Tianjin MacHinery Import & Export Corp. v. United States

752 F. Supp. 2d 1336, 33 I.T.R.D. (BNA) 1020, 2011 Ct. Intl. Trade LEXIS 5, 2011 WL 248542

• Court: United States Court of International Trade
• Decided: January 4, 2011
• Docket: Slip Op. 11-1; Court 05-00522
• Status: Published

Opinion

OPINION AND ORDER

EATON, Judge.

At issue in this action are the United States Department of Commerce’s (“Commerce” or the “Department”) final results of the thirteenth administrative review of four antidumping duty orders applicable to imports into the United States of heavy forged hand tools (“HFHTs”) from the People’s Republic of China (“PRC”). See HFHTs, Finished or Unfinished, With or Without Handles, From the PRC, 70 Fed. Reg. 54,897 (Dep’t of Commerce Sept. 19, 2005) (final Results of antidumping duty administrative reviews) (the “Final Results”); see also HFHTs, Finished or Unfinished, With or Without Handles From the PRC, 56 Fed.Reg. 6,622 (Dep’t of Commerce Feb. 19, 1991) (notice of anti-dumping duty orders) (covering axes/adzes, bars/wedges, hammers/sledges and picks/mattocks) (the “HFHTs Orders”). In Tianjin Machinery Import & Export Corp. v. United States, 31 CIT 1416, 2007 WL 2701368 (2007) (not reported in the Federal Supplement) (“Tianjin I”), the court sustained certain aspects of the Final Results and remanded several issues to the Department.

Subsequently, the court also ordered the Department to reopen the record as to one of those issues and to address the merits of plaintiffs’ chart submissions, which purportedly demonstrated “that TMC and Huarong’s steel type and the steel surrogate values differed in recent reviews ... and ... that these differences significantly impacted the resulting margins relied upon by Commerce in its AFA analysis.” Tianjin Machinery Import & Export Corp. v. United States, Ct. No. 05-00522, Order at 4 (June 8, 2009) (directing Commerce to place plaintiffs’ charts on the record and to offer a response to them) (“Tianjin II”).

Now before the court are Commerce’s Final Results of Redetermination as supplemented. See Final Results of Redetermination (Dep’t of Commerce Mar. 11, 2008) (the “First Remand Results”); Final Results of Redetermination Pursuant to Court Order (Dep’t of Commerce Sept. 16, 2009) (the “Supplemental Remand Results”) (collectively, the “Remand Results”). The court has jurisdiction pursuant to 28 U.S.C. § 1581(c) (2006) and 19 U.S.C. § 1516a(a)(2)(B)(iii). For the following reasons, the Remand Results are sustained, in part, and remanded.

STANDARD OF REVIEW

“The court shall hold unlawful 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)(l)(B)(I).

DISCUSSION

I. Legal Framework for Adverse Facts Available Rates

When periodically reviewing antidumping duties for an uncooperative party, Commerce “may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available.” 19 U.S.C. § 1677e(b). While the court has previously sustained 1 Commerce’s threshold decision to use Adverse Facts Available 2 (“AFA”), questions remain about the manner in which the adverse inferences, arising from that decision, were used “in selecting from among the facts otherwise available.” 19 U.S.C. § 1677e(b). In particular, plaintiffs contest the Department’s application of adverse inferences when choosing from the facts otherwise available to select their individual AFA rates.

II. AFA Rate for TMC’s and Huarong’s Sales of Bars/Wedges

In Tianjin I, the court found that Commerce did not present substantial evidence to justify its decision to assign an AFA rate of 139.31 percent to Tianjin Machinery Import & Export Corp.’s (“TMC”) and Shandong Huarong Machinery Co., Ltd.’s (“Huarong” and, collectively, “plaintiffs”) for their sales of bars/wedges. This rate had been calculated for TMC in the eighth administrative review of the HFHTs Orders and is the highest calculated rate under the Orders. See Tianjin I, 31 CIT at 1434, 2007 WL 2701368. On remand, the court directed the Department to either “explain ... how the 139.31 percent rate applied to TMC’s and Huarong’s sales of bars/wedges is a reasonably accurate estimate of TMC’s actual rate with a built-in increase to deter non-compliance ... and ... [to] explain in detail how any rate assigned to Huarong is reliable and bears a rational relationship to the company itself’ or “reopen the record and calculate an AFA rate to be applied to Huarong’s and TMC’s sales of bars/wedges, with an additional amount to deter future non-compliance.” Id. at 1435, 2007 WL 2701368.

In the First Remand Results, Commerce did not change the 139.31 percent rate for TMC’s and Huarong’s sales of bars/wedges, nor did it formally reopen the record, as permitted by the court, to calculate new AFA rates. Instead, the Department insisted that it complied with the court’s remand instructions by further explaining the applicability of the 139.31 percent rate to TMC and Huarong, and providing additional factual support for this determination. See First Remand Results at 2, 6.

Responding to the First Remand Results, plaintiffs claimed that Commerce’s new justifications were based on information not on the record. Tianjin II, Ct. No. 05-00522, at 4. In making their argument, plaintiffs primarily objected to Commerce’s reliance on newly produced data from United States Customs and Border Protection. Id. at 3-4. As a result, they sought an opportunity to put on the record information of their own, consisting of several charts purportedly demonstrating “that TMC[’s] and Huarong’s steel type and the steel surrogate values differed in recent reviews! ] ... and ... these differences significantly impacted the resulting margins relied upon by Commerce in its AFA analysis.” Id. at 4.

The court agreed with plaintiffs that Commerce had “independently accessed Customs data to substantiate its results.” Id. at 5. Likewise, the court also agreed that the plaintiffs’ charts were “relevant to the probative value of Commerce’s weighted-average unit value analysis” and that the information “should be fully vetted by the Department at the administrative level.” Id. at 5-6. As a result, the court remanded the matter again, and ordered Commerce to “address the impact of plaintiffs’ charts ... to their selection of AFA rates for TMCf’s] and Huarong’s sales of bars/wedges.” Id. at 6.

