CP Kelco US, Inc. v. United States

Slip Op. 15-
UNITED STATES COURT OF INTERNATIONAL TRADE
OPINION
[The court remands the final determination in an antidumping investigation of xanthan gum from
the People’s Republic of China.]
Dated:
Nancy A. Noonan and Matthew L. Kanna, Arent Fox LLP, of Washington, DC, argued
for plaintiff. With them on the brief was Matthew J. Clark.
Alexander O. Canizares, Trial Attorney, Commercial Litigation Branch, Civil Division,
U.S. Department of Justice, of Washington, DC, argued for defendant. With him on the brief
were Stuart F. Delery, Assistant Attorney General, Jeanne E. Davidson, Director, Patricia M.
McCarthy, Assistant Director, and L. Misha Preheim, Senior Trial Counsel. Of counsel on the
brief was Melissa M. Brewer, Office of the Chief Counsel for Trade Enforcement & Compliance,
U.S. Department of Commerce, Washington, DC.
Mark E. Pardo, Grunfeld, Desiderio, Lebowitz, Silverman & Klestadt LLP, of
Washington, DC, argued for defendant-intervenors Neimenggu Fufeng Biotechnologies Co., Ltd.
and Shandong Fufeng Fermentation Co., Ltd. With him on the brief were Andrew T. Schutz,
Dharmendra Choudhary, and Kavita Mohan.
Goldberg, Senior Judge: CP Kelco US (“Kelco”), Neimenggu Fufeng Biotechnologies,
Co., Ltd., and Shandong Fufeng Fermentation Co., Ltd. (collectively, “Fufeng”) challenge the
CP KELCO US, INC.,
Plaintiff,
v.
UNITED STATES,
Defendant,
and
NEIMENGGU FUFENG
BIOTECHNOLOGIES CO., LTD., and
SHANDONG FUFENG FERMENTATION,
CO., LTD.,
Defendant-Intervenors.
Before: Richard W. Goldberg, Senior Judge
Court No. 13-00288
PUBLIC VERSION
.BSDI 
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 1 of 36

Consol. Court No. 13-00288 	Page 2
final determination of the Department of Commerce (“Commerce”) in its antidumping
investigation of xanthan gum from the People’s Republic of China. Xanthan Gum from the
People’s Republic of China, 78 Fed. Reg. 33,351 (Dep’t Commerce June 4, 2013) (final determ.)
and accompanying Issues & Decision Mem. (“I&D Mem.”); Xanthan Gum from the People’s
Republic of China, 78 Fed. Reg. 43,143 (Dep’t Commerce July 19, 2013) (am. final determ.); see
also Final Determination Analysis Mem. for Neimenggu Fufeng Biotechnologies Co., Ltd.
(“Fufeng Final Determination Analysis Mem.”), PD 432 (May 28, 2013).
Together, Kelco and Fufeng claim that Commerce made several incorrect decisions: (1)
Commerce’s decision to treat the bacterial strain Xanthomonas Campestris (“X. Campestris”) as
an asset, rather than as a direct material input, (2) Commerce’s selection of the Thai Ajinomoto
financial statements over the Thai Fermentation statements for calculating surrogate financial
ratios, (3) Commerce’s selection of the Doing Business 2013: Trading Across Borders (“Doing
Business 2013”) report over competing data from a website called Dxplace for valuing Fufeng’s
truck freight, (4) Commerce’s decision to value Fufeng’s corn as a direct input, instead of
valuing the cornstarch milk produced therefrom as an intermediate input, (5) Commerce’s
decision to value Fufeng’s corn as corn imported under the Thai tariff heading for corn “fit for
animal feed,” (6) Commerce’s decisions, when allocating energy at Fufeng’s Neimenggu facility,
(a) to include all of the energy consumed at Neimenggu’s cornstarch workshop in the numerator
of Neimenggu’s xanthan-gum energy-consumption rate, and (b) to use the full amount of
unfinished xanthan gum produced at Neimenggu as the denominator of Neimenggu’s xanthan-
gum energy-consumption rate. The court remands the second of these decisions for further
explanation. As for the last two ((6)(a) and (b)), the government has requested voluntary
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 2 of 36

Consol. Court No. 13-00288 	Page 3
remand, which the court grants. The court sustains the agency’s reasoning on the remaining four
decisions.
GENERAL BACKGROUND
When foreign exporters sell their goods in the United States at less than fair value and to
the detriment of U.S. industry, the U.S. Government imposes duties on those goods. See 19
U.S.C. § 1673 (2006). These duties are called “antidumping duties.” Antidumping duties are
calculated by subtracting the foreign product’s “export price,” or the product’s price in the
United States, from its “normal value” (“NV”), or the product’s price in the exporting country.
See id.
Commerce’s method for calculating the NV of goods depends on whether the goods
come from a country with a market economy (“ME”) or from a country with a nonmarket
economy (“NME”). For market-economy goods, Commerce generally uses the goods’ price in
the exporting country as NV. See id. § 1677b(a)(1)(B)(i). For NME exports, however, the
export-country price cannot be used: The law presumes that government intervention distorts
prices in the home market. See Blue Field (Sichuan) Food Indus. Co. v. United States, 37 CIT
__, __, 949 F. Supp. 2d 1311, 1317 (2013). Nor can Commerce calculate the NV of NME
exports by adding the cost of the resources used to make those goods (called “inputs”), because
input costs may be just as distorted as the final costs of the goods. See id. To calculate NV for
goods made in NME countries, then, Commerce assigns each of the goods’ direct material inputs
an artificial market price or surrogate value. See 19 U.S.C. § 1677b(c)(1). Commerce selects
sources for artificial market prices based on which ones provide the “best available information.”
The prices generally come from a market-economy country that Commerce has selected because
it produces significant amounts of subject or similar merchandise and is economically
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 3 of 36

Consol. Court No. 13-00288 	Page 4
comparable to the NME country. 19 C.F.R. § 351.408(c)(2). Commerce then totals the
surrogate cost of the inputs used, and adds to this total input cost an amount intended to capture
any noninput costs of production, such as factory overhead, selling, general, and administrative
expenses (“SG&A”), and profit. Id. § 1677b(c)(1); 19 CFR § 351.408(c)(4). These noninput
costs are added through the application of a series of ratios known as the “surrogate financial
ratios.” The ratios, like inputs, are selected from sources deemed to provide the “best available
information.” See 19 U.S.C. § 1677b(c)(1).
In the investigation underlying this case, Commerce sought to determine the appropriate
antidumping duty to impose upon Chinese xanthan-gum manufacturers, including Fufeng. In a
preliminary determination, Commerce chose Thailand as the surrogate country from to draw
input values, and reached several tentative conclusions about the proper antidumping duty to
impose on Chinese exporters. Prelim. Surrogate Value Determination, PD 306 (Jan. 9, 2013);
Prelim. Surrogate Country Determination, PD 307 (Jan. 9, 2013). Kelco and Fufeng contested
Commerce’s initial conclusions, and Commerce revised some of them in its final determination.
Kelco and Fufeng filed separate suit, each challenging Commerce’s final-determination
conclusions on various grounds. See Order, ECF No. 23. The court consolidated the cases into
one action. Id.
JURISDICTION AND STANDARD OF REVIEW
The court has jurisdiction to hear this appeal under 28 U.S.C. § 1581(c). The court will
uphold the agency’s decisions unless those decisions are “unsupported by substantial evidence
on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B)(i).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 4 of 36

