2014 WL 1426444
Ct. Intl. Trade
2014

OPINION and ORDER

GORDON, Judge:

This consolidated action involves the U.S. Department of Commerce’s (“Commerce”) fifth administrative review of the antidumping duty order covering Floor-Standing, Metal-Top Ironing Tables from China. See Floor-Standing, Metal-Top Ironing Tables and Certain Parts Thereof from the People’s Republic of China, 76 Fed.Reg. 15,297 (Dep’t of Commerce Mar. 21, 2011) (final results admin, review), as amended by 76 Fed.Reg. 23,543 (Dep’t of Commerce Apr. 27, 2011) (amended final results admin, review) (collectively, “Final Results ”); see also Issues and Decision Memorandum for Ironing Tables from China, A-570-888 (Mar. 22, 2011), available at http://ia.ita.doc.gov/frn/summary/ PRC/2011-6558-1.pdf (last visited this date) (“Decision Memorandum ”). Before the court are the Final Results of Redetermination, ECF No. 113 (“Second Remand Results ”) filed by Commerce pursuant to Since Hardware (Guangzhou) Co. v. United States, 37 CIT -, 911 F.Supp.2d 1362 (2013) (“Since Hardware II”); see also Final Results of Redetermination, ECF No. 85 (“First Remand Results”); Since Hardware (Guangzhou) Co. v. United States, Consol. Court No. 11-106, ECF No. 81 (“Since Hardware I”) (order remanding to Commerce). The court has jurisdiction pursuant to Section 516A(a)(2)(B)(iii) of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2006), 1 and 28 U.S.C. § 1581(c) (2006). Familiarity with the prior judicial and administrative decisions in this action is presumed.

Plaintiffs Since Hardware (Guangzhou) Co., Ltd. (“Since Hardware”) and Foshan Shunde Yongjian Housewares & Hard-wares Co., Ltd. (“Foshan Shunde”) both challenge Commerce’s financial statement selection; Foshan Shunde challenges Commerce’s brokerage and handling surrogate valuation. 2 See Since Hardware Comments on Remand Results, ECF No. 119; Foshan Shunde Comments on Remand Results, ECF No. 118; Foshan Shunde Reply Comments on Remand Results, ECF No. 128. Defendant and Defendant-Intervenor, Home Products International, Inc. (“Home Products” or “HPI”), oppose these challenges and argue that the Second Remand Results should be sustained. See Def.’s Comments on Remand Results, ECF No. 126; Home Products’ Comments on Remand Results, ECF No. 127; Def.’s Surreply to Comments on Remand Results, ECF No. 145; Home Products’ Surreply to Comments on Remand Results, ECF No. 144. For the reasons that follow, the court sustains Commerce’s financial statement selection, but remands the brokerage and handling issue to Commerce for further consideration.

I. Standard of Review

When reviewing Commerce’s antidumping determinations under 19 U.S.C. § 1516a(a)(2)(B)(iii) and 28 U.S.C. § 1581(c), the U.S. Court of International Trade sustains Commerce’s “determinations, findings, or conclusions” unless they are “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i). More specifically, when reviewing agency determinations, findings, or conclusions for substantial evidence, the court assesses whether the agency action is reasonable given the record as a whole. Nippon Steel Corp. v. United States, 458 F.3d 1345, 1350-51 (Fed.Cir.2006). Substantial evidence has been described as “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Consol. Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938). Substantial evidence has also been described as “something less than the weight of the evidence, and the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence.” Consolo v. Fed. Mar. Comm’n, 388 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966). Fundamentally, though, “substantial evidence” is best understood as a word formula connoting reasonableness review. 3 Charles H. Koch, Jr., Administrative Law and Practice § 9.24[1] (3d ed. 2014). Therefore, when addressing a substantial evidence issue raised by a party, the court analyzes whether the challenged agency action “was reasonable given the circumstances presented by the whole record.” Edward D. Re, Bernard J. Babb, and Susan M. Koplin, 8 West’s Fed. Forms, National Courts § 13342 (2d ed. 2013).

Separately, the two-step framework provided in Chevron, U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842-45,104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), governs judicial review of Commerce’s interpretation of the antidumping statute. See United States v. Eurodif S.A., 555 U.S. 305, 316, 129 S.Ct. 878, 172 L.Ed.2d 679 (2009) (Commerce’s “interpretation governs in the absence of unambiguous statutory language to the contrary or unreasonable resolution of language that is ambiguous.”).

II. Discussion

A. Financial Statement Selection

When selecting financial statements to calculate the financial ratios for respondents’ margins, Commerce is guided by a general regulatory preference for publicly available, non-proprietary information. 19 C.F.R. § 351.408(c)(1), (4) (2009). Beyond that, Commerce generally considers the quality, specificity, and contemporaneity of the available financial statements. See Fresh Garlic from the People’s Republic of China, 67 Fed.Reg. 72, 139 (Dep’t of Commerce Dec. 4, 2002) (final results new shipper review). During the administrative review, Commerce had a choice from among four Indian financial statements: '06-'07 Infiniti Modules Private Ltd. (“Infiniti Modules”); '08-'09 Omax Autos Ltd. (“Omax”); and '07-'08 and '08-'09 Maximaa Systems Ltd. (“Maximaa”). In the Final Results Commerce chose the '06-'07 Infiniti Modules financial statements alone as the best available information from which to calculate the financial ratios.