In the Supplemental Remand Results, Commerce retained its initial rates, finding that plaintiffs’ charts and arguments did not “provide!] information or argument sufficient to change the Department’s selection of 139.31 percent rate for either TMC’s or Huarong’s sales of bars/wedges.” Supplemental Remand Results at 2.

A. TMC

In Tianjin I, the court concluded that the 139.31 percent rate did not bear a rational relationship to TMC’s sales of bars/wedges during the period of review (“POR”), i.e., February 1, 2003 to January 31, 2004. Tianjin I, 31 CIT at 1434-35, 2007 WL 2701368. In reaching this conclusion, the court found that Commerce could not rely on TMC’s rate from the eighth administrative review of the HFHTs Orders (covering the POR February 1, 1998 to January 31, 1999) without explaining its relevance to the thirteenth administrative review, which took place some five years later.

On remand, Commerce has chosen to comply with Tianjin I’s directives by seeking “additional information to determine whether TMC’s sales during the eighth administrative review continue[ ] to reflect TMC’s commercial activity during the [POR] for the thirteenth review.” See Def.’s Resp. to Comments Regarding the Remand Redetermination 5 (citing First Remand Results at 2). In order to demonstrate how the 139.31 percent rate is reflective of TMC’s recent commercial activity, Commerce cites the following additional factual support:

The Department obtained information from the Automated Commercial System (ACS) of U.S. Customs and Border Protection (CBP) regarding sales values of TMC’s merchandise classifiable under harmonized tariff schedule subheading 8205.59.30, the subheading applicable to the merchandise subject to the bars/wedges order. The Department specifically queried the two review periods at issue: February 1, 1998, through January 31, 1999 [the eighth review], and February 1, 2003, through January 31, 2004 [the thirteenth review]. Using this information, the Department calculated a weighted-average unit value (AUV) for each period for TMC’s sales of merchandise subject to the bars/wedges order. The Department compared the AUV from each period and found that TMC’s AUV for the subject merchandise declined ... from the earlier to the later period. This change in TMC’s AUVs contrast^] with little to no change in the production process used by the PRC industry to produce bars/wedges over the last five years, as demonstrated by respondent questionnaire responses and verifications from multiple administrative proceedings. Thus, because the production process of the industry has generally stayed constant, while TMC’s U.S. sales values have declined, the Department concludes that this information further substantiates the relevance of the 139.31 percent margin as AFA for TMC’s sales of merchandise under the bars/wedges order.

First Remand Results at 6-7 (citations omitted). Thus, Commerce insists that a steep decline 3 in the price at which the merchandise was sold into the United States, while the production process used to make the bars/wedges stayed constant, is “probative evidence of TMC’s continued dumping of hand tools on the United States market.” Def.’s Resp. To Court’s Letter Dated June 11, 2010 3 (emphasis removed). In other words, for Commerce the price decline from the eighth review to the thirteenth review, during which period nothing changed in the way the merchandise was made, indicates that, not only did TMC continue to dump its products, but that the dumping margin, if anything, had increased.

Next, the Department points to, what it calls, the volatility of TMC’s margins in past reviews as further evidence of the relevance of the 139.31 percent rate. Specifically, Commerce notes that: (1) in the seventh review, it assigned TMC an AFA rate of 47.88 percent; (2) in the eighth review, it calculated the 139.31 percent rate (an increase of ninety-two percentage points); (3) in the ninth review, it calculated a 0.56 percent rate (a 248-fold decrease); (4) in the tenth review, it calculated a 0.48 percent rate (a de minimus change from the ninth review); and, (5) in the twelfth review, sustained by the court in Shandong Huarong Machinery Co., Ltd. v. United States, 31 CIT 1815, 1820 (2007) (not reported in the Federal Supplement) (“Shandong ”), Commerce, using AFA, assigned the 139.31 percent rate calculated in the eighth review. See First Remand Results at 7-8. Accordingly, for Commerce, the “wide swings” experienced by TMC in its prior administrative reviews of bars/wedges demonstrates the continued reliability of the 139.31 percent rate. Id. at 8; see also id. at 7 (“[A] review of the volatility of TMC’s margins in past reviews provides further factual support for the relevance of the 139.31 percent rate.”). Thus, Commerce asserts that the 139.31 percent rate is consistent with TMC’s varying rate history.

As to why the 139.31 percent rate is more accurate than more recent lower rates calculated for TMC, the Department argues:

Because TMC’s sales of bars/wedges are receiving total adverse facts available 4 due to its involvement in the “agent” sales scheme, we do not find that using one of TMC’s prior lower rates would be appropriate as an AFA rate. In administrative reviews where TMC received lower rates, not only did TMC fully participate in those proceedings, but TMC did not receive total AFA. As a result, it would be difficult to reconcile applying a lower rate to TMC as AFA in this review when those lower rates were not calculated based on AFA.

First Remand Results at 9.

With respect to the steel valuation charts plaintiffs have now placed on the record, Commerce maintains that it “cannot compare the steel used for TMC or Huarong in the [thirteenth] administrative review with the [eighth] administrative review” because the Department believes, based on the companies’ participation in the fraudulent invoicing scheme that the new data, like all of the companies’ questionnaire responses, is not reliable. Supplemental Remand Results at 8. “As such, the Department has no reliable information upon which to conclude that the steel surrogate values would be different — or how they would be different — than the surrogate values applied in the 8th administrative review.” Id. Put another way, because of the plaintiffs’ failure to disclose their true business relationship, 5 Commerce has concluded that it cannot trust any of the steel valuation data they have offered.

Further, Commerce contends that, even if it did consider the explanatory charts, they would not be sufficient to explain why the margin for the thirteenth review would not be the same as or higher than the eighth. Id. In making this argument, the Department stresses that the dumping margin is not derived solely on the surrogate steel value, but is based on several factors of production. See id. at 9 (“Plaintiffs’ argument examines only one component of the normal value.... Without considering the differences in consumption rates for all [factors of production] (e.g., material inputs, energy and labor) between segments, we cannot assume, as plaintiffs argue, that the difference in margins would simply be due to a change in the surrogate values for steel.”). Commerce also notes the margin is sensitive to the U.S. price. See id. at 8-9 (“While plaintiffs seem to claim that the surrogate values that would have been used in the 13th administrative review would have been lower than the surrogate value for steel used in the 8th administrative review, U.S. prices at least for TMC’s bars/wedges also declined.”). Thus, even if normal value did decrease between the eighth and thirteenth reviews, Commerce argues that the margin would not necessarily be smaller, because the U.S. price also declined.