Consol. Court No. 13-00288 	Page 5
DISCUSSION
In light of these standards, the court remands three decisions to Commerce. First, the
court remands Commerce’s selection of the Thai Ajinomoto financial statements over the Thai
Fermentation statements for calculating surrogate financial ratios. Commerce has also requested
remand as to two other decisions, both having to do with the allocation of energy at Fufeng’s
Neimenggu facility. The court remands both per Commerce’s request, and sustains the agency
as to all other decisions.
I. 	Commerce’s Decision to Treat Bacterial Strain X. Campestris as an Asset
Rather than a Direct Material Input Is Supported by Substantial Evidence and
Otherwise in Accordance with Law
Commerce treated the bacterial strain X. Campestris as an asset accounted for in the
surrogate financial ratios, rather than as a direct material input. Kelco challenges Commerce’s
approach. Kelco makes three claims: (1) the decision did not attune to statute, (2) the decision
was unsupported by substantial evidence, and (3) the decision was out of accord with past
agency practice. Pl. CP Kelco US, Inc.’s Mem. of Law in Support of R. 56.2 Mot. for J. on
Agency R. 11–18, ECF No. 29 (“Pl. Kelco’s Br.”). All three of Kelco’s claims fail.
A. 	Background
Before recounting the reasons that Commerce gave for treating X. Campestris as an asset
to be compensated for in the surrogate financial ratios, it is first helpful to outline the purpose of
the surrogate financial ratios and how assets are accounted for therein. Commerce uses surrogate
financial ratios to account for those production inputs that cannot be wholly attributed to a finite
batch of subject merchandise. For example, a honey factory’s jars and corks might not be wholly
attributable to subject honey produced at the factory, if the factory also makes use of the jars and
corks when producing other goods. See Shanghai Eswell Enter. Co. v. United States, 32 CIT
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 5 of 36

Consol. Court No. 13-00288 	Page 6
1233, 1242–44 (2008). Similarly, the factory’s honey machines are assets that can be used over
and over to produce honey, so—unlike the raw honey inputs—the machines’ costs do not
correspond to any standalone batch of honey. 	See id. at 1235–40 (discussing raw inputs). As
such, the costs of the jars, corks, and honey machines need to be priced into the NV of the honey
in some other way than by simply tallying their full cost. That is part of what surrogate financial
ratios are for. 1
Surrogate financial ratios account for asset costs (such as the cost to acquire honey
machines) through depreciation 2 or amortization 3 figures. See Seamless Refined Copper Pipe
and Tube from the People’s Republic of China, 75 Fed. Reg. 60,725 (Dep’t Commerce Oct. 1,
2010) (final determ.) (“Copper Pipe and Tube”) and accompanying I&D Mem. at cmt. 2. These
figures can be itemized under factory overhead or SG&A.
As noted, Commerce chose to classify X. Campestris as an asset, rather than as a direct
material input. I&D Mem. at 35–37. Commerce reasoned that a certain physical property of X.
Campestris—its capacity to self-regenerate—made the bacteria look like an asset. The bacteria’s
self-regeneration meant that it was never “used up” in the production of xanthan gum: Although
manufacturing xanthan gum required fermenting X. Campestris cells—a process that killed
individual cells—the cells grew back. Id.; Pl. Kelco’s Br. 4 –5, 12–13. This property allowed
Fufeng to acquire its X. Campestris before the period of investigation (“POI”) had even started,
1 Compare 19 U.S.C. § 1677b(c)(1) (instructing Commerce to include in NV “an amount for general
expenses and profit”), with 19 C.F.R. § 351.408(c)(4) (redescribing this amount as accounting for “manufacturing
overhead, general expenses, and profit”), and, e.g., Seamless Refined Copper Pipe and Tube from the People’s
Republic of China, 75 Fed. Reg. 60,725 (Dep’t Commerce Oct. 1, 2010) (final determ.) and accompanying I&D
Mem. at cmt. 2 (explaining Commerce’s practice of using surrogate financial ratios to cover the amount described in
statute and regulation).
2 “Depreciation is defined as the accounting process of allocating the cost of tangible assets to expense in a
systematic and rational manner to those periods expected to benefit from the use of these assets.” Fuyao Glass
Indus. Grp. Co. v. United States, 27 CIT 1892, 1908 (2003) (internal quotation marks omitted).
3 Amortization is the same as depreciation, only with respect to intangible assets. See Hyundai Elecs.
Indus. Co. v. United States, 28 CIT 517, 536 n.7, 342 F. Supp. 2d 1141, 1158 n.7 (2004).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 6 of 36

Consol. Court No. 13-00288 	Page 7
via a one-time payment. I&D Mem. at 36. Because Fufeng paid for its X. Campestris only once,
the bacteria’s cost was not attributable to any discrete batch of xanthan gum. Commerce
concluded that the bacteria was therefore best classified as an asset.
Commerce also rebutted an argument previously made by Kelco, that classifying X.
Campestris as an asset impermissibly omitted certain asset-related costs from the xanthan-gum
NV. Id. at 37. In making this argument, Kelco had noted that asset-related costs were normally
deemed accounted for in the surrogate financial ratios. Yet, argued Kelco, the ratios chosen by
Commerce—drawn from a company called Thai Ajinomoto, which also produced a bacteria-
based product, monosodium glutamate (“MSG”)—did not account for two particular asset-
related costs. Namely, the Thai Ajinomoto ratios did not cover (1) research-and-development
costs and (2) bacterial acquisition costs. Id.
Commerce responded that Fufeng’s X. Campestris did not have any research and
development costs, because Fufeng simply bought it once and kept it. Id. As for the cost to
acquire X. Campestris, Commerce noted that there was simply no good way to account for this
cost. The bacteria’s self-regeneration imparted it with an indefinite useful life. This made
attributing the acquisition cost over any finite period of time (that is, depreciating the cost)
inaccurate. In sum, Commerce concluded that X. Campestris was best treated as an asset fully
accounted for in the Thai Ajinomoto surrogate financial ratios. Id.
B. 	Analysis
Kelco first claims that Commerce decision to classify X. Campestris as an asset did not
accord with its statutory mandate. Pl. Kelco’s Br. 12–14, 16; 19 U.S.C. § 1677b(c). One of the
reasons Commerce gave for classifying X. Campestris as an asset was that Fufeng had acquired
the bacteria before the POI had started, via a one-time payment. Yet, Kelco argues, the statute
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 7 of 36

Consol. Court No. 13-00288 	Page 8
precluded Commerce from considering “the frequency or amount of . . . [Fufeng’s] payments”
when deciding how to classify X. Campestris. Pl. Kelco’s Br. 16.
This claim fails. Section 1677b(c)(4) requires Commerce to value factors of production
(“FOPs”) using surrogate data, thereby implicitly disallowing Commerce from using subject
producers’ actual payments or costs for valuation purposes. But the statute in no way forbids
Commerce from using payment data to decide whether a particular input is better described as an
asset or as a direct material prior to attaching the appropriate surrogate value (a question
determined, in part, by whether an input is an asset or a direct material). Because the statute
simply does not say whether or not using subject producers’ payments at this prior step is
permissible, Commerce’s only statute-based obligation is to comport itself reasonably. Chevron
U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984). It was reasonable
for Commerce to use the time that Fufeng paid for the X. Campestris as evidence that the
bacteria was an asset: Fufeng’s one-time purchase and the bacteria’s self-regenerating properties
made it look like an asset. Kelco’s first statutory argument therefore fails.
Kelco next makes a substantial-evidence claim, taking issue with Commerce’s conclusion
that Fufeng “acquired” its stock of X. Campestris before the POI. Pl. Kelco’s Br. 14–16. 4 Kelco
points out that Fufeng did not own the bacteria outright, but rather licensed the “rights to exploit”
it. Pl. Kelco’s Br. 15 (quoting Case Br. Rebuttal of Neimenggu Fufeng Biotechnologies Co.,
Ltd. (“Fufeng’s Administrative Rebuttal Br.”) at 43, PD 416 (Mar. 20, 2013). Kelco then argues
that the undisputed fact of Kelco’s licensing controverted Commerce’s conclusion that Fufeng
had “acquired” the bacteria, rendering Commerce’s decision unsubstantiated in evidence. Pl.
4 In its substantial-evidence claim, Kelco does not quibble with Commerce’s conclusion that Fufeng paid
for its X. Campestris just once; Kelco’s substantial-evidence challenge is limited to Commerce’s conclusion that
Fufeng “acquired” the bacteria by paying for it. Pl. Kelco’s Br. 14–16.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 8 of 36