1. Infiniti Modules

When first reviewing the issue of Commerce’s selection of the '06-'07 Infiniti Modules financial statements, the court could not sustain Commerce’s conclusion that those statements were publicly available. See Since Hardware I at 4-5. On remand Commerce acknowledges that it erred in the Final Results when it concluded that the Infiniti Modules financial statements were available through a website. First Remand Results at 7. Commerce clarified, though, that it still believed the financial statements were publicly available because they were used in a prior administrative réview and available on the public administrative record of that review. Id. at 29. Commerce also explained that Commerce and all interested parties had significant experience with Infiniti Modules’ financial statements. Id. at 5-6.

In reviewing the First Remand Results, the court acknowledged Commerce’s reasonable desire to continue to use a data source with which all parties were well acquainted, but could not sustain Commerce’s continuing insistence that the Infiniti Modules financial statements were “publicly available.” Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1368-69. The problem undermining Commerce’s decision was its reliance upon Catfish Farmers of America v. United States, 33 CIT 1258, 1272, 641 F.Supp.2d 1362, 1377 (2009), as providing “the standard for public availability established in our practice.” First Remand Results at 29. The court noted that Catfish Farmers nowhere explains Commerce’s standards or criteria for public availability. Foshan Shunde, on the other hand, identified a contemporaneous proceeding in which Commerce had applied fairly rigorous standards of public availability: Final Results of Redetermination Pursuant to Court Remand, Yantai Xinke Steel Structure Co. v. United States, Court No. 10-00240, ECF No. 83 at 18-23 {“Steel Grating Remand Results ”); see also Certain Steel Grating from the People’s Republic of China, 75 Fed.Reg. 32,-366 (Dep’t of Commerce June 8, 2010) (final LTFV determ.). The court therefore directed Commerce to reconcile its approach here with the Steel Grating Remand Results.

In the Second Remand Results Commerce provides a comprehensive and reasonable justification for its continued reliance on the Infiniti Modules financial statements as among the best available information. Second Remand Results at 20-25. In doing so Commerce reasonably distinguishes the Steel Grating Remand Results and other administrative decisions relied upon by Foshan Shunde and Since Hardware. Id. at 21-23. Specifically, Commerce explains that “[i]n contrast to the cases cited by Foshan Shunde and Since Hardware, the instant case does not involve the introduction of a new financial statement or the selection of data from a new surrogate country; this case deals with a familiar financial statement whose provenance has never been called into question by even a scintilla of evidence.” Id. at 23. Also, “Infiniti’s financial statements were put on the record of multiple reviews of [the antidumping duty orders on] Folding Metal Tables and Chairs [from China] and Hand Trucks [from China], and ... no party contested the public availability” in those proceedings. Id. at 25 (footnotes omitted). Commerce concludes that the Infiniti Modules financial statements “are publicly available within the meaning of section 351.408(c)(1) of [its] regulations.” Id.

Commerce also provides a more in depth analysis of its applicable regulation, 19 C.F.R. § 351.408(c). Id. at 23-24 (analyzing promulgation of regulation). Commerce explains that the use of publicly available information is relatively more important to value material inputs than it is to value overhead, general expenses and profit because use of public information for material inputs tends to yield more representative data reflecting numerous transactions between many buyers and sellers. Id. at 23. Commerce further explains that the same imperative does not exist for overhead, general expenses and profit “because such data do not exist on an aggregated basis, and because [Commerce] uses overhead, general expenses and profit on a company-specific basis, as it must, since that is the only data available.” Id. at 24 (citing 19 C.F.R. § 351.408(c)(4)). Instead, Commerce explains that the primary purpose for obtaining publicly available information for financial statements “is to ensure that all interested parties have access to such information, and are able to comment on the reliability and relevance of such information in the particular case, and not as much for purposes of obtaining broader information that reflects numerous transactions as is the case for material inputs.” Id. And here, Foshan Shunde and Since Hardware had access to, and were able to comment upon, the financial statements at issue, finding “no basis to question the reliability of the data,” which in turn led Commerce to conclude that “the purpose of the regulation is fulfilled in this case.” Id.

Unlike the Final Results and First Remand Results, the court cannot identify any unreasonableness in Commerce’s determination here. Commerce addressed those administrative precedents in which it applied rigorous public availability criteria to new financial statements, reasonably distinguishing them given the widespread past use of Infiniti Modules’ financial statements in prior segments of the Ironing Tables order, respondents’ own substantive reliance on the financial statements in those prior segments, as well the financial statements’ use in proceedings under other antidumping duty orders. Indeed, their “provenance has never been called into question.” Id. at 23. 3 Also, to the extent Commerce’s decision implicates an interpretation of its regulation, that interpretation is not “plainly erroneous or inconsistent with the regulation,” Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414, 65 S.Ct. 1215, 89 L.Ed. 1700 (1945), and is therefore entitled to deference. See American Signature, Inc. v. United States, 598 F.3d 816, 827 (Fed.Cir.2010). Accordingly, the court -will sustain Commerce’s selection of the Infiniti Modules financial statements.

2. Maximaa

In Since Hardware II the court determined that Commerce had not reasonably distinguished the Final Results, in which it selected Infiniti Modules’ financial statements and rejected Maximaa’s, from another administrative proceeding in which it did the exact opposite, Folding Metal Tables and Chairs from the People’s Republic of China, 74 Fed.Reg. 68,568 (Dep’t of Commerce Dec. 28, 2009) (final results admin. review) (“Folding Metal Tables and Chairs ”); see also Issues and Decision Memorandum for Folding Metal Tables and Chairs from China, A-570-868 (Dec. 18, 2009), available at http://ia.ita.doc.gov/frn/summary/prc/E9-30695-l.pdf (last visited this date). Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1371.