Finally, the Department claims that plaintiffs’ charts themselves undermine their theory of a direct correlation between steel values and margins. Specifically regarding TMC, Commerce notes that the company’s margin declined at a much steeper rate than its steel surrogate value did from the eighth to the ninth to the tenth administrative review. See id. at 10 (“While the surrogate value of steel declined from the 8th administrative review value of 37.52 Rs/kg to a range of values between 5.43 Rs/kg and 9.327 Rs/ kg, the [calculated] margins declined at a much steeper rate: from [a calculated rate of] 139.31 percent in the 8th administrative review to [calculated rates of] 0.56 percent and 0.48 percent in the 9th and 10th administrative reviews, respectively.”). As a result, the Department contends that something other than the surrogate values for steel alone influenced the size of TMC’s margin.

In support of its challenge to the Remand Results, TMC states that the Department’s 139.31 percent rate is unreliable because Commerce “merely selected data to support that conclusion without meaningful and required corroboration.” Comments on the Department of Commerce’s Final Results of Redetermination Pursuant to Court Remand (“Pis.’ Supplemental Comments”) 5. Central to TMC’s various claims is its contention that Commerce refuses to acknowledge “how important it is to match the input steel surrogate values with reasonable margins.” Pis.’ Supplemental Comments 10.

TMC relies heavily on its contention that the steel surrogate value chosen for the eighth administrative review was “aberrational in comparison to all other input steel surrogate values.” Pis.’ Supplemental Comments 8. In other words, the company claims that the steel surrogate value used for the eighth review was markedly more expensive than in other years, because in the eighth review Commerce chose a specific Indian surrogate steel value it had not used before, and that it has not used since. According to TMC, this surrogate value choice resulted in a much higher antidumping margin in the eighth review than in subsequent reviews. To support their argument, plaintiffs present the two charts, described supra, purporting to demonstrate “that the calculated margins for TMC and Huarong during the [eighth through thirteenth reviews] varied in direct proportion to the steel input surrogate value.” Pis.’ Supplemental Comments 9.

Commerce’s selection of the 139.31 percent rate for TMC’s sales of bars/wedges during the POR is sustained. In reaching this decision, the court notes that many of the arguments made in this case were made in the earlier Shandong case, where this Court sustained the 139.31 percent rate for TMC’s sales of bars/wedges in the twelfth administrative review. In that case, the Court stressed that all of TMC’s proffered sales data were tainted by the company’s participation in the fraudulent invoicing scheme. See Shandong, 31 CIT at 1819-20 (“Because of TMC’s participation in the invoicing scheme, all of its sales data was necessarily tainted.... Ac cordingly, despite plaintiffs’ insistence that Commerce should calculate a rate for TMC’s bars/wedges, Commerce had no reliable information from which to do so.”).

As an initial matter, Commerce’s AUV analysis tends to support its contention that “TMC’s declining United States sales values (without commensurate production process changes) constituted probative evidence of TMC’s continued dumping of hand tools on the United States market.” Def.’s Resp. to Court’s Letter Dated June 11, 2010 3 (emphasis removed). As the Department explains:

A direct relationship exists between an industry’s production process and the subject merchandise’s cost of production because significant changes in the production process, such as the introduction of new technology, can affect the cost of production. Hypothetically, if the cost of production decreases because of new technologies or efficiencies, such cost decreases could be reflected in declining United States sales values. If no such changes occurred, however, this would be some indication that costs may have remained steady, corroborating, to the extent practicable, the selected facts available dumping margin because declining sales values reflected continued dumping. Accordingly, Commerce in this case examined whether changes in the industry’s production process might have explained some of the decline in U.S. sales values, and found that there were no significant changes in the production process of the industry. Therefore, Commerce reasonably determined that any change in the cost of production for the subject merchandise and the United States sales values was not attributable to changes in the production process of the industry.

Def.’s Resp. to Court’s Letter Dated June 11, 2010 1-2. While the analysis does not take into account any changes in the surrogate cost of manufacturing inputs, it is, as Commerce puts it, “some indication” that TMC’s manufacturing costs remained steady while its U.S. price declined, thus eliminating production changes as a cause of the change in price.

Moreover, TMC has a history of significant rate changes among reviews. That is, the company has received calculated rates that vary from 139.31 percent in the eighth, to 0.48 percent in the tenth review. Thus, while TMC’s more recent rates have been dramatically lower, the calculated rate of 139.31 percent tends to confirm the Department’s finding that assigning this rate in this thirteenth review is not inconsistent with TMC’s calculated rate history. See Shandong, 31 CIT at 1820.

TMC’s primary objection to Commerce’s analysis is that the Department has failed to recognize the significance of differences, across reviews, of the steel surrogate value to the resulting margin. At the outset, it must be noted that Commerce’s claim that it can ignore plaintiffs charts based on the use of total AFA overstates the uses of 19 U.S.C. § 1677e(b). See, e.g., Gerber Food (Yunnan) Co., Ltd. v. United States, 29 CIT 753, 771, 387 F.Supp.2d 1270, 1286-87 (2005) (“Commerce failed to provide a rational explanation of how Green Fresh’s participation in the export agency agreement, and the circumstances surrounding its reporting of that agreement, affected the unrelated information needed to calculate an antidumping duty rate for application to all shipments by Green Fresh of mushrooms subject to the administrative review.”). Put another way, the use of AFA resulting from the fraudulent invoicing scheme may taint TMC’s pricing information, but cannot be used as an excuse to disregard unrelated information. See 19 U.S.C. § 1677e(b).

