Union Steel that "when it comes to reviews, which are intended to more accurately reflect commercial reality, Commerce is permitted to unmask dumping behavior in a way that is not necessary at the investigation stage." Union Steel, 36 CIT at ___, 823 F.Supp.2d at 1359. Commerce's decision to adjust its methodology to seek overall pricing behavior in investigations and more accurate duties in reviews, by zeroing in reviews but not in investigations, is a reasonable interpretation of the statute. Because Commerce has reasonably interpreted the statute, its explanation for zeroing in administrative reviews but not in investigations is affirmed.

II. Individual Review of Grobest as a Voluntary Respondent

A. Background

In the Final Results, Commerce denied a revocation request from Plaintiff Grobest because Grobest was not individually reviewed. In Grobest I, Grobest challenged this determination, arguing, inter alia, that it should have been reviewed as a voluntary respondent pursuant to 19 U.S.C. § 1677m(a).11 The court found that Commerce had failed to apply the § 1677m(a) standard when it denied Grobest's request for voluntary respondent status but had, instead, improperly applied the § 1677f-1(c)(2) standard for choosing mandatory respondents. Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1362-63. Therefore, the court remanded this challenge to Commerce to evaluate Grobest's request for voluntary respondent status consistent with the standard set forth in § 1677m(a). Id. at 1364.

B. Analysis

In the Remand Results, Commerce again rejected Grobest's request for voluntary respondent status. On remand, Commerce determined that individual review of Grobest "would have been unduly burdensome and inhibited the timely completion of the review." Remand Results at 15. Commerce first contends that the workload in this case along with the workload in other antidumping and countervailing duty cases handled by the office in charge of this case made individual review of Grobest unduly burdensome. Second, Commerce contends that having fully extended the time for the preliminary results and partially extended the time for the final results, it could not individually review Grobest without rendering completion of the administrative review untimely.

Grobest challenges the Remand Results as an unreasonable interpretation of § 1677m(a). In particular, Grobest contends that "the Department says that even one company can be a `large' number under the statute. This reading of the statute ignores [the plain language of the provision in] that the term `so large' refers to a `number' not a description of the respondents." Pls.' Comments at 16 (citations omitted). Therefore, according to Grobest, Commerce cannot refuse to individually review one or two voluntary respondents because these are not "large numbers." Id. at 16-17.

It is not the role of the court to second guess Commerce's allocation of its resources. See Longkou Haimeng Mach. Co. v. United States, 32 CIT 1142, 1151, 581 F.Supp.2d 1344, 1353 (2008). "[A]ny assessment of Commerce's operational capabilities or deadline rendering must be made by the agency itself." Id.; see also Torrington Co. v. United States, 68 F.3d 1347, 1351 (Fed.Cir.1995) ("[A]gencies with statutory enforcement responsibilities enjoy broad discretion in allocating investigative and enforcement resources."). Because allocation of resources is committed to the sound discretion of the agency, so long as Commerce complies with statutory requirements the court reviews its allocation decision for abuse of discretion, i.e., "whether Commerce's decision was `based on a consideration of the relevant factors and whether there has been a clear error of judgment.'" Torrington, 68 F.3d at 1351 (quoting Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971)).

Insofar as Grobest challenges Commerce's interpretation of the statute pursuant to the initial step in the Chevron framework, its argument is unavailing. Contrary to Grobest's assertion, the statute does not unambiguously define "a number so large." Rather, the statute conditions consideration of "a number so large" on whether review of such a number of respondents would be unduly burdensome and inhibit the timely completion of the review. The fact that the statute sets out a standard for interpreting "a number so large" means that there is no definitive number contemplated under the statute. Rather, "number" and "large" are ambiguous statutory terms; thus, the court must consider whether Commerce's interpretation is reasonable. Chevron, 467 U.S. at 842-45, 104 S.Ct. 2778.

In the Remand Results, Commerce chose to examine the burden imposed and the effect upon the timeliness of the review posed by individual investigation of one voluntary respondent. In doing so, Commerce interpreted the statute not to set a floor for the number of voluntary respondents to be reviewed; rather, Commerce interpreted the statute to render every voluntary respondent request subject to an undue burden and timely completion analysis. Such interpretation is reasonable as it contemplates the standard set forth by the statute itself — whether the number of respondents is so large as to be unduly burdensome and inhibit the timely completion of the review. Cf. F.C.C. v. AT & T Inc., ___ U.S. ___, 131 S.Ct. 1177, 1184, 179 L.Ed.2d 132 (2011) ("[S]tatutory interpretation turns on the language itself, the specific context in which that language is used, and the broader context of the statute as a whole." (quoting Nken v. Holder, 556 U.S. 418, 426, 129 S.Ct. 1749, 173 L.Ed.2d 550 (2009) (internal quotation marks omitted))). In contrast, Grobest's argument that the number of voluntary respondents must reach some arbitrary threshold of largeness fails to consider the relative burdens that may be imposed by review of any one respondent, thereby disregarding the statutorily established standard.