As with its treatment of Infiniti Modules’ financial statement, in the Second Remand Results Commerce provides a comprehensive and reasonable justification for its rejection of Maximaa’s financial statements. Second Remand Results at 6-11, 26-28. The court now has a better understanding of what transpired during the administrative proceeding and how the interested parties developed the administrative record. In response to respondent’s addition of the Maximaa financial statements to the administrative record, petitioner augmented the record with documentation and argumentation that enabled Commerce to reasonably pursue an alternative financial statement selection than it had in Folding Metal Tables and Chairs. Id. at 7. Importantly, the record in this proceeding established that Infiniti Modules was a not just an assembler (a conclusion Commerce reached in Folding Metal Tables and Chairs), but a manufacturer. Id. at 9-10. Petitioner also added documentation and information that undermined the suitability of Maximaa as a surrogate (information that apparently was not present in Folding Metal Tables and Chairs). This information and associated reasonable inferences support Commerce’s determination that Maximaa was apparently transitioning from furniture assembly to other lines of business like information technology services. Id. at 7-8. In short, Commerce has reasonably identified several problems with the Maximaa financial statements that render them unsuitable for use, precluding the court from ordering Commerce to incorporate them in the financial ratio calculation.

Interested parties bear the burden of developing the administrative record. QVD Food Co. v. United States, 658 F.3d 1318, 1325 (Fed.Cir.2011). Here, petitioner was equal to the task, aggressively countering respondents’ efforts to supplement the record with additional surrogate financial statements. The result is that respondents have failed to develop an administrative record that would mandate a reasonable mind to include Maximaa’s financial statements within the financial ratio calculation. Accordingly, the court will sustain Commerce’s rejection of the Maximaa financial statements.

B. Brokerage and Handling

When the court first reviewed the brokerage and handling issue, it had difficulty understanding exactly what Commerce did when calculating this typically routine and well-known component of most international trade transactions. The seeming impenetrability of Commerce’s calculation aroused the court’s suspicions about the reasonableness of Commerce’s approach. Since Hardware I at 8. The court thus directed Commerce “to prepare a clear, complete public summary of its calculation of Foshan Shunde’s B & H expense.” Id. Commerce obliged, attaching a summary to the First Remand Results revealing that it divided a $645 baseline cost described in the World Bank’s Doing Business 2010: India publication by “the estimated weight of [Foshan Shunde’s] product shipped in 20-foot containers.” First Remand Results at Att. A. This formula did not appear to comport with record evidence appearing to show that B & H costs are actually lower than $645 in coastal cities and do not increase proportionately with container size, leading the court to remand again. Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1374; see First Remand Results at Att. A.

In the Second Remand Results, Commerce continues to use the $645 Doing Business 2010: India data point as a basis for calculating Foshan Shunde’s B & H costs, but adjusts its treatment of the three components underlying that figure in response to the court’s observations:

In Doing Business India — 2010, total brokerage and handling expenses for exporting a 20 foot container is listed as follows:
1) Document Preparation $350
2) Customs Clearance and Technical Control $120
3) Ports and Terminal Handling $175
Because [Commerce] had available data pertaining to shipments in 20 foot containers, while Foshan Shunde shipped ironing tables in 40 foot containers, Commerce adjusted the weight of the ironing tables shipped in a 40 foot container to an estimated weight that would correspond to a 20 foot container size quote from the Doing Business India — 2010 study.... Therefore, the following adjustment was made to determine the estimated comparable weight of the ironing tables had they been shipped in 20 foot containers (D):
D=(A*B)/C
A represents the cubic capacity of a 20 foot container, which is 33 cubic meters.
B represents the weight of ironing tables shipped in 40 foot containers, which is [ ] kg.
C represents the cubic capacity of a 40 foot container (the size in which both respondents shipped merchandise), which is 67.3 cubic meters.
In this case D yields an estimated weight of [ ] kilograms for ironing tables shipped in a 20 foot container.
D= 33*[ ]/67.3= [].
In the second redetermination, Commerce determined that Foshan Shunde’s Document Preparation and Customs Clearance charges increased proportionately with container size. That is, this cost, for use of a 40 foot container increases 100 percent, relative to this cost for use of a 20 foot container. However, for Ports and Terminal Handling Charges, Commerce determined that it increased by only 50 percent as a result of using of a 40 foot container in lieu of a 20 foot container. Thus, Commerce used the following variables and formula to calculate Foshan Shunde’s brokerage and handling expense (B & H) (per kilogram):
E=Documents Preparation Charge, which is $350.
F=Customs Clearance and Technical Control Charges, which is $120.
G=Ports and Terminal Handling Charges, which is $175.
B & H represents the calculated expense for brokerage and handling (per kilogram).
B & H= ((E + F)/D) + ((G*1.5*0.5)/ D)=$[ ] per kilogram. 8
In this redetermination, this formula equates to:
B & H = ((350 +120)/[ ]) + ((175*1.5*0.5)/[ ] )=$[ ] per kilogram.