Consol. Court No. 13-00288 	Page 9
Kelco’s Br. 15; Pl. CP Kelco US, Inc.’s R. 56.2 Reply Br. for J. on Agency R. 11–12, ECF No.
60 (“Pl. Kelco’s Reply Br.”). 5
Kelco’s substantial-evidence claim fails. It is coherent to acknowledge that Fufeng
licensed its X. Campestris, but to nonetheless conclude that Fufeng had acquired the license that
it owned before the POI. And that is precisely what Commerce did when it noted that Fufeng’s
acquisition “included the right to further grow and exploit” the bacteria. I&D Mem. at 36.
In any case, identifying the precise rights that Fufeng had in its X. Campestris—a license
versus full ownership—was not essential to Commerce’s ultimate conclusion that the bacteria
was an asset. Commerce reached that conclusion on grounds that the bacteria self-regenerated,
such that Fufeng only needed to pay for the bacteria once. Id. It was this set of features that
rendered X. Campestris unlike direct material inputs used up in the production process. Kelco’s
substantial-evidence claim therefore fails.
Kelco’s final claim is that Commerce’s past practice mandates treating X. Campestris as
an FOP, and that 19 U.S.C. § 1677b(c) requires Commerce to assign all FOPs a surrogate value.
Pl. Kelco’s Br. 12–16. According to Kelco, Commerce has consistently applied a five-factor test
to determine whether an alleged input is in fact a direct material input or something else (like an
asset), but that the agency did not proceed through these factors in its analysis below. Id.
Kelco’s proposed test was set forth in Copper Pipe and Tube:
[T]he Department will typically value a material as a direct material input if it is
1) consumed continuously with each unit of production, 2) required for a
particular segment of the production process, 3) essential for production, 4) not
used for “incidental purposes,” or 5) otherwise a “significant input into the
manufacturing process rather than miscellaneous or occasionally used materials.”
5 In its reply, Kelco raises another substantial-evidence claim: If classifying X. Campestris as an asset
meant that the bacteria’s acquisition cost would not be priced into the xanthan-gum NV (because the bacteria’s
indefinite useful life made cost depreciation inaccurate) then asset classification was unreasonable. Pl. Kelco’s
Reply Br. 8–10. Kelco did not make this argument in its lead brief, so the argument is waived. USCIT R. 81(l).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 9 of 36

Consol. Court No. 13-00288 	Page 10
Copper Pipe and Tube and accompanying I&D Mem. at cmt. 7. Had Commerce applied this
five-factor test, it would have found that X. Campestris was a direct material input, for which a
surrogate value was statutorily required. See Pl. Kelco’s Br. 13.
Kelco’s claim fails. When an agency establishes a consistent practice, this can bind it to
at least explain any departure therefrom. E.g., Consol. Bearings Co. v. United States, 348 F.3d
997, 1007 (Fed. Cir. 2003) (holding that agency acts arbitrarily and capriciously when it
“consistently followed a contrary practice in similar circumstances and provide[s] no reasonable
explanation for the change in practice”). However, contrary to Kelco’s claim, Commerce has not
used one monolithic test to evaluate whether or not an item is a direct material input or not, but
has instead proceeded case by case. In Copper Pipe and Tube, Commerce did provide a list of
considerations “typically” made. Copper Pipe and Tube and accompanying I&D Mem. at cmt.
7. But this was simply an aggregation of different methods used in various past cases. 6 And
Commerce avoided ossifying those methods into a practice by designating them as typical, thus
welcoming the possibility of other considerations (such as, for example, frequency of payment).
Id. In sum, then, Commerce’s past practice did not require it to consider the five factors listed in
Copper Pipe and Tube. Those factors therefore could not compel Commerce to conclude that X.
Campestris was a direct material input, for which a surrogate value was statutorily required.
Commerce’s past-practice claim as to Commerce’s valuation of X. Campestris fails and
Commerce’s decision to treat the bacteria as an asset withstands this court’s review.
6 Compare Copper Pipe and Tube and accompanying I&D Mem. at cmt. 7 & nn.97–101 (citing as the
primary authority for the list of considerations a similar listing set forth in Certain New Pneumatic Off-the-Road
Tires from the People’s Republic of China, 73 Fed. Reg. 40,485 (Dep’t Commerce July 15, 2008) (final determ.) and
accompanying I&D Mem. at cmt. 27), with Bridgestone Americas, Inc. v. United States, 710 F. Supp. 2d 1359, 1364
(2010) (describing the listing from Pneumatic Off-The-Road Tires as “merely a survey of various criteria taken into
consideration in different past determinations,” as opposed to a “hard-and-fast four-prong standard”)
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 10 of 36

Consol. Court No. 13-00288 	Page 11
II. 	Commerce’s Decision to Use Thai Ajinomoto’s Financial Statements to
Calculate Surrogate Financial Ratios Was Unsubstantiated in Evidence
Fufeng challenges Commerce’s decision to calculate the surrogate financial ratios using
Thai Ajinomoto’s financial statements, rather than those of Thai Fermentation, another Thai
MSG producer. Commerce rejected Thai Fermentation’s financial statements on grounds that
two paragraphs in one footnote of the statements were left untranslated. I&D Mem. at 16 &
n.70; see Pl. Fufeng’s Br. in Support of Pls.’ Mot. for J. on Agency R. 19–20, ECF No. 26 (“Pl.
Fufeng’s Br.”); Def.’s Resp. to Pls.’ Mot. for J. on Agency R. 17, ECF Np. 43 (“Gov’t Resp.
Br.”). Once Commerce had rejected the Thai Fermentation statements, it then accepted the
Ajinomoto statements as the only statements left on record, despite evidence that Ajinomoto had
received subsidies from the Thai government. Fufeng claims that Commerce broke with past
agency practice when it rejected the Thai Fermentation statements on grounds of incompleteness
alone. Fufeng further claims that Commerce’s decision to adopt the Thai Ajinomoto statements
was contrary to substantial evidence.
The court remands to Commerce for further explanation. Commerce never addressed
why the weakness of the Thai Fermentation statements—incompleteness—was worse than the
weakness of the Thai Ajinomoto statements: evidence of subsidies. Rather, by first considering
(and rejecting) the Thai Fermentation statements and then subsequently accepting the Thai
Ajinomoto statements for lack of an alternative, Commerce effectively ignored the weakness of
the Thai Ajinomoto statements. The substantial-evidence standard does not permit such one-
sided evaluation of potential data sources. See Blue Field, 37 CIT at __, 949 F. Supp. 2d at
1328–31.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 11 of 36