TMC’s arguments about the differences in the type of steel used as an input in various reviews and hence its cost, however, are not persuasive. The data found in plaintiffs’ charts only considers one component of the normal value calculation, i.e., the input for steel. Without looking at all the factors of production together, e.g., the cost of labor, the court has no way to discover if the reduction in the cost of the steel surrogate value would have reduced TMC’s margin. In addition, while the cost of producing the merchandise may have decreased from review to review, so did the U.S. price. Thus, it is not at all clear that TMC’s margin between the eighth and the thirteenth review would have decreased based solely on a reduction in the surrogate value of the steel inputs. 6

Next, an examination of TMC’s charts reveals that they do not show a direct correlation between the steel surrogate value and antidumping margin. According to the charts, TMC’s margin decline was much more exaggerated than the decline in its steel surrogate values between the eighth, ninth, and tenth reviews. Therefore, the data suggests that factors other than the cost of the steel input were at work and resulted in declining prices.

Here, the Department has chosen a rate calculated for TMC in a prior review. Reliance on a previously calculated rate for the company itself was reasonable because, although the Department had no reliable pricing information for the POR, it was able to look to a rate arrived at using TMC’s own verified data in an earlier review. See Shandong, 31 CIT at 1820. The Department then took steps to demonstrate that the data remained relevant to the “commercial reality” of TMC during the thirteenth administrative review. Gallant Ocean (Thailand) Co., Ltd. v. United States, 602 F.3d 1319, 1324 (Fed.Cir.2010) (“Gallant Ocean”); see also KYD, Inc. v. United States, 607 F.3d 760, 767 (Fed.Cir. 2010) (reaffirming Gallant Ocean’s “commercial reality” test) (“KYD ”).

Specifically, Commerce (1) engaged in an AUY analysis to demonstrate that a rate calculated using TMC’s data from the thirteenth review could have been as high or higher than 139.31 percent, and (2) employed a volatility analysis to show that TMC’s rates have varied greatly in past reviews, and that the 139.31 percent rate, though high, is consistent with TMC’s rate history. In the AFA context, Commerce is permitted to add a “built-in increase intended as a deterrent to non-compliance,” as long as that increase does not become “punitive, aberrational, or uncorroborated.” F.lli De Ceceo Di Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000) (“De Ceceo ”). By using a previously calculated rate for TMC itself and justifying the rate with substantial evidence, it is apparent that Commerce has not selected an “unreasonably high rate[ ] having no relationship to the respondent’s actual dumping margin.” Gallant Ocean, 602 F.3d at 1323. Thus, Commerce has satisfied the instructions in Tianjin I to “explain ... how the 139.31 percent rate applied to TMC’s ... sales of bars/wedges is a reasonably accurate estimate of TMC’s actual rate with a built-in increase to deter noncompliance .... ” Tianjin I, 31 CIT at 1435, 2007 WL 2701368. While hardly overwhelming, taken as a whole, the Department has martialed sufficient substantial evidence to support the rate. See Shandong Huarong Gen. Corp. v. United States, 25 CIT 834, 842, 159 F.Supp.2d 714, 722-23 (2001). Accordingly, the court finds that Commerce’s weighing of the evidence and deriving inferences therefrom was not unreasonable and sustains the Department’s application of the 139.31 percent rate to TMC’s sales of bars/wedges in this thirteenth administrative review.

B. Huarong

In Tianjin I, the court found that in applying TMC’s 139.31 percent rate from the eighth review to Huarong, the Department did not “articúlate! ] how the 139.31 percent rate is a reasonable estimate of what Huarong’s rate would have been had it complied together with a built-in increase as a deterrent,” and, therefore, directed Commerce to “explain in detail how any rate assigned to Huarong is reliable and bears a rational relationship to the company itself.” 31 CIT at 1435, 2007 WL 2701368.

On remand, the Department insists that the volatility of Huarong’s margins in past reviews provides factual support for the relevance of the 139.31 percent rate to the company. First Remand Results at 11; Supplemental Remand Results at 11.

Specifically, the Remand Results state that: (1) in the sixth review, Huarong received a calculated rate of 34.00 percent; (2) in the seventh review, a calculated rate of 1.27 percent; (3) in the eighth review, a calculated rate of 27.28 percent; (4) in the ninth review, an assigned AFA rate of 47.88 percent; (5) in the tenth review, a calculated rate of 18.99 percent; (6) in the eleventh review, a calculated rate of 30.02 percent; and, (7) in the twelfth review, Huarong was assigned the 139.31 percent AFA rate. 7 See First Remand Results at 11. For Commerce, these “wide swings” justified the continued relevance of the 139.31 percent rate to Huarong. First Remand Results at 11. The Department maintains this position even though Huarong has never had a calculated rate higher than 30.02 percent.

In addition, the Department examined Huarong’s sales data from the eleventh review (in which the company cooperated and received its highest calculated rate of 30.02 percent) in order to demonstrate the relevance of the 139.31 percent to Huarong’s sales during this POR. First Remand Results at 11-12; Supplement Remand Results at 11. In doing so, the Department noted that Huarong had some transaction-specific margins in the eleventh review that were nearly as high as 139.31 percent and claims that those transactions did “not appear to be aberrant or unusual in any way.” First Remand Results at 12. 8 For the Department, these transactions were “representative of the margins [it] would have calculated for ... [Huarong] in the thirteenth review (with a built-in incentive to encourage cooperation) had it not received total AFA.” First Remand Results at 12. In other words, the Department claims that the 139.31 percent rate is shown to be relevant to Huarong’s commercial activity during the POR by looking only at the high end of a selected group of sales for a time period two years removed from the POR. For Commerce, this analysis was significant because it not only relied upon Huarong’s relatively recent data, but also served to support a rate high enough to deter future non-cooperation.

Finally, the Department asserts “that if an uncooperative respondent could have demonstrated that its dumping margin is lower than the highest prior margin[,] it would have provided information showing the margin to be less.” First Remand Results at 12-13 (citation omitted). To justify this approach, the Department relies on the “common sense” presumption first confirmed by the Federal Circuit in Rhone Poulenc, Inc. v. United States, 899 F.2d 1185 (Fed.Cir.1990) (“Rhone Poulenc ”). See First Remand Results at 12-13; Rhone Poulenc, 899 F.2d at 1190.