Def.’s Resp. to Ct.’s Feb. 27, 2014 Order 1-3 (footnotes omitted, emphasis in original), ECF No. 140 (“Calculation Submission”). 4

Foshan Shunde now objects to four aspects of Commerce’s revised B & H calculation. First, Foshan Shunde again disputes the $645 data point, albeit for reasons premised upon new evidence in the administrative record. Second, Foshan Shunde objects to Commerce’s selection of a 50% increase to convert ports and terminal handling costs for 20-foot to 40-foot containers in light of record evidence showing such cost increases can be as low as 30%. Third, Foshan Shunde argues that Commerce ignores record evidence demonstrating that Foshan Shunde actually incurred document preparation and customs clearance costs only once every 6.2 40-foot containers it shipped. Lastly, Foshan Shunde insists that Commerce’s reliance on an estimated 20-foot container weight is unreasonable because it implies a relationship between B & H costs and container weight that is not supported by the record. The court largely agrees and therefore remands this issue to Commerce for clarification or reconsideration, as may be appropriate.

1. The World Bank’s Doing Business 2010 Publication

As a preliminary matter, the court in Since Hardware II directed Commerce to address evidence appearing to show that B & H costs are lower on average for Indian companies that, like Foshan Shunde, are located near a seaport. In so doing, the court made the following observation: “The data that Commerce relied upon, the World Bank’s Doing Business in India: 2010, is composed of the B & H costs of 17 Indian cities/regions[.] ... [BJased on the aggregate data of all 17 cities, Commerce calculated $645 in B & H costs.” Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1379. What is now clear, however, is that the $645 figure is not based on the aggregate data of 17 Indian cities. $645 is in fact the estimated cost for one city: Mumbai. The court’s misunderstanding stemmed in large part from Foshan Shunde’s inaccurate but uncontested representations of the Doing Business 2010 evidence. Compare Mot. for J. on the Agency R. of Foshan Shunde Yongjian Housewares & Hardwares Co., 27-30, ECF No. 44, and Pl. Foshan Shunde’s Comments on the Commerce Department’s Remand Determination 16, ECF No. 89, with First Remand Results at 38-41 and Def.’s Resp. to Pls.’ Comments Concerning Remand Results 14-21, ECF No. 100. 5

To clarify, the World Bank’s Doing Business 2010 publication compares the costs of doing business in 183 different economies based upon surveys of local experts. Foshan Shunde Surrogate Values for the Final Results Ex. 8 at 26 (Dep’t of Commerce Oct. 18, 2010) PD 96 6 (“Foshan Shunde SV Submission”). These surveys “are built on the basis of standardized case scenarios,” which evaluate the costs a hypothetical business would incur when undertaking various activities in an economy. See id. at 2. The “trading across borders” segment of each economy-specific study details the costs a hypothetical business located within that economy’s largest city would incur when exporting product in a single 20-foot shipping container. Id. at 6. The “trading across borders segment” of the India-specific Doing Business 2010: India thus estimates the following costs a hypothetical company would incur when exporting a single 20-foot container from India’s largest city, Mumbai:

City Document Preparation Customs Clearance Ports & Terminal Handling Inland Transportation Total

Mumbai_$350_$120_$175_$300_$945

See id. Ex. 4 at 10. Commerce subtracted the $300 inland transportation component, resulting in the $645 baseline cost used in both remand determinations. Second Remand Results at 13.

The World Bank’s methodology “come[s] at the expense of generality,” as costs in an economy’s largest city “may not be representative of regulation [costs] in other parts of the economy.” Foshan Shunde SV Submission Ex. 8 at 6, 26. To address this limitation, the World Bank also produces “subnational” reports for additional cities in several large economies, including India. Id. at 6, 26. The 2010 subnational reports for India estimate the following “trading across borders” costs for 16 additional cities:

City Document Preparation Customs Clearance Ports & Terminal Handling Total (Excluding Inland Transportation)

Chennai_$252_$61_$125_$438_

Kochi $210 $57 $108 $375

Kolkata $224 $95 $143 $462

Ranchi $252 $78 $143 $473

Patna $230 $91 $143 $464

Jaipur $187 $227 $318 $732

Indore $226 $57 $175 $458

Bhubaneswar $217 $59 $81 $357

Ahmedabad $217 $93 $318 $628

Ludhiana $213 $13 $175 $401

Guwahati $204 $60 $143 $407

New Delhi $230 $65 $175 $470

Noida $230 $65 $175 $470

Gurgaon $230 $65 $175 $470

Bengaluru $206 $66 $125 $397

Hyderabad $228 $57 $125 $410

See Foshan Shunde Comments on Remand Results Att. 1 Ex. 2; Second Remand Results at 13. 7 Confusingly, the record also contains a subnational report for Mumbai, which repeats the data summarized in the broader study without clarification. Foshan Shunde SV Submission at Ex. 4. For ease of reference, the court refers to Commerce’s adjusted India-specific “trading across borders” value as the “Doing Business 2010: India ” or “Mumbai-only” data point.

At first, the administrative record only included the Doing Business 2010: India study detailing costs in Mumbai as a proxy for India as a whole, as well as subnational reports detailing costs in the seaport cities of Chennai, Kochi, Kolkata, and Mumbai. Foshan Shunde SV Submission Exs. 3, 4; see Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1373-80. During the second remand proceedings, Commerce supplemented the record with subnational reports for an additional seven inland cities: Ludhiana, Guwahati, New Delhi, Noida, Gurgaon, Bengaluru, and Hyderabad. Second Remand Results at 12-13 & n. 49; Foshan Shunde Comments on Remand Results Att. 1 at 24. Foshan Shunde responded by submitting subnational reports for the six remaining inland cities, namely, Ahmedabad, Bhubaneswar, Indore, Jaipur, Patna, and Ranchi. Foshan Shunde Comments on Remand Results Att. 1 at Ex. 1.