Consol. Court No. 13-00288 	Page 12
A. 	Background
As noted above, surrogate financial ratios are calculated by drawing upon the financial
statements of an appropriate company, usually a producer of similar goods from the surrogate
country. Commerce’s statutory task in selecting the surrogate-ratios company is to ensure that
the company provides the “best available information.” 19 U.S.C. § 1677b(c)(1)(B). In the
investigation below, Commerce had three potential surrogate-ratio companies to choose from:
Thai Ajinomoto, Thai Fermentation, and Thai Churos. I&D Mem. at 14–16. Commerce rejected
the financial statements of Thai Churos and Thai Fermentation on grounds of incompleteness.
The Thai Churos statements were “missing several footnotes” and the Thai Fermentation
statements lacked “complete English translations.” Id. at 16. Although Commerce did not go
into detail about what exactly was missing from the Thai Fermentation statements, the record
undisputedly shows that they were incomplete insofar as two paragraphs at the bottom of
accounting note twelve, concerning depreciation of assets, were untranslated. Pl. Fufeng’s Br.
19–20; Gov’t Resp. Br. 17. Accounting note twelve nonetheless contained a fully translated
depreciation schedule, complete with a line item for “Depreciation in statement income” for
2011: 140,861,456.76 Thai baht. Pl. Fufeng’s Br. Attach. 2.
After rejecting the Thai Fermentation and Thai Churos statements, Commerce then found
Thai Ajinomoto’s statements to be the best available. Commerce reasoned that, although the
Thai Ajinomoto statements were imperfect insofar as they showed “evidence of the receipt of
countervailable subsidies,” they were the only statements left to use. I&D Mem. at 16–17.
Commerce acknowledged that it had a general practice of disregarding such statements, but cited
a past instance in which it had resorted to subsidy-affected statements given nothing better on
record. Id. (citing Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled into
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 12 of 36

Consol. Court No. 13-00288 	Page 13
Modules, from the People’s Republic of China, 77 Fed. Reg. 63,791 (Dep’t Commerce Oct. 17,
2012) (final determ.) and accompanying I&D Mem. at cmt. 2)
B. 	Analysis
Fufeng claims that Commerce’s rejection of the Thai Fermentation financial statements
was contrary to both past agency practice and substantial evidence. See Pl. Fufeng’s Br. 13–24.
According to Fufeng, Commerce’s past practice is to “only disregard incomplete financial
statements . . . where the statement is missing key sections . . . that are vital to [Commerce’s]
analysis and calculations.” Pl. Fufeng’s Br. 17 (quoting Galvanized Steel Wire from the People’s
Republic of China and Mexico, 76 Fed. Reg. 23,548, 23,551 (Dep’t Commerce Apr. 27, 2011)
(initiation of investigation) (“Galvanized Steel Wire”)). Fufeng argues that Commerce’s decision
to reject the Thai Fermentation financial statements was out of accord with this past practice
because Commerce never explained why the information missing from the Thai Fermentation
statements was vital. See id. at 17–22.
Fufeng also makes a substantial-evidence claim. Fufeng argues that, by eliminating the
Thai Fermentation financial statements and then accepting the Thai Ajinomoto statements as the
only ones available, Commerce effectively ignored the weakness in the Ajinomoto statements:
evidence that Ajinomoto had received countervailable subsidies. Id. at 22–24.
The court accepts Fufeng’s substantial-evidence claim, and remands to Commerce to
further explain why it selected the Thai Ajinomoto financial statements. When presented with
multiple imperfect potential surrogate-data sources, Commerce must faithfully compare the
strengths and weaknesses of each before deciding which to use. Blue Field, 949 F. Supp. 2d at
1328–31. Nor is this general rule of any less import when it comes to selecting the best set of
financial statements to use for surrogate financial ratios. See, e.g., Tianjin Magnesium Intern.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 13 of 36

Consol. Court No. 13-00288 	Page 14
Co., Ltd. v. United States, 34 CIT __, __, 722 F. Supp. 2d 1322, 1340 (2010) (remanding
Commerce’s financial-statement choice on grounds that Commerce insufficiently explained why
it rejected one imperfect set of statements, yet accepted another). In this case, Commerce
rejected the Thai Fermentation financial statements because two paragraphs at the bottom of
accounting note twelve were left untranslated. But Commerce then accepted the Thai Ajinomoto
financial statements despite evidence that Ajinomoto had received countervailable subsidies.
Commerce’s only reason for accepting the Thai Ajinomoto statements was that there were no
other financial statements left on record. I&D Mem. at 17. That was a conundrum created by
Commerce itself, when the agency chose to preemptively reject the Thai Fermentation
statements. Rather than fashioning itself a pigeonhole by considering and then rejecting the Thai
Fermentation statements before ever reaching the Thai Ajinomoto statements, Commerce should
have compared the two side-by-side. On remand, Commerce must explain why, on the whole,
the Thai Ajinomoto statements were a better source than the Thai Fermentation statements.
Fufeng also claims that Commerce’s decision to reject the Thai Fermentation financial
statements was out of accord with a past practice of only rejecting incomplete financial
statements when the missing information is “vital.” Pl. Fufeng’s Br. 17–22. As already noted,
an agency’s consistent practice can bind it to explain departures. See, e.g., Consol. Bearings,
348 F.3d at 1007. But Commerce has not bound itself to a practice of only rejecting financial
statements when they are missing vital information. To be sure, Commerce has occasionally
characterized its rule as such—even before this court. Ass’n of Am. Sch. Paper Suppliers v.
United States, 35 CIT __, __, 791 F. Supp. 2d 1292, 1304 (2011) (recounting Commerce’s
argument that the agency had the aforementioned past practice, wherein Commerce cited
Galvanized Steel Wire at 23,551). Notwithstanding these characterizations, the fact remains that
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 14 of 36

Consol. Court No. 13-00288 	Page 15
Commerce has often rejected incomplete financial statements without finding that the statements
lacked vital information. See id. at __ & n.15, 791 F. Supp. 2d at 1302–03 & n.15 (citing several
such instances). These examples show that Commerce does not really reject financial statement
as sparingly as Fufeng claims, even though Commerce has sometimes suggested otherwise.
And, perhaps precisely for this reason, this court has before declined the invitation to tie
Commerce’s hands to a practice of rejecting incomplete financial statements only when they lack
vital information. See id. at 1304 (holding only that Commerce does not have a practice of
always rejecting incomplete financial statements).
In keeping with this court’s own past rulings, then, the court will not now hold that
Commerce rejects financial statements only when missing vital information. As such, on
remand, Commerce will not be bound to either accept the Thai Fermentation financial statements
or else to specifically find the statements to be lacking vital information: Rather, Commerce’s
only duty will be to compare and contrast the Thai Fermentation and Thai Ajinomoto financial
statements, and to explain why the Thai Ajinomoto statements constitute a better source. 7
7 Alternatively, if Commerce finds that the Thai Fermentation statements are missing “vital information,”
then Commerce should follow its past practice of rejecting such statements. See, e.g., Wooden Bedroom Furniture
from the People’s Republic of China, 71 Fed. Reg. 70,739 (Dep’t Commerce Dec. 6, 2006) (final results of new
shipper review) (“Wooden Bedroom Furniture”) and accompanying I&D Mem. at cmt. 2. (Although Commerce
does not have a past practice of only rejecting financial statements that are missing key sections of vital information,
see Ass’n of Am. Sch. Paper Suppliers, 35 CIT at __ & n.15, 791 F. Supp. 2d at 1302 & n.15, Commerce does have a
past practice of rejecting those statements that are missing such information. Id.) In keeping with this practice,
Commerce has often deemed financial statements to be unusable when they are missing all or many accounting
notes. See Floor-Standing Metal-Top Ironing Tables and Certain Parts Thereof from the People’s Republic of
China, 77 Fed. Reg. 14,499 (Dep’t Commerce Mar. 12, 2012) and accompanying I&D Mem. at cmt. 2; Certain Steel
Nails from the People’s Republic of China, 75 Fed. Reg. 34,425 (Dep’t Commerce June 17, 2010) (final new shipper
review) and accompanying I&D Mem. at cmt. 4; Wooden Bedroom Furniture and accompanying I&D Mem. at cmt.
2; Silocomanganese from Kazakhstan, 67 Fed. Reg. 15,535 (Dep’t Commerce Apr. 2, 2002) (final determ.) and
accompanying I&D Mem. at cmt. 3.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 15 of 36