Huarong’s arguments in response to Commerce’s conclusions concerning Huarong are not entirely dissimilar to those made by TMC. Thus, the company asserts that there is a direct correlation between the surrogate steel values from its various reviews and its calculated antidumping margins. Pis.’ Supplemental Comments 9. Moreover, Huarong insists that Commerce did not comply with the court’s remand instruction to explain how TMC’s prior rate is rationally related to Huarong. The company argues that the Department’s analysis was “inadequate and selective,” and that Commerce “improperly used TMC’s prior rate as a benchmark for Huarong.” Comments on the Dept, of Commerce’s Final Results of Redetermination Pursuant to Court Remand (“Pis.’ First Comments”) 25. Additionally, Huarong claims that Commerce’s volatility analysis does not provide substantial evidence that Huarong should receive a 109.29 percentage point increase from its last calculated rate in the eleventh review. Pis.’ First Comments 28. It also notes that this Court invalidated the 139.31 percent AFA rate for Huarong in the ninth review. Pis.’ Supplemental Comments 14; see also Shandong Huarong Gen. Grp. Corp. v. United States, 29 CIT 1227, 1237 (2005) (not reported in the Federal Supplement) (“Huarong ”) (characterizing the use of the 139.31 percent rate in the ninth review as “punitive in nature” based on Commerce’s “strained efforts to demonstrate [its] validity”).

Although the court finds Huarong’s arguments concerning the correlation between steel surrogate values and its anti-dumping margins unconvincing for the same reasons set out in the TMC discussion supra, the question of Huarong’s rate must again be remanded. First, the court is not persuaded by the Department’s citation of Rhone Poulenc for the idea that the use of AFA dispenses with the Department’s statutory mandate to corroborate secondary information. See 19 U.S.C. § 1677e(e). The Rhone Poulenc case is most often cited for its statement on the assignment of the highest prior margin to an uncooperative respondent: “[I]t reflects a common sense inference that the highest prior margin is the most probative evidence of current margins because, if it were not so, the importer, knowing of the rule, would have produced current information showing the margin to be less.” Rhone Poulenc, 899 F.2d at 1190. In other words, the case stands for the proposition that a respondent can be assumed to make a rational decision to either respond or not respond to Commerce’s questionnaires, based on which choice will result in the lower rate.

It is important to keep in mind, however, that Rhone Poulenc was a pre-Uruguay Round Agreements case. As a result, it reflected the state of the law prior to the enactment of the Act 9 that implemented the Agreements’ negotiated terms. See generally Gerber Food (Yunnan) Co., Ltd. v. United States, 31 CIT 921, 947, 491 F.Supp.2d 1326, 1351 (2007) (“Rhone Poulenc ... involved the application of the ‘best information available’ standard under 19 U.S.C. § 1677e prior to the substantial amendment of that statutory provision by the Uruguay Round Agreements Act....”). As part of the implementing legislation, however, Congress directed Commerce to make additional findings in AFA cases. See 19 U.S.C. § 1677e(c) (“When the administering authority or the Commission relies on secondary information rather than on information obtained in the course of an investigation or review, the administering authority or the Commission, as the case may be, shall, to the extent practicable, corroborate that information from independent sources that are reasonably at their disposal.”). It appears that Commerce is now trying to dispense with this corroboration requirement by employing the Rhone Poulenc presumption. Because the case predated the Uruguay Round Agreements Act additions to § 1677e, however, it necessarily did not hold that the presumption could replace actual corroboration.

In addition, in the most recent cases where the presumption is mentioned, the Federal Circuit appears to restrict its use to situations where a respondent has not answered Commerce’s questionnaire at all, rather than when the questionnaire responses were found wanting for one reason or another. In fact, in the most recent case citing the Rhone Poulenc presumption, the Federal Circuit paid particular attention to the fact that the exporter put nothing on the record. See KYD, 607 F.3d at 764 (“King Pac had elected not to cooperate at all in the review.”); see also id. at 767 (“King Pac’s failure to cooperate deprived Commerce of the most direct evidence of King Pac’s actual dumping margin.”). Thus, the KYD case seems to confirm that “common sense” restricts the Rhone Poulenc presumption to cases where a respondent can be assumed to have chosen not to respond to a questionnaire at all, in order to achieve a lower rate.

Furthermore, the Federal Circuit recently analyzed the outer limits of Commerce’s discretion in assigning AFA rates. In Gallant Ocean, the plaintiff was a Thai exporter of shrimp that refused to participate in an administrative review. Gallant Ocean, 602 F.3d at 1321-22. Having no company-specific information to use, Commerce assigned Gallant a 57.64 percent AFA rate, based on the petition rate. Commerce claimed that the rate was corroborated by transaction-specific margins for three other cooperating exporters. Id. at 1322. The rate, however, was ten times higher than the average rate for all cooperating respondents and five times higher than the highest rate calculated for any cooperating respondent. Id. at 1324. The Federal Circuit rejected Commerce’s chosen rate because “nothing in the record tie[d] the ... rate to Gallant,” making the rate “unrelated to [Gallant’s] commercial activity.” Id. In addition, the Gallant Ocean Court characterized the rate as “ ‘punitive, aberrational, or uncorroborated’” because of the large disparity between the assigned rate and the cooperative respondents’ calculated dumping rates. Id. (quoting De Ceceo, 216 F.3d at 1032). On remand, the Court instructed Commerce that it “must select secondary information [i.e., a rate] that has some grounding in [Gallant’s] commercial reality.” Id.

In the subsequent KYD case, the Federal Circuit reaffirmed the “commercial reality” requirement. See KYD, 607 F.3d at 767 (“Moreover, in Gallant Ocean this court concluded that the AFA margin that Commerce selected ‘did not and does not represent commercial reality’ and ruled that Commerce ‘may not use the petition rate to establish the dumping margin when its own investigation revealed that the petition rate was not credible.’ ”). Thus, KYD and Gallant Ocean, taken together, confirm that an AFA rate must be shown by substantial evidence to have a rational relationship to the “commercial reality” of the respondent during the POR.