In the table below the court summarizes the Doing Business 2010: India and 16 subnational report data for analysis in the B & H calculation:

_Data Source Document _Preparation Customs Clearance Ports & Terminal Handling Total (Excluding Inland Transportation)

Mumbai only (ie., Doing Business 2010: India)_ $350.00 $120.00 $175.00 $645.00

Average of all seaport cities (ie., Mumbai, Kochi, Kolkata, and Chennai) $259.00 $83.25 $137.75 $480.00

Average of Commerce’s inland cities (ie., Ludhiana, Guwahati, New Delhi, Noida, Gurgaon, Bengaluru and Hyderabad)_ $220.14 $55.86 $156.14 $432.14

Average of all inland cities (ie., the 13 cities other than the • seaport cities above)_ $220.77 $76.62 $174.69 $472.08

Average of all 17 cities_$229.76_$78.18 $166.00_$473.94

See id. at Att. 1 Ex. 2; Foshan Shunde SV Submission Exs. 3, 4; Second Remand Results at 13.

2. The $645 Mumbai-Only Data Point

Turning to the substance of Commerce’s revised B & H calculation, when reviewing substantial evidence issues involving Commerce’s selection of the best available surrogate values, the court evaluates “whether a reasonable mind could conclude that Commerce chose the best available information.” Goldlink Indus. Co. v. United States, 30 CIT 616, 619, 431 F.Supp.2d 1323, 1327 (2006); see also CI-TIC Trading Co. v. United States, 27 CIT 356, 366 (2003) (“[W]hile the standard of review precludes the court from determining whether [Commerce’s] choice of surrogate values was the best available on an absolute scale, the court may determine the reasonableness of Commerce’s selection of surrogate prices.”).

In the court’s view, no reasonable mind would conclude that the Mumbai-only data point is the “best available” information on the administrative record to provide the baseline for calculating Foshan Shunde’s B & H costs. Commerce’s announced criteria for selecting surrogate values in accordance with 19 U.S.C. § 1677b(c)(l) is to “select surrogate values which are product-specific, representative of a broad-market average, publicly available, contemporaneous with the POR, and free of taxes and duties.” First Remand Results at 17-18 (citing Certain Polyester Staple Fiber from the People’s Republic of China, 75 Fed.Reg. 1336 (Dep’t of Commerce Jan. 11, 2010) (final results admin, review)). The World Bank apparently derived and published the Mumbai-only data point and the subnational report data points using the same methodology, meaning all 17 data points are equally available to the public, specific to the costs in question, and contemporaneous to the POR. The Mumbai-only data point, however, is limited to a hypothetical shipment of goods from one city, whereas the subnational reports offer data points for hypothetical shipments of goods from 16 additional cities. In other words, all 17 data points on the record are qualitatively equal in all respects except that they, in aggregate, represent a broader market average than the Mumbai-only data point in isolation.

Commerce attempts to justify its selection by explaining that “the World Bank assigned importance to the accessibility of a larger port relative to the accessibility of other, smaller ports within that country,” and that it would “decline to second-guess the statistical assumptions underlying the design of [the Doing Business 2010: India ] study.” Second Remand Results at 33-34. But the weakness Commerce “decline[s]” to consider is precisely that which led the World Bank to issue subnational reports for other Indian cities in the first place:

The Doing Business methodology has 5 limitations that should be considered when interpreting the data. First, the collected data refer to businesses in the economy’s largest business city and may not be representative of regulation [costs] in other parts of the economy. To address this limitation, subnational Doing Business indicators were created for 17 economies in 2008/2009[, including] ... India____

Foshan Shunde SV Submission Ex. 8 at 26 (emphasis added). The World Bank recognized that the Indian economy is too large to support the assumption that costs in Mumbai alone are the most useful approximation of costs in India as a whole. Indeed, the administrative record appears to bear this out. Costs vary from as low as $357 in Bhubaneswar to as high as $732 in Jaipur. Costs in Mumbai are the second highest of any city on the administrative record, and almost 27% higher than the $473.94 average cost of all 17 cities. See Foshan Shunde Comments on Remand Results Att. 1 Ex. 2. Commerce’s selective reliance on the “statistical assumptions” underlying the Doing Business 2010: India data point is therefore not a reasonable basis to ignore the 16 additional and identical-quality data points for other Indian cities on the administrative record.

Nevertheless, Commerce reasonably declined to use Foshan Shunde’s preferred alternative figure, the $480 average seaport city cost. As Commerce correctly explains, “within the four Subnational Report data points [for seaport cities], [B & H] charges range from range from a low of $375 (Kochi) to a high of $645 (Mumbai),” a 72% difference. Second Remand Results at 14. Moreover, brokerage and handling charges in the other seven inland cities that Commerce analyzed “range from a low of $397 (Bengaluru) to a high of $469 (Gurgaon, New Delhi, Noida),” all of which “are substantially lower than the $645 brokerage and handling charges associated with Mumbai, a data point that is close to a seaport.” Id. (emphasis in original). As Home Products points out, costs in the most remote city, Ludhiana, are lower than costs at three of the four seaport cities on the record. Home Products’ Surreply to Comments on Remand Results 7, 13, ECF No. 144. Beyond the limited set of inland cities Commerce analyzed, the average cost of all 13 inland cities on the record is $472.08, $7.92 lower than the seaport city average. There does not appear to be any meaningful connection between distance from a seaport and B & H costs, at least outside of the inland transportation costs Commerce excluded from the calculation.