Consol. Court No. 13-00288 	Page 16
III. 	Commerce’s Selection of the World Bank’s Doing Business 2013: Trading Across
Borders Thailand as its Surrogate Source for Valuing Fufeng’s Truck Freight Is
Supported by Substantial Evidence and Otherwise in Accordance with Law
Next, Fufeng challenges Commerce’s decision to value Fufeng’s truck-freight costs using
data from Doing Business 2013 rather than competing data from a website called Dxplace. To
select Doing Business 2013, Commerce considerations included a multifactor test that the
department has established through past practice. See Xiamen Intern. Trade and Indus. Co., Ltd.
v. United States (XITIC), 37 CIT __,__, 953 F. Supp. 2d 1307, 1312–13 (2013): Commerce
concluded that Doing Business 2013 better satisfied this test than the competing data from
Dxplace. Commerce reached this conclusion by tallying each factor in the multifactor test that
the data sources met or did not meet. Because Doing Business 2013 met more factors,
Commerce selected it. I&D Mem. 38–39.
Fufeng claims that Commerce’s tallying approach to the multifactor test was unsupported
by substantial evidence because it ignored qualitative differences in the degree to which Doing
Business 2013 and the Dxplace data satisfied one of the test’s factors. Fufeng also claims that
Commerce’s selection of Doing Business 2013 was contrary to past agency practice. Pl.
Fufeng’s Br. 24–36.
The court rejects Fufeng’s claims. Commerce’s tallying approach was reasonable in this
particular case. And Commerce’s decision-making did not run afoul of past practice, because
Fufeng has not established that there was any practice.
A. 	Background
In its preliminary determination, Commerce used a report called Costs of Doing Business
in Thailand as its source for surrogate truck-freight values. I&D Mem. at 38. But both Fufeng
and Kelco objected to the use of this report, and put on record alternate data sources: Fufeng, the
Dxplace data, and Kelco, the Doing Business 2013 report. So, to make its final determination,
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 16 of 36

Consol. Court No. 13-00288 	Page 17
Commerce had to choose which of the two sources to use. After administrative briefing on the
matter, Commerce concluded that Doing Business 2013 was the better source because it met
more factors of the multifactor test than did Dxplace. See id. at 38–39.
In its multifactor test, Commerce considers whether each source (1) provides a broad
market average covering a range of prices, (2) is publicly available, (3) is specific to the input in
question, (4) is tax and duty exclusive, and (5) is contemporaneous with the review period.
XITIC, 37 CIT at __, 953 F. Supp. 2d at 1312–13. Applying this test, Commerce found that
“[b]oth the 2010 Dxplace and [Doing Business 2013] data satisfy the criteria of public
availability, broad market average and tax and duty exclusivity,” but that the Doing Business
2013 report was more contemporaneous than the Dxplace data. I&D Mem. at 38–39. In other
words, Doing Business 2013 satisfied four factors in the multifactor test, whereas the Dxplace
data satisfied only three.
Commerce also provided an extended explanation of why both data sets satisfied the
broad-market-average criterion:
[T]he Doing Business 2013: Thailand report provides information for the inland
freight cost of shipping a container on a route from Bangkok to the port 133
kilometers away. On the other hand, the 2010 Dxplace data provide price points
for three types of trucks from multiple companies and include the cost to ship
from Bangkok to 76 different cities throughout the country, yielding a total of 228
price points. 	However, the Dxplace data come from a single date in June 2010
and it is unclear if these prices are six-month averages or a snapshot in time.
Additionally, it appears that the Dxplace website is still currently used for
shipping rates, but no other historical data are provided. . . . 	Additionally, as
stated in Certain Polyester Staple Fiber, the Department prefers Doing Business
2013: Thailand despite the fact that it provides freight costs solely from the main
city to the port because it reflects freight costs for multiple vendors and users (i.e.,
shipping lines, customs brokers and banks).
I&D Mem. at 38 –39 (footnotes omitted). Put more succinctly, Commerce believed that both
data sets satisfied the broad-market-average criterion, but with different kinds of breadth.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 17 of 36

Consol. Court No. 13-00288 	Page 18
Dxplace data satisfied the broad-market-average criterion because it included multiple truck
types, multiple companies, and multiple routes. On the other hand, it was unclear whether the
Dxplace data was temporally broad. Doing Business 2013 satisfied the broad-market-average
factor because it included multiple vendors and users.
B. 	Analysis
In making its substantial-evidence claim, Fufeng argues that Commerce’s tallying
approach ignored differences in the degree to which Doing Business 2013 and the Dxplace data
satisfied the broad-market-average factor. Pl. Fufeng’s Br. 29–36. According to Fufeng
Dxplace featured a much broader market average than Doing Business 2013, but tallying ignored
this strength. As for its past-practice claim, Fufeng cites three past investigations where
Commerce took a different approach, and argues that these investigations evidence binding past
practice. Id. at 32–33; Pl. Fufeng’s Reply Br. 15–16, ECF No. 59.
Both of Fufeng’s claims fail. With respect to Fufeng’s substantial-evidence claim, the
court cannot hold that Commerce’s tallying approach was unreasonable in this particular case.
As a general matter, “Commerce has not identified a hierarchy among the[ multifactor test’s]
factors, and the weight accorded to a factor varies depending on the facts of each case.” XITIC,
37 CIT at __, 953 F. Supp. 2d at 1313. The tallying approach was reasonable in this case
because the supposed difference in the extent to which the two data sources satisfied broad
market average was not as extreme as Fufeng claims. To be sure, Dxplace was far more
geographically broad than Doing Business 2013, and Commerce acknowledged as much. I&D
Mem. at 39. But it was unclear whether Dxplace featured temporal breadth, and Doing Business
2013 boasted breadth of a different kind: it tracked “multiple vendors and users.” Id. Therefore,
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 18 of 36