Other Federal Circuit AFA decisions bolster the conclusion that Commerce must establish the relevance of a chosen rate to the respondent. In De Ceceo, the court stressed that Commerce’s discretion is “not unbounded” and that “Congress could not have intended for Commerce’s discretion to include the ability to select unreasonably high rates with no relationship to the respondent’s actual dumping margin.” 216 F.3d at 1032. In particular, the De Ceceo Court noted that “the statute has no requirement that Commerce is limited to the highest rate imposed on a cooperating company when selecting a rate for a non-cooperating respondent.” Id.

The Federal Circuit did uphold an AFA rate corroborated by a small percentage of a company’s transactions in PAM, S.p.A. v. United States, 582 F.3d 1336, 1340 (Fed. Cir.2009) {“PAM”). The Court, however, discussed PAM in Gallant Ocean, noting that “[substantial evidence [still] requires Commerce to show some relationship between the AFA rate and the actual dumping margin” of the company being reviewed. Gallant Ocean, 602 F.3d at 1325. The Gallant Ocean Court emphasized that an analysis of selected transactions was not sufficient to constitute substantial evidence of what a calculated rate would be when “a large body of reliable information suggests] the application of a much lower margin.” Id.; see also id. at 1324 (“Commerce used a very small percentage of the mandatory respondents’ transactions as corroborative evidence even though most transactions during the period of review had significantly lower dumping margins.”).

Here, Commerce’s justifications recall the Department’s reasoning in the litigation following the final determination in the ninth administrative review, where the Department also selected the 139.31 percent AFA rate for Huarong. Huarong, 29 CIT at 1230. As in this case, in Huarong, Commerce relied on volatility and a limited number of transaction-specific margins to justify the same 139.31 percent rate. Id. at 1233-35, 2005 WL 2365322. The Huarong Court noted, however, that “[w]hile changes in antidumping duty rates from one review to the next may be consistent with the ‘volatile nature’ of the rates for bars/wedges, Commerce has still failed to demonstrate the validity of such a large absolute increase.” Id. at 1233, 2005 WL 2365322. That is, the jump from calculated rates ranging from 1.27 percent to 34.00 percent to a 139.31 percent AFA rate needed a thorough explanation backed by substantial evidence. Thus, the Huarong Court concluded that “Commerce has failed to justify the 139.31% rate with substantial evidence” and found the rate to be “punitive in nature.” Id. at 1237, 2005 WL 2365322; see also id. (“[T]he court finds that Commerce has failed to justify the 139.31% rate assigned to the Companies, and further finds that the rate is punitive .... ”). Since it had remanded the case twice, the Court directed Commerce “to no longer employ [the 139.31 percent] rate.” Id. Commerce then selected a 47.88 percent rate on the third remand. The Huarong Court found that Commerce had finally “explained adequately the reliability and relevance of the ... AFA rate” by selecting a rate that was 13.88 percent higher than Huarong’s previous highest calculated rate in order to encourage future compliance. Shandong Huarong Gen. Grp. Corp. v. United States, 31 CIT 42, 45-46 (2007) (not reported in the Federal Supplement).

Based on the foregoing, it is clear that Commerce has again failed to support with substantial evidence the 139.31 percent rate for Huarong. In choosing an AFA rate on remand, Commerce must select one grounded in the “commercial reality” of the company under review during the POR. Commerce has failed to do so here. First, the volatility argument for Huarong is weak, considering that the company has never had a calculated rate anywhere near 139.31 percent. That is, Huarong’s calculated rates in earlier reviews varied from 1.27 percent to 34.00 percent. Thus, while Huarong has had various calculated rates, none has approached the rate assigned by Commerce.

Moreover, considering the persuasive evidence represented by these lower calculated rates, Commerce’s decision to look only at selected sales to justify a rate over one hundred points higher than Huarong’s highest calculated rate does not constitute substantial evidence of what Huarong’s calculated rate would have been. See, e.g., Gallant Ocean, 602 F.3d at 1325(stating that “a small percentage of ... transactions [does not] represent [ ] a reasonably accurate estimate of Gallant’s actual dumping margin” when “a large body of reliable information suggest[ ] the application of a much lower margin.”). This is particularly the case where an examination of Commerce’s analysis reveals that the Department looked at 447 transactions, of which 395(or roughly eighty-eight percent) had margins of less than one hundred percent, while over 75 percent of the sample had margins below fifty percent. First Remand Results at App. 3.

As this Court recently noted, to “avoid imposition of a punitive rate,” the court must consider whether “the AFA rate [is] so far from what has been demonstrated by actual rates ... that it must be rejected.” Qingdao Taifa Grp. Co., Ltd., v. United States, 34 CIT-, -, — F.Supp.2d -, -, 2010 WL 4704464, *5-6 (2010). Furthermore, this Court has held that more than doubling a previously-determined AFA rate “must necessarily be punitive.” Shandong Mach. Imp. & Exp. Co. v. United States, 34 CIT-, -, — F.Supp.2d -, -, 2010 WL 3199830, *4 (2010). Thus, a rate that is over one hundred percentage points higher and more than four times greater than Huarong’s highest previously calculated rate, as well as a rate dramatically higher than the vast majority of specific transactions Commerce looked at, renders Commerce’s choice “punitive in nature” based on Commerce’s “strained efforts to demonstrate [its] validity.” Huarong, 29 CIT at 1237. See also De Cecco, 216 F.3d at 1032 (affirming the rejection of a punitive AFA rate that “was many times higher than [the company’s] actual dumping margin”).

Because the court finds that Commerce has failed to justify with substantial evidence the 139.31 percent rate after having had two previous opportunities to do so, and further finds that the rate is punitive, the Department is directed, on remand, to no longer employ this rate and to choose a lower rate. Therefore, Commerce shall choose and support, with substantial evidence, one of the following: (1) a calculated rate from a previous review, that reflects the actual rate during the POR, with a built-in increase to deter non-compliance; or (2) reopen the record and calculate a rate that accurately reflects what the rate would have been had Huarong cooperated, with a built-in increase as a deterrent to non-compliance.