Commerce may not have intended to undercut its selection of the $645 Mumbai-only data point when it took the risk of adding subnational reports for inland cities to the administrative record. But now that it has, Commerce must reconsider its calculation of Foshan Shunde’s B & H costs. Relying on the Mumbai-only data point in isolation is not reasonable in light of identical-quality record evidence of B & H costs for 16 additional Indian cities, which when averaged with the Mumbai-only data point yield the broadest B & H cost data on the record. It therefore appears that a reasonable mind would conclude that the only reasonable option on remand would be to select the average of the data from all 17 cities as its baseline for calculating Foshan Shunde’s B & H costs. This therefore is what Commerce must do.

3. Foshan Shunde’s Rate Schedule Evidence

In the Second Remand Results, Commerce evaluated rate schedules for various port fees at seaport cities appearing to demonstrate that costs for handling 40-foot containers are not double the costs for handling 20-foot containers. Second Remand Results at 11-12, 31-32; see Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1380-81. Commerce found that this evidence was relevant to only one of the three components of its preferred $645 B & H baseline cost, namely, ports and terminal handling charges. Second Remand Results at 11-12. Commerce thus opted to treat the three components of its baseline cost as discreet elements of the B & H calculation, and limited its use of the rate schedule evidence to the ports and terminal handling component. To this extent, Commerce’s treatment of the rate schedule evidence is reasonable. See id.

Commerce acted unreasonably, however, in how it applied the rate schedule evidence to the ports and terminal handling component. Commerce acknowledged “that the rate schedule information ... establishes that the ports and terminal handling charges associated with use of a 40-foot container increased from approximately 30 to 50 percent relative to a 20-foot container rather than proportionately [ie., by roughly 100 percent].” Id. at 11. Rather than apply an increase based on the average of 30 and 50 percent as Foshan Shunde suggested, or the average of actual cost differences listed in the rate schedules, Commerce selected the highest available data point in that range. According to Commerce, “a 50 percent increase in [container] costs is within the range of experience set forth” in the rate schedules on the record, and there is “nothing in the record ... that renders our estimate of a 50 percent increase in container charges to be unreasonable.” Id. at 31-32. To the contrary, there is evidence on the administrative record of cost increases as low as 30 percent, and that evidence renders Commerce’s selection of 50 percent unreasonable. Commerce’s selection of the highest available value feels more like the application of an adverse inference to derive a higher margin than a reasonable attempt to determine the best available value on the record. See 19 U.S.C. § 1677e(b); cf. Ta Chen Stainless Steel Pipe, Inc. v. United States, 298 F.3d 1330, 1340 (Fed.Cir.2002) (sustaining application of adverse facts available rate featwring a built-in increase intended as a deterrent to noncompliance because it was “within the range of Ta Chen’s actual sales data”).

On remand, Commerce must reconsider its application of a 50 percent increase in ports and terminal handling costs to account for evidence demonstrating that such costs may increase by as little as 30 percent. In other words, Commerce should replace the “1.5” multiplicand in its formula with a lower, reasonable value, such as the 1.4 average value that Foshan Shunde suggests.

4. Foshan Shunde’s Bill of Lading Evidence

Next, Foshan Shunde argues, as it has at every possible opportunity, that Commerce should alter its B & H calculation to reflect evidence indicating that Foshan Shunde actually incurred document preparation and customs clearance fees once every 6.2 containers it shipped. Commerce, however, chose not to address this argument at all in the Second Remand Results: “Regarding Foshan Shunde’s argument that we should apply only one single document preparation fee and one single Customs clearance fee for every 6.2 containers (based on Foshan Shunde’s claims it shipped an average of 6.2 containers per bill of lading used), we note that these arguments are not part of the Foshan Shunde surrogate value information identified by the Court in Since Hardware II, [namely, port fee schedules attached as exhibits 1 and 2 to Foshan Shunde’s August 24, 2010 surrogate value submission,] and thus not at issue in this redetermination.” Second Remand Results at 31-32 & n. 113.

Commerce’s refusal to address Foshan Shunde’s evidence contravenes the court’s finding that “Commerce unreasonably concluded [in the First Remand Results] that Foshan Shunde has failed to demonstrate which, if any, of the costs ... do not increase proportionately with volume.’ ” Since Hardware II, 37 CIT at -, 911 F.Supp.2d at 1380-81. Nowhere did the court state that this finding was limited to ports and terminal handling charges, only one of the three cost components of the $645 data point. More to the point, the court did not sustain Commerce’s treatment of document preparation and customs clearance fees, the other two cost components. See id. In fact, given that Commerce treated all three cost components as a single value in the First Remand Results, there was no occasion for the court to do so in the first place. Compare First Remand Results at Att. A with Calculation Submission at 1-5. Commerce’s position is therefore untenable. See Changzhou Wujin Fine Chem. Factory Co. v. United States, 701 F.3d 1367, 1374-75 (Fed.Cir.2012) (noting that “limited remands that restrict Commerce’s ability to collect and fully analyze data on a contested issue” are generally disfavored); cf. Am. Silicon Techs. v. United States, 334 F.3d 1033, 1039 (Fed.Cir.2003) (CIT remand order “deficient” because it “prevented Commerce from undertaking a fully balanced examination that might have produced more accurate results”).

On remand, Commerce must address Foshan Shunde’s arguments regarding document preparation and customs clearance costs. Commerce should address in particular record evidence appearing to demonstrate that' Foshan Shunde actually incurred such costs only once per 6.2 containers it shipped. If Foshan Shunde’s representations prove accurate, Commerce could correct its formula by inserting “(1/6.2)” as a multiplier into the documents preparation and customs clearance cost numerator. See Calculation Submission at 1-5.