Consol. Court No. 13-00288 	Page 19
the facts of this case do not clearly indicate a discrepancy in broad market average so grave that
Commerce could not take a tallying approach. 8
Fufeng’s past-practice claim also fails. Fufeng offers three investigations as evidence of
three different purported past practices that it argues contravene Commerce’s selection. Pl.
Fufeng’s Br. 32–33; Pl. Fufeng’s Reply Br. 15–16. But citing isolated investigations does not
prove the existence of past practices; it just proves that Commerce thought differently on
different facts at different times. Fufeng’s past-practice claim is therefore as unavailing as its
substantial-evidence claim, and Commerce’s decision to use Doing Business 2013 to value truck
freight stands.
IV. 	Commerce’s Decision to Value Fufeng’s Corn as a Direct Input, Rather Than to
Value Cornstarch as an Intermediate Input, Is Supported by Substantial
Evidence and Otherwise in Accordance with Law
Kelco claims that Commerce’s decision to value Fufeng’s corn as a factor of production
was unsupported by substantial evidence. According to Kelco, Commerce should have instead
8 Fufeng also supports its substantial-evidence claim with three other arguments: (1) Commerce
unreasonably ignored the specificity factor, a factor that would have favored Dxplace, Pl. Fufeng’s Br. at 30–31, (2)
Commerce unreasonably focused “almost entirely” on the contemporaneity factor, id. at 31–33, and (3) Commerce
unreasonably discussed its use of Doing Business 2013 in other investigations, id. at 33–36.
The court declines to address Fufeng’s first argument that Commerce unreasonably ignored the specificity
criterion. Fufeng did not exhaust its administrative opportunity to raise specificity as a concern. 28 U.S.C.
§ 2637(d); Yangzhou Bestpak Gifts & Crafts Co. v. United States, 716 F.3d 1370, 1381 (Fed. Cir. 2013); see Case
Br. of Neimenggu Fufeng Biotechnologies Co., Ltd., PD 401 (Mar. 12, 2013); Fufeng’s Administrative Rebuttal Br.
Fufeng argues that it had no opportunity to voice its objection, because Commerce did not adopt Doing Business
2013 until its final determination. Pl. Fufeng’s Reply Br. 12. But Fufeng was on notice that Doing Business 2013
was under consideration by Commerce, such that it even took administrative briefing as an opportunity to the
compare Doing Business 2013 with Dxplace. Fufeng’s Administrative Rebuttal Brief at 36–40. In its comparison,
Fufeng analyzed other of the factors in the multifactor test. There is therefore no reason Fufeng could not have
analyzed the specificity of Doing Business 2013 versus Dxplace. It did not do so then, so it cannot now.
Fufeng’s second argument fails because its premise is false. Commerce did not focus “almost entirely” on
the contemporaneity factor. Pl. Fufeng’s Br. 31–33. As discussed, contemporaneity was one of four factors
Commerce weighed. Furthermore, Commerce’s most in-depth explanation went to the broad-market-average
factor—not contemporaneity.
Fufeng’s third and final argument—that Commerce unreasonably discussed the fact that it had used Doing
Business 2013 in other investigations—also fails. Even assuming that Commerce should not have discussed its prior
use of Doing Business 2013, Commerce provided a de novo multifactor analysis of Doing Business 2013.
Commerce’s use of Doing Business 2013 was justified on the basis of this first-impression analysis.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 19 of 36

Consol. Court No. 13-00288 	Page 20
applied its intermediate-input method, treating the product that Kelco made out of its corn,
cornstarch milk, as an input. Commerce declined to apply the intermediate-input method
because Commerce found that neither of the method’s predicate elements was met, and Kelco
argues this finding was unreasonable. See Pl. Kelco’s Br. 18–20. The court rejects Kelco’s
claim.
A. 	Background
In determining what counts as a factor of production, Commerce’s general policy is to
follow producers’ actual production experience. See Drill Pipe from the People’s Republic of
China, 76 Fed. Reg. 1966 (Dep’t Commerce Jan. 11, 2011) and accompanying I&D Mem. at
cmt. 12. So, if a producer self-manufactures one of the products used to construct the export,
then Commerce will value the items used to manufacture the intermediate product as inputs. By
contrast, if a producer buys a necessary product readymade, then Commerce will value the
product itself as an input. But there is an important exception to this rule: the intermediate-input
method. Under the intermediate-input method, Commerce will occasionally treat a self-
produced product as an input even though it has been made in house. Id. Commerce applies this
exception when “it is clear that attempting to value the factors used in a production process
yielding an intermediate product would lead to an inaccurate result because a significant element
of cost would not be adequately accounted for in the overall factors buildup.” Id. In short,
Commerce uses the intermediate-input method when (1) certain costs are not accounted for in the
input buildup and (2) those costs are significant.
In administrative briefing, Kelco asked that Commerce apply the intermediate-input
approach to Fufeng’s cornstarch milk, which Kelco manufactured in house from off-grade corn it
had purchased. Case Brief of CP Kelco US, Inc. (“Kelco’s Case Br.”) at 4–10, PD 407–09 (Mar.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 20 of 36

Consol. Court No. 13-00288 	Page 21
13, 2013). Kelco argued that the intermediate-input method was appropriate because (1) the
overhead costs of wet-milling the corn (to produce the cornstarch milk) would not be captured in
the Thai Ajinomoto surrogate financial ratios, and (2) those wet-milling costs were significant.
As evidence of cornstarch milk’s significant, unaccounted-for costs, Kelco offered an Energy
Star report. But, notably, the Energy Star report detailed the costs of producing different
products: dried cornstarch (made by drying cornstarch milk), as well as more complex products
like ethanol and corn-based sweeteners. I&D Mem. at 50. Relatedly, Kelco suggested that
Commerce remedy the alleged undervaluation by treating dried cornstarch—not cornstarch
milk—as the input in Fufeng’s xanthan-gum-production process. See Kelco’s Case Br. at 4–10
(designating “cornstarch”—not cornstarch milk—as the proper intermediate input); see also I&D
Mem. at 49.
Commerce addressed Kelco’s intermediate-input proposal in its final determination.
Commerce first pointed out that, even were it to find both of the intermediate-input elements
met, the appropriate intermediate input to supply would be cornstarch milk, not dried cornstarch.
I&D Mem. at 49. Having so specified, Commerce then considered the elements of the
intermediate-input test.
Addressing the second element first, Commerce found no evidence that the overhead
costs of producing cornstarch milk were significant. To evaluate overhead-cost significance,
Commerce compared the amount of capital equipment and energy used to produce cornstarch
milk to the amount used to produce xanthan gum. I&D Mem. at 49. If producing cornstarch
milk used a low proportion of capital-equipment and energy, then the cornstarch-milk production
process was likely to be simple relative to the xanthan-gum process. And if cornstarch milk’s
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 21 of 36

Consol. Court No. 13-00288 	Page 22
production process was relatively simple, then the overhead costs of the process were likely to be
relatively low (or, in other words, insignificant).
Commerce obtained the capital-equipment and energy-usage figures needed for this
comparison from Fufeng’s data submissions. I&D Mem. at 49 n.215. That is, Commerce did
not accept Kelco’s suggestion that it use figures from the Energy Star Report. Commerce
explained that the Energy Star report addressed the production of a different product—dried
cornstarch—whose production carried additional overhead costs. Id. at 50. Looking to Fufeng’s
data submissions, then, Commerce found that “Fufeng’s starch-making facility require[d] less
capital equipment and less electricity to operate than d[id] the rest of the xanthan gum production
process”; in fact, producing cornstarch milk required less than four percent of the total energy
needed to produce xanthan gum. Id. at 49 & n.215 (citing Section D Resp. for Neimenggu
Fufeng Biotechnologies Co., Ltd. (“Section D Resp.”) at 5–6, PD 118 (Sept. 27, 2012)); see also
Def.-Intervenor Fufeng’s Resp. Br. in Opp’n to CP Kelco US, Inc.’s R. 56.2 Mot. for J. on
Agency R. 19, ECF No. 41 (“Def.-Intervenor Fufeng’s Resp. Br.”). Given these proportions,
Commerce could not conclude that the overhead costs to produce cornstarch milk were
significant.
Turning to the first element of the intermediate-input test, Commerce also found that the
overhead costs of producing cornstarch milk were accounted for in the Thai Ajinomoto surrogate
financial ratios. Id. at 49–50. Commerce admitted that Thai Ajinomoto did not have an
analogous starch-making process in its MSG plant (Thai Ajinomoto started with tapioca starch,
rather than with unprocessed cassava root), but noted that Thai Ajinomoto also faced a number
of production costs that Fufeng did not. Commerce’s point was that any failure of the Thai
Ajinomoto surrogate to match cornstarch-production costs specifically was offset by
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 22 of 36