Having considered whether Commerce's statutory interpretation is reasonable, we must now consider whether Commerce's decision amounted to an abuse of discretion. The Court of Appeals has stated that "an agency abuses its discretion where its `decision (1) is clearly unreasonable, arbitrary, or fanciful; (2) is based on an erroneous conclusion of law; (3) rests on clearly erroneous fact findings; or (4) follows from a record that contains no evidence on which the [agency] could rationally base its decision.'" Sterling Fed. Sys., Inc. v. Goldin, 16 F.3d 1177, 1182 (Fed.Cir.1994) (quoting Gerritsen v. Shirai, 979 F.2d 1524, 1529 (Fed.Cir. 1992)); see also Robert Bosch LLC v. Pylon Mfg. Corp., 659 F.3d 1142, 1148-49 (Fed.Cir. 2011) (noting that a clear error of judgment occurs when an action is "arbitrary, fanciful or clearly unreasonable"). When the record cannot support the agency's determination or the agency's exercise of discretion exceeds the limits of the statute, the agency has abused its discretion.

As we explained in Grobest I, § 1677m(a) sets a heightened standard that Commerce must meet when denying individual review to a voluntary respondent. See Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1363. That heightened standard anticipates voluntary respondents placing some burden on the agency and requires that the voluntary respondents receive an individual review in those circumstances. Such an expectation was contemplated by Congress when it noted in the Statement of Administrative Action for the Uruguay Rounds Agreements Act that "Commerce, consistent with Article 6.10.2 of the Agreement, will not discourage voluntary responses and will endeavor to investigate all firms that voluntarily provide timely responses in the form required...." Uruguay Round Agreements Act, Statement of Administrative Action, H.R. Doc. No. 103-316, vol. 1, at 873 (1994), reprinted in 1994 U.S.C.C.A.N. 4040, 4201 ("SAA"). If this were not the case, then no analysis beyond that required in § 1677f-1(c)(2) would be required. Cf. Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1362 ("To limit voluntary respondents through § 1677f-1(c)(2) is to foreclose review under § 1677m(a) barring the unexpected and irregular."). We held in Grobest I that the agency cannot make a voluntary respondent decision based on its analysis pursuant to § 1677f-1(c)(2); likewise, the agency cannot draw its § 1677m(a) analysis so narrowly that it mirrors the analysis under § 1677f-1(c)(2). It is only when the burden becomes undue that Commerce may decline to individually review voluntary respondents.

In this case Commerce has failed to show undue burden and has exercised its discretion in a way that renders the statute meaningless. On remand, Commerce found individual review of one voluntary respondent to be unduly burdensome. However, the facts that Commerce put forward to support that conclusion do not distinguish this case from the paradigmatic review of an antidumping or countervailing duty order.12 Rather, the burdens Commerce names in the Remand Results are the same burdens that occur in every review. In this regard, Commerce's decision that the burden in this case is undue sets the bar for undue burden too low because it would make individual review of voluntary respondents in any typical antidumping or countervailing duty review unduly burdensome, and such a determination renders § 1677m(a) meaningless. Cf. Grobest I, 36 CIT at ___, 815 F.Supp.2d at 1363; see also SAA, vol. 1 at 873, 1994 U.S.C.C.A.N. at 4201.

When Commerce can show that the burden of reviewing a voluntary respondent would exceed that presented in the typical antidumping or countervailing duty review, the court will not second guess Commerce's decision on how to allocate its resources. See Longkou Haimeng Mach., 32 CIT at 1151, 581 F.Supp.2d at 1353. However, Commerce's failure to make such a showing in this case, thereby rendering § 1677m(a) meaningless, is an abuse of discretion. Therefore, this case is remanded to Commerce to individually review Grobest as a voluntary respondent and, if appropriate in light of the review, to consider Grobest's request for revocation.

CONCLUSION

In light of the foregoing opinion, the Final Results, 75 Fed.Reg. at 47,771, as explained by the Remand Results, are affirmed in part and remanded in part. Commerce's explanation for why it continued to zero in this review after ceasing zeroing in investigations is affirmed. Commerce's assignment of the separate rate to Amanda Foods is also affirmed. Commerce's rejection of Grobest's request for voluntary respondent status is remanded to Commerce to conduct an individual review of Grobestas a voluntary respondent and to reconsider Grobest's revocation request in light of the results of that review.

Commerce shall have until September 14, 2012, to complete and file its remand redetermination. Plaintiffs and Defendant-Intervenors shall have until September 28, 2012, to file comments. Plaintiffs, Defendant, and Defendant-Intervenors shall have until October 8, 2012, to file any reply.

It is SO ORDERED.