5. Foshan Shunde’s Estimated 20-Foot Container Weight

Despite the court’s finding in Since Hardware II, Commerce continued to divide its baseline costs by “D,” an estimate of the weight Foshan Shunde would have shipped in 20-foot containers, to convert the per-container World Bank data into a per-kilogram value more readily combined with other surrogate values on the record. Commerce explained that it relied upon Foshan Shunde’s estimated 20-foot container weight because “the container size assumed in the [Doing Business ] study is for a 20 foot full container load.” First Remand Results att. A at 2; see Calculation Submission at 2. Commerce’s explanation thus rests entirely upon the presumption that the per-container World Bank costs bear some relationship to the weight of product inside.

As Foshan Shunde correctly argues, “[n]o shred of evidence suggests that the container costs presented by Doing Business, or any other source, are dependent on the kilograms inside the container. Rather, the evidence submitted by Foshan Shunde indicates that the fee structure is per container; not per kilograms in a container.” Foshan Shunde Comments on Remand Results 16-17; see also Foshan Shunde’s Reply at 16-17, ECF No. 128. Commerce’s reliance on the parameters of the World Bank study is inapposite. The fact that the World Bank expressed all “trading across borders” costs on a per-20-foot-container basis establishes nothing about the relationship between costs of 20-foot containers versus 40-foot containers. Since Hardware II, 37 CIT at-, 911 F.Supp.2d at 1380-81; see Foshan Shunde Comments on Remand Results 16- 17; see also Foshan Shunde’s Reply at 16-17. Commerce admitted as much in a similar context during the course of this litigation, noting that “$645 is not derived from the weight of the container.” Def.’s Resp. to Pis.’ Mots, for J. upon the Agency R. 31, ECF No. 49 (emphasis added); see also HPI Case Brief, at 16 (Dep’t of Commerce Nov. 15, 2010), PD 107 (noting that Commerce adjusted the per-container World Bank figure “to derive” a per-kilogram cost). Commerce never explains how costs “not derived” from container weight nevertheless increase on the basis of container weight. On the other hand, Foshan Shunde identifies bill of lading evidence suggesting that document preparation and customs clearance costs accrue on a per-container basis, as well as fee schedules demonstrating that ports and terminal handling costs increase slightly with container capacity (but not proportionately with weight). The only evidence on the record with respect to the relationship between container size and B & H costs thus does not support increasing any cost component relative to container weight. See Since Hardware II, 37 CIT at-, 911 F.Supp.2d at 1380-81.

In fact, by insisting on using Foshan Shunde’s estimated 20-foot container weight as its conversion factor, Commerce forces an unexplained increase into Foshan Shunde’s B & H surrogate value. Commerce used the formula “(G*1.5*0.5)/D” to calculate Foshan Shunde’s ports and terminal handling costs. Calculation Submission at 3 & n. 8. According to Commerce, “ ‘0.5’ represents the shipment weight of a 20 foot container (which is half that of a 40 foot container).” Id. at 3 n. 8. In calculating “D,” however, Commerce multiplied Foshan Shunde’s 40-foot container weight by 33/67.3, or approximately 0.49. Substituting “D” for its mathematical equivalent reveals the problem: (0.5/0.49)* ((G* 1.5)/ W), or more simply, 1.02*((G*1.5)/W), where W represents Foshan Shunde’s 40-foot container weight. In other words, Commerce applied two different downward conversion factors, 0.5 and approximately 0.49, to account for the same concept, resulting in a facially unreasonable 2% increase in ports and terminal handling costs. Cf. id. at 3 n. 9 (declining to “double the Documents Preparation and Customs Clearance and Technical Control charges and then hal[ve] that total” because the result would be identical).

It may seem reasonable to adjust Foshan Shunde’s actual container weight to be consistent with the parameters of the study, especially since Foshan Shunde’s estimated 20-foot container shipping weight reflects certain quantifiable aspects of its shipping experience not present in the World Bank data. See Final Results at 17-19 (selecting an estimated 20-foot container weight over the hypothetical weight used in the World Bank study because of the nature of Foshan Shunde’s per-container shipping experience); Calculation Submission at 2 (discussing proprietary information underpinning Foshan Shunde’s estimated 20-foot container weight); HPI Case Brief, at 16-17 & n. 12 (arguing in favor of using an estimated 20-foot container weight). But by using Foshan Shunde’s estimated 20-foot container weight, Commerce implicitly relies upon a relationship between B & H costs and container weight that, as Foshan Shunde argues, does not appear to find support in the record. Since Hardware II, 37 CIT at-, 911 F.Supp.2d at 1380-81. Therefore, there appear to be only two reasonable alternatives remaining on the record. First, as Foshan Shunde suggests, Commerce could use Foshan Shunde’s actual 40-foot container weight, which would yield a per-kilogram value free of any unreasonable presumption regarding the relationship between the World Bank’s estimated costs and container weight. Second, there may be evidence on the record concerning the average number of units Foshan Shunde shipped per 40-foot container. See Final Results at 17-19; HPI Case Brief, at 16-17 & n. 12. The court wonders what prevents Commerce from simply using Foshan Shunde’s average number of units shipped per 40-foot container instead of weight. Given that Commerce converted Foshan Shunde’s final per-kilogram B & H value into a per-unit price to achieve consistency with other surrogate values, Final Results at 17-19, such an approach could spare Commerce the additional conversion effort as well as the additional risk of further error.