Consol. Court No. 13-00288 	Page 23
overcompensation in other cost centers. In sum, Commerce concluded that applying the
intermediate-input exception was unwarranted, because neither of the two elements necessary to
trigger it had been met.
B. 	Analysis
Kelco now claims that Commerce’s decision to value corn as a direct input, rather than
cornstarch milk as an intermediate input, was unsupported by substantial evidence. According to
Kelco, record evidence demonstrated that both intermediate-input factors were met, such that
Commerce was bound by its intermediate-input method to treat cornstarch milk as a factor of
production. Pl. Kelco’s Br. 18–20. In making this claim, Kelco essentially repeats its
administrative claim below, except that it replaces cornstarch milk with cornstarch as the
intermediate input it seeks. 9
Kelco’s claim must fail, because Commerce had the support of substantial evidence when
it concluded that neither element of the intermediate-input test was met. With respect to the
second element, Commerce concluded that the overhead costs of cornstarch milk were
insignificant because producing cornstarch milk required less capital equipment, and far less
energy, than producing finished xanthan gum. I&D Mem. at 49; see also Def.-Intervenor
Fufeng’s Resp. Br. 19. Commerce’s logic—that a production process with relatively low capital-
equipment and energy costs was likely to be relatively simple, and therefore to bear insignificant
overhead costs—was reasonable.
9 It is far from clear that exhaustion doctrine permits Kelco to switch claims like this. Whether or not
Kelco meant to directly ask Commerce to use cornstarch milk as an intermediate input, it did not. Because Kelco
did not do so, Kelco’s license to seek such a remedy from this court is at best suspect. 28 U.S.C. § 2637(d).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 23 of 36

Consol. Court No. 13-00288 	Page 24
In any case, Kelco does not take issue with Commerce’s logical approach, at least not
directly. 10 Rather, Kelco second-guesses its factual underpinnings, suggesting that Commerce
was wrong to conclude that starch-making really uses less capital equipment and energy than the
rest of the xanthan-gum process. See Kelco’s Reply Br. 16. Kelco again commends the Energy
Star report, which suggests that the entire xanthan-gum-production process uses just 88% of the
energy that producing dried cornstarch does, and surmises that producing cornstarch milk must
be similarly energy intensive. But, as Commerce has already pointed out, this is question
begging. See I&D Mem. at 50. Dried cornstarch requires further processing than cornstarch
milk, and the energy consumption involved in producing dried cornstarch could well come from
this additional work. Kelco has provided no evidence to the contrary, and there is no reason to
doubt Commerce’s conclusion that the first element of the intermediate-input test was not met. 11
Even assuming Kelco had so undermined Commerce’s first-element conclusion,
Commerce’s overarching intermediate-input finding would still stand. That is because it was
reasonable for Commerce to find, under the first intermediate-input element, that the cost of
converting corn to cornstarch milk was assimilated in other NV factors. As discussed above,
Commerce found that the conversion cost was accounted for in the Thai Ajinomoto surrogate
10 In its reply, Kelco offers a number of alternative measures of the overhead costs of producing cornstarch
milk, which Kelco believes show such costs to be significant. Kelco references (1) the amount of capital equipment
used to produce cornstarch milk (apparently viewed in isolation, without comparison to the amount used to produce
xanthan gum), (2) the number of steps in the cornstarch-milk production process versus in the xanthan-gum process,
and (3) the amount of labor used in each process. Pl. Kelco’s Reply Br. 13–16. But Kelco failed to make any
argument about these indicia either before Commerce or in its lead brief before this court. Therefore, any potential
argument is both unexhausted and waived. See 28 U.S.C. § 2637(d); USCIT R. 81(l).
11 At oral argument, Kelco provided an additional argument against Commerce’s conclusion that producing
cornstarch milk was less energy intensive than the rest of xanthan-gum production. According to Kelco, Commerce
made this relativistic finding using energy data from the preliminary determination, even though it had changed its
energy methodology by the time the final determination came around. Had Commerce used its final-determination
energy data, says Kelco, it would not have found the energy costs of cornstarch production to be lower than general
production costs. Whether or not this argument has merit, Kelco raised it neither in its lead nor its reply briefing. It
is therefore waived. USCIT R. 81(l).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 24 of 36

Consol. Court No. 13-00288 	Page 25
financial ratios—not because those ratios included an analogous cost for tapioca production, but
instead because it included offsetting costs that Fufeng did not face. I&D Mem. at 49–50.
Nothing in Kelco’s briefing undermines this reasoning: Kelco emphasizes that the Thai
Ajinomoto ratios did have a tapioca-conversion cost in particular, Pl. Kelco’s Br. 19–20, but
Commerce never said it did, choosing instead to rely on other offsetting costs. Therefore, Kelco
has failed to convince the court that Commerce’s first-prong and second-prong findings were
unsupported by substantial evidence. Kelco’s substantial-evidence claim is accordingly rejected
in full.
V. 	Commerce’s Decision to Value Fufeng’s Corn as Corn Imported Under the Thai
Tariff Heading for Corn “Fit for Animal Feed” Is Supported by Substantial
Evidence and Otherwise in Accordance with Law
Kelco claims that Commerce’s decision to value Fufeng’s corn as corn imported under
the “fit for animal feed” heading of the Thai Tariff Schedule was unsupported by substantial
evidence. 12 According to Kelco, Commerce’s selection of the “animal feed” heading was
unreasonable because the [[
]]. Pl. Kelco’s Br. 21. Kelco further argues that Commerce unreasonably rejected
Kelco’s suggestion that Commerce consult the Thai Agricultural Standard for Maize (“Thai Corn
Standard”) before deciding how to value Fufeng’s corn. Pl. Kelco’s Br. 23–25. Kelco’s claim
fails because Commerce’s selection of the “animal feed” heading was reasonable.
A. 	Background
After determining that Fufeng’s corn could be valued as a direct input, Commerce was
left with the task of assigning a surrogate value for the input. In order to choose the surrogate
12 Commerce determined the value of corn imported under the “fit for animal feed” heading of the Thai
Tariff Schedule, as well as under other corn headings, by consulting the Global Trade Atlas. Prelim. Surrogate
Value Determination at 2, 7.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 25 of 36

Consol. Court No. 13-00288 	Page 26
value, Commerce needed to match the actual corn used by Fufeng with the value that it would
have in the surrogate country, Thailand. See 19 U.S.C. § 1677b(c)(1), (4); Pl. Kelco’s Br. 28
(describing the relevant corn as the “actual corn consumed by Fufeng”). Commerce’s source for
potential surrogate values was the Thai Tariff Schedule. According to that schedule, corn
imported under tariff heading 1005.90.90001, entitled “Maize Corn, Fit For Human
Consumption,” had an average unit value (“AUV”) of $0.897 per kilogram, (“/kg”), corn
imported under HTS 1005.90.90090, entitled “Maize Corn, Other,” had an AUV of $0.726/kg,
and corn imported under HTS 1005.90.90002, entitled “Maize Corn, Fit For Animal Feed,” had
an AUV of $0.114/kg. Fufeng’s Post-Prelim. Surrogate Value Rebuttal Submission (“Fufeng’s
Post-Prelim. Surrogate Values”) at Ex. 1, PD 387 (Mar. 5, 2013). Commerce had to choose
which heading’s AUV (or combination of heading AUVs) best captured the value of Fufeng’s
corn.
In its preliminary determination, Commerce valued Fufeng’s corn according to the AUV
of HTS 1005.90.90002, the heading designated to corn “fit for animal feed.” Prelim. Surrogate
Value Determination at 5. As noted, the “animal feed” AUV was $0.114/kg, the lowest of the
three tariff-heading AUVs. Fufeng’s Post-Prelim. Surrogate Values at Ex. 1. Commerce chose
this heading “based on Fufeng’s description of the corn it purchase[d].” Prelim. Surrogate Value
Determination at 5 (citing Supplemental Section D Resp. for Neimenggu Fufeng Biotechnologies
Co., Ltd. at 3–4 and Ex. SD-8, PD 257 (Nov. 28, 2012) (“Supplemental Section D. Resp.”)).
Fufeng had described its corn as “off grade,” and noted that the [[
]]. Supplemental Section D
Resp. at 3–4 and Ex. SD-8.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 26 of 36