II. AFA Rate for TMC’s Sales of Picks/Mattoeks

In Tianjin I, the court found unsupported by substantial evidence Commerce’s se lection, as the AFA rate for TMC’s sales of picks/mattocks, a 98.77 percent rate calculated in the fifth review for a different respondent, Funjian Machinery Import and Export Corp. (“FMEC”). 31 CIT at 1437-38, 2010 WL 3199830; Pis. First Comments 34. Commerce applied AFA to TMC’s sales of picks/mattocks “based on the company’s failure to have available for inspection at verification its sole supplier’s factors of production data.” Id. at 1437, 2010 WL 3199830. The court found that Commerce’s justification that the rate was “calculated for another respondent in a prior segment of these proceedings” was “not sufficient for the court to find that the selected rate was a reasonably accurate reflection of what TMC’s actual rate would be during the POR,” together with a built-in increase to deter non-compliance. Id. The court remanded the issue to the Department with instructions to:

(1) explain (a) how the 98.77 percent rate for TMC’s picks/maddocks is a reasonably accurate estimate of TMC’s actual rate with a built-in increase to deter non-compliance; and (b) why it did not select as an AFA rate for TMC’s sales of picks/mattocks one of the previously assigned lower rates, albeit with a built-in increase to deter future non-compliance; or (2) reopen the record and obtain evidence to support an actual calculated rate for TMC’s sales of picks/mattocks.

Id. at 1438, 2010 WL 3199830.

On remand, the Department claims that it did not reopen the record. Nonetheless, it did place on the record additional factual support in an effort to demonstrate that the 98.77 percent rate was relevant to TMC. Specifically, Commerce analyzed the company’s sales in the twelfth review and found that TMC had transaction-specific margins in that review that were considerably above 98.77 percent. 10 First Remand Results at 14. The Department added:

We also compared the [twelfth review] transactions that approximate the proposed AFA rate of 98.77 percent to other U.S. sales of [picks/mattocks] made by TMC. The U.S. transactions corroborating the AFA rate do not appear to be aberrant or unusual in any way. They appear to be made in commercial quantities. Because we are making an adverse inference with regard to TMC, we regard these transactions as representative of the margins we would have calculated for this company in the thirteenth review (with a built-in incentive to encourage cooperation) had it not received total AFA.

Id.

In addition, as in its bars/wedges analysis, Commerce relies on the Rhone Poulenc presumption to corroborate the 98.77 percent rate. Id. at 15. Commerce argues that if TMC could have demonstrated that its rate would have been lower than the highest previous margin, it would have provided such information. Further, to explain why the Department did not select one of TMC’s previously assigned lower rates as an AFA rate, the Department maintains that it has discretion to choose the information on which it relies. Id. at 15. (“Because TMC’s sales of picks/mattocks are receiving total AFA because it could not provide factors of production information, we do not find that using one of TMC’s prior rates would be appropriate as an AFA rate. In administrative reviews where TMC received calculated rates, TMC fully participated in those proceedings and received a calculated rate based on its own data. As a result, it would be inappropriate to apply a lower rate to TMC as AFA in this review.”).

For TMC, Commerce has still not demonstrated the reliability nor the relevance of the rate to the company. TMC argues that using a different company’s rate from a review almost a decade removed from the POR is not reflective of TMC’s commercial activity during the POR. Pis.’ First Comments 35.

In addition, TMC claims that the Department has not adequately explained why it did not select one of TMC’s previously assigned lower rates and add to it a built-in increase for non-compliance. More specifically, TMC notes that the highest prior margin for a review in which TMC fully participated was 4.76 percent, and that 4.76 percent should have been the starting point for a margin in this case, plus an addition to deter future non-compliance. See Pis.’ First Comments 39.

The court is not persuaded that Commerce has supplied the substantial evidence to sustain the 98.77 percent AFA rate for TMC’s sales of picks/mattocks. First, as TMC argues, “nothing in the record ties the ... rate to” TMC, as Commerce chose a rate calculated for a different party in a completely different review. See Gallant Ocean, 602 F.3d at 1324. By not using a rate calculated for TMC, but rather one calculated for another company, Commerce must produce evidence that this rate is relevant to TMC. “Although Commerce has discretion in choosing from a list of secondary information to support its adverse inferences, Commerce must select secondary information that has some grounding in commercial reality.” Id.

Second, Commerce’s analysis of a small fraction of TMC’s sales in a time period outside of the POR does not constitute substantial evidence that FMEC’s rate from another POR was relevant to TMC during the POR. This is because the selected sales from the twelfth review simply do not corroborate a rate of 98.77 percent. Of the twenty-two sales Commerce examined, only two had a margin in excess of 98.77 percent, and those margins were for only one of the six products included in the analysis. Furthermore, fourteen of the twenty-two sales showed no dumping, and the six other sales had margins close to zero. First Remand Results at App. 4. Thus, the selected sales do not constitute substantial evidence, or indeed very much evidence at all, that TMC’s margin during the POR was anything approaching 98.77 percent.

While the Department has been sustained in its use of a small number of sales to corroborate rates in the past, the manner in which these sales were used in this case distinguishes it from cases like Ta Chen and PAM. For instance, in Ta Chen, Commerce used the company’s “sales data from the relevant review period.” Gallant Ocean, 602 F.3d at 1324. Thus, the rate was corroborated by Ta Chen’s own sales for the POR. Furthermore, as the Federal Circuit noted in Gallant Ocean, “[u]nlike PAM, Commerce in the present [Gallant Ocean] case did not show that a small percentage of the mandatory respondents’ transactions represented a reasonably accurate estimate of Gallant’s actual dumping margin.” Id. at 1325. That is, in PAM, the Federal Circuit found that “in view of the entire record, ... the transactions were reasonably tied to PAM’s actual dumping margin.” Id. Such is not the case here where the selected transactions point to a substantially lower rate than the one assigned.