6. Summary of B & H Remand Instructions

The court expects Commerce’s efforts on remand to be a straightforward exercise in adjusting its formula and making simple mathematical substitutions in accordance with the discussion above. As a demonstration, the table below provides a summary of what changes Commerce could make to bring its current calculation into alignment with the evidence on the record. As above, the variable “W” represents Foshan Shunde’s 40-foot container weight, “X” represents a reasonable conversion factor somewhere between 1.3 and 1.5, perhaps 1.4, and the remaining variables are the same as Commerce defined them in its Calculation Submission.

Element Commerce’s First Formula Commerce’s Second Formula Adjustment Suggestions Result (Changes in Bold)

Baseline Costs $645 E = $350 Use 17-city average instead of E = $229.76

F = $120 Mumbai-only data point F = $78.18

G = $175 G = $166.00

Augend Not applicable; $645/1) used alone (E + F)/D Replace estimated 20-foot weight with actual weight and insert multiplier to account for multiple containers per bill of lading ((E + F) * (1/6.2))/ W

Addend Not applicable; $645/D used alone (G*1.5*0.5)/D Replace estimated 20-foot weight with actual weight and replace “1.5” with a lower, reasonable conversion factor (G*X)/W

After incorporating all changes, Commerce’s current formula, (($350 + $120)/D) + (($175*1.5*0.5)/D), would become (($229.76 + $78.18)*(l/6.2)/W) + (($166*X)/ W). 8 Commerce may also substitute “W” for Foshan Shunde’s average units per container value, should such an approach comport with the evidence available on the administrative record. Having been through multiple remands on this issue, the court merely offers this approach in the interest of efficiency as a reasonable means of calculating Foshan Shunde’s B & H costs that the court could ultimately sustain.

C. Zeroing

In a prior order the court delayed its remand on Foshan Shunde’s zeroing issue to coincide with its decision on the Second Remand Results. Since Hardware (Guangzhou) Co. v. United States, Consol. Court No. 11-106, ECF No. 115 (order). Accordingly, the issue of zeroing is remanded to Commerce to address Foshan Shunde’s arguments about zeroing in the nonmarket economy context. See Foshan Shunde Submission, ECF No. 110.

III. Conclusion

Accordingly, it is hereby

ORDERED that Commerce’s financial statement selection is sustained; it is further

ORDERED that Commerce’s brokerage and handling calculation is remanded for reconsideration; it is further

ORDERED that Commerce prepare and attach to its remand results a clear, complete public summary of its calculation of Foshan Shunde’s B & H expense suitable for the court to incorporate into a written disposition of this action; it is further

ORDERED that the zeroing issue is remanded for Commerce to address in the first instance Foshan Shunde’s arguments about zeroing in the nonmarket economy context; it is further

ORDERED that Commerce shall report its remand results on or before May 29, 2014; and it is further

ORDERED that, if applicable, the parties shall file a proposed scheduling order with page limits for comments on the remand results no later than seven days after Commerce files its remand results with the court.

Notes

1

. Further citations to the Tariff Act of 1930, as amended, are to the relevant provisions of Title 19 of the U.S.Code, 2006 edition.

2

. Since Hardware also attempted to challenge Commerce’s brokerage and handling ("B & H”) valuation, but the court deemed the issue waived for inadequate briefing and argument. Since Hardware I ail; see also Home Prods. Int’l, Inc. v. United States, No. 11-00104 (Jan. 3, 2012), ECF No. 62 (order waiving challenge to B & H calculation), as amended, ECF No. 63; Home Prods. Int’l, Inc. v. United States, 36 CIT -, -, 837 F.Supp.2d 1294, 1300-02, opinion after remand, 36 CIT -, 853 F.Supp.2d 1257 (2012).

3

. Nothing in Yantai Steel Structure Co. v. United States, 38 CIT -, Slip Op. 14-38 at 20-29 (Apr. 9, 2014) detracts from the reasonableness of Commerce’s explanation of its standards for public availability in the Second Remand Results.

8

In the formula, "1.5" represents the 50 percent proportionate increase in Ports and Terminal Handling Charges through the use of a 40 foot container in lieu of a 20 foot container, and "0.5” represents the shipment weight of a 20 foot container (which is half that of a 40 foot container).

4

. It is unclear whether "the weight of ironing tables shipped in 40 foot containers” is itself the result of a conversion based on the average number of units Foshan Shunde shipped per 40-foot container. See Decision Memorandum at 16-19.

5

. Neither Commerce nor Home Products filed a motion to amend or correct Since Hardware H’s treatment of the World Bank evidence.

6

. ''PD” refers to a document contained in the public administrative record.

7

. Foshan Shunde and Commerce’s summaries contain immaterial discrepancies, apparently due to differing treatment of rounded values in the subnational reports. See, e.g., Foshan Shunde SV Submission Ex. 4 at 1-2, 4 (listing the total export cost for Chennai as “541,” even though the components' sum is only 540).

8

. Some of these suggested changes may individually have a small impact on Foshan Shunde’s overall dumping margin. See Home Products' Surreply to Comments on Remand Results 16, ECF No. 144. But in aggregate, these changes appear to be significant. Assuming that X= 1.4, and substituting "D” for the equivalent (33/67.3)*W, the drastic difference between Commerce’s calculated per-kilogram B & H surrogate value in both remand determinations and the court’s model alternative becomes obvious. Expressed in terms of W and numerals rounded to the nearest integer, Commerce’s calculations in the first and second remand determinations yield 1315/W and 1226/W respectively, whereas the court’s model alternative calculation yields 282/W.

Read the detailed case summary