Consol. Court No. 13-00288 	Page 27
Kelco objected to Commerce’s preliminary selection of the “animal feed” heading.
Kelco’s Case Br. at 10–26. Kelco argued that Fufeng’s corn description (upon which Commerce
had based its selection of the “animal feed” heading) was misleading. In Kelco’s own words,
Notwithstanding [[
]],
Fufeng has continued throughout this investigation to publicly refer to its corn as
“off-grade,” [[
]] By using the technical
term “off-grade” in its filings (while simultaneously disavowing that the term
“off-grade” conveys any information regarding the nature of its corn inputs,
[[
]], to position its corn inputs to be valued as
“feed-grade” by [Commerce] in its normal value calculation, in an attempt to
reduce [Fufeng’s] calculated margin.
Id. at 11–12. Put another way, Kelco was concerned that Fufeng had publicly described its corn
as “off grade” to imply that Fufeng’s corn [[
]]. Then, in its confidential submissions, Fufeng had [[
]]. Commerce had chosen to value Fufeng’s corn using the
“animal feed” AUV because of Fufeng’s deceptive description, not because the AUV actually
matched the quality of Fufeng’s corn. Commerce’s decision to use the “animal feed” heading
was therefore erroneous in Kelco’s view.
To remedy the problem, Kelco proposed valuing Fufeng’s corn through a weight-
averaging of the tariff-heading AUVs. Id. at 23. Each tariff-heading AUV would be weighted
according to how much of Fufeng’s corn could actually be imported under the relevant tariff
heading. See id. at 19–24. Kelco said Commerce could determine which corn matched which
heading by referring to a third source, the Thai Corn Standard. The Thai Corn Standard is a
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 27 of 36

Consol. Court No. 13-00288 	Page 28
voluntary certification standard: It sets out a series of four corn categories along with the criteria
corn-selling merchants must meet if they want their corn to be certified within a particular
category. Id. at 22 (citing Post-Prelim. Surrogate Value Submission at Ex. 10 §1 & tbl.2, PD 357
(Feb. 22, 2013); see also Pl. Kelco’s Br. 23–26. The categories do not expressly align with the
Thai tariff headings: That is, the categories are not labeled as covering corn for “human
consumption,” “animal feed,” and “other corn,” but are instead enumerated one through four.
Post-Prelim. Surrogate Value Submission at Ex. 10 tbl.2. However, the standard’s standalone
scope provision does specify that the standard covers corn “for human consumption, feed, [and]
food and feed raw materials,” and the table setting forth the corn categories bears the following
note: “Maize kernels for food or food raw material shall not be lower than class 2.” Id. at Ex. 10
§ 1 & tbl.2.
Kelco argued that the categories in the Thai Corn Standard correlated with the Thai tariff
headings, such that Commerce could determine which tariff heading corn would enter under by
categorizing the corn under the standard. Kelco’s Case Br. at 19–24. Kelco based this argument
on the standard’s scope provision and note: Reading those sources together, Kelco inferred that
the two more stringent certification categories were reserved for human-consumption corn, while
the two less stringent categories were for feed-grade corn. Id. at 22–23. So interpreted, the
standard’s categories dovetailed with the tariff headings. Corn certifiable under the standard’s
two more-stringent human-consumption categories would be imported under the “fit for human
consumption” heading, while corn certifiable under the feed categories would be imported as
corn “fit for animal feed.” Id. And corn too poor in quality to be certified under any of the Thai
Corn Standard’s four categories would be imported under the “other corn” heading. Id. Kelco
further argued that [[ 	]] of Fufeng’s corn would fail certification under any of
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 28 of 36

Consol. Court No. 13-00288 	Page 29
the Thai Corn Standard’s four categories. Id. Therefore, using the Thai Corn Standard to decide
what portion of Fufeng’s corn would enter under each tariff heading, and weight-averaging
accordingly, would result in a corn value [[ 	]] the AUV for [[
]]. See Fufeng’s Post-Prelim. Surrogate Values at Ex. 1.
In the final determination, Commerce stood by its choice to value Fufeng’s corn
according to the “animal feed” AUV, and accordingly rejected Kelco’s proposed weight-
averaging approach. See Fufeng Final Determination Analysis Mem. at 9–10. Commerce
explained that, contrary to Kelco’s contention, it had not selected the “animal feed” heading
based on a mistaken belief that Fufeng’s corn satisfied any tier of the Chinese commercial
standard (including the “off grade” or “feed grade” tiers). Rather, Commerce had chosen the
“animal feed” heading based on the fact that the “[[
]].” Id. at 9. In other words,
Commerce intended the “animal feed” AUV to reflect the fact that Fufeng’s corn was of such
poor quality that it did not satisfy the Chinese commercial standard in any respect.
Commerce further explained that Kelco’s proposed weight-averaging approach was
“distortive.” Id. As just noted, Kelco’s methodology resulted in a corn value [[ 	]]
the AUV of corn imported under the [[ 	]]. Yet Commerce did not know
what corn was actually imported under this heading. Id. And Commerce did not agree with
Kelco that the Thai Corn Standard proved that the “other corn” heading was for low-quality corn
that could not enter under the “human consumption” or “animal feed” headings: It was not clear
to Commerce that there was any correlation whatsoever between the Thai Corn Standard and the
Thai tariff headings—much less one whereby the “human consumption” and “animal feed” tariff
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 29 of 36

Consol. Court No. 13-00288 	Page 30
headings were reserved for standard-certifiable corn and the “other corn” heading covered only
uncertifiable, lower quality remains. Id. As Commerce put it,
Petitioner has not demonstrated that [[
]] is reasonable or
that there is a correlation between the [[
]]. . . .
[R]ecord evidence shows that the[] standard[ is] voluntary. There is also no record
evidence on what type of corn is required to produce xanthan gum in Thailand, so
the 	voluntary Thai 	standard[ 	is] 	not 	determinative 	and 	do[es] 	not support
Petitioner’s argument of using the[] standard[] as a basis for selecting the
appropriate [surrogate value] for Fufeng’s corn input.
Id. at 9–10. Commerce accordingly disregarded the Thai Corn Standard and the weight-
averaging approach, instead continuing to value Fufeng’s corn using the “animal feed” AUV.
B. 	Analysis
Kelco now claims that Commerce’s decision to value Fufeng’s corn using the “animal
feed” AUV was unsupported by substantial evidence. Pl. Kelco’s Br. 21–29. Kelco makes two
arguments in support of this claim: (1) Commerce’s selection of the “animal feed” AUV was
unreasonable because the [[
]], Pl. Kelco’s Br. 21; see id. at 22–23, 27–29, and (2) Commerce’s decision to
reject the Thai Corn Standard was unreasonable because Commerce’s stated reasons for doing so
were inadequate, id. at 26–28. Both arguments fail.
Kelco’s first argument fails because Commerce selected the “animal feed” AUV
precisely to account for the low quality of Fufeng’s corn, and Commerce’s decision to do so was
reasonable. Fufeng Final Determination Analysis Mem. at 9–10. Put another way, Commerce
did not select the “animal feed” AUV because it mistakenly believed that Fufeng’s corn was of
sufficient quality to be fed to animals; rather, Commerce selected the AUV because it matched
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 30 of 36

Consol. Court No. 13-00288 	Page 31
the low quality of Fufeng’s corn. Id. This made sense, because the “animal feed” AUV, at
$0.114/kg, was the lowest of the three available AUVs. Fufeng’