Moreover, as TMC also emphasizes, there is an even wider gulf here than in Gallant Ocean between the AFA rate for this review and TMC’s previous non-AFA rates. TMC’s highest previous calculated rate was 4.76 percent. The rate chosen here is ninety-four percentage points greater than the highest calculated rate for TMC and almost twenty times greater than that rate. This disparity, together with Commerce’s failure to cite any serious evidence corroborating it, is enough for the court to conclude that the 98.77 percent AFA rate chosen by Commerce is aberrational and punitive. See id. at 1324(stress-ing the difference between cooperating respondents’ actual dumping margins of 2.58 to 10.75 percent and Gallant’s AFA rate of 57.64 percent). Although Commerce insists that it could not use one of TMC’s previous rates as a baseline because none of them were AFA rates, that argument does not justify Commerce’s selection of a punitive rate. See also Huarong, 29 CIT at 1237(finding a 139.31 percent AFA rate punitive when the highest prior rate was 34.00 percent).

It is also worth noting that Commerce’s reliance on Rhone Poulenc in this context continues to be misplaced. First, as noted, the Rhone Poulenc presumption does not excuse Commerce from corroborating secondary evidence as required by statute. Second, no common sense inference can be drawn from TMC’s making the effort to answer Commerce’s questionnaires and then being unable to produce its supplier’s records at verification because they were seized by local authorities.

Because the court finds that Commerce has twice failed to produce substantial evidence demonstrating the relevance of the 98.77 percent rate to TMC during the POR, and because it further finds that the rate is punitive, Commerce is directed upon remand to no longer employ this rate and to choose a lower rate. On remand, Commerce must choose an AFA rate based on the “commercial reality” of TMC’s sales of picks/mattocks. Therefore, Commerce shall choose and support, with substantial evidence, one of the following: (1) select a calculated rate from a previous review, that reflects TMC’s actual rate during the POR, with a built-in increase to. deter non-compliance; or (2) reopen the record and calculate a rate that accurately reflects what the rate would have been had TMC cooperated, with a built-in increase as a deterrent to non-compliance.

CONCLUSION

For the foregoing reasons, Commerce’s First Remand Results and Supplemental Remand Results are sustained in part and remanded. The remand results shall be due on May 4, 2011; comments to the remand results shall be due on June 3, 2011; and replies to such comments shall be due on June 17, 2011.

1 . The court sustained the application of AFA to TMC’s and Huarong's sales of bars/wedges: ”[A]s this Court has previously held, plaintiffs’ 'failure initially to provide the relevant information with respect to their invoicing arrangement, information that was fully within their command, justified Commerce's application of AFA' to plaintiffs' claimed 'agent' sales.” Tianjin I, 31 CIT at 1424, 2010 WL 3199830 (quoting Shandong Huarong Mach. Co. v. United States, 30 CIT 1269, 1278, 435 F.Supp.2d 1261, 1270 (2006)). As to TMC’s sales of picks/mattocks, the court found that "based on the company’s failure to have available for inspection at verification its sole supplier’s factors of production data,” "Commerce's application of AFA ... [was] supported by the record.” Tianjin I, 31 CIT at 1437, 1432, 2007 WL 2701368 (citation omitted).

2 . Pursuant to 19 U.S.C. § 1677e(a), if:

(1) necessary information is not available on the record, or

(2) an interested party or any other person—

(A) withholds information that has been requested by the administering authority or the Commission under this subtitle,

(B) fails to provide such information by the deadlines for submission of the information or in the form and manner requested, ...

(C) significantly impedes a proceeding under this subtitle, or

(D) provides such information but the information cannot be verified ...,

the administering authority and the Commission shall, subject to section 1677m(d) of this title, use the facts otherwise available in reaching the applicable determination under this subtitle.

If Commerce determines that the above criteria are met, and makes the separate subjective determination that the respondent has “failed to cooperate by not acting to the best of its ability to comply with a request for information,” then, under 19 U.S.C. § 1677e(b), the agency "may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available.” 19 U.S.C. § 1677e(b).

3 . TMC’s AUV "declined by 34.33 percent from the earlier [eighth review] to the later period [thirteenth review].” First Remand Results at 7.

4 . While the phrase "total adverse facts available” is not referenced in either the statute or the agency’s regulations, it can be understood, within the context of this case, as referring to Commerce’s application of the "facts otherwise available” and "adverse inferences” provisions of 19 U.S.C. § 1677e after rejecting as untrustworthy all information submitted by respondents in this review.

5 . As explained in Tianjin I:

Plaintiffs fail to acknowledge ... that Commerce discovered, after the issuance of several supplemental questionnaires, that the business relationship between Huarong and TMC was nothing more than a scheme apparently directed toward circumventing the antidumping duties applicable to Huarong’s HFHTs sales to the United States.... [T]he mere statement that sales were made through an agent when, in reality, the agent’s role was simply to provide the principal with blank invoices, is not enough to preclude Commerce from resorting to facts otherwise available.

31 CIT at 1423, 2007 WL 2701368.

6 . "While plaintiffs seem to claim that the surrogate values that would have been used in the 13 th administrative review would have been lower than the surrogate value for steel used in the 8th administrative review, U.S. prices at least for TMC’s bars/wedges also declined.” Supplemental Remand Results at 8-9.

7 . Huarong’s rate for the twelfth review was not judicially reviewed.

8 . The Department examined 447 transactions from the eleventh review. None were as high 139.31 percent, and only twelve, or roughly .03 percent of the transactions, were above 130 percent. In addition, only fifty-two of the sales, or approximately twelve percent of the sample, rose above one hundred percent. More than seventy-five percent of sales had less than a fifty percent margin. First Remand Results at App. 3.

9 . The Uruguay Round Agreements Act, Pub.L. No. 103-465, 108 Stat. 4809 (1994), changed United States law to conform to the provisions agreed upon at the Uruguay Round of negotiations of the General Agreement on Tariffs and Trade.

10 . The First Remand Results cites two sales with a margin of 363.91 percent. First Remand Results at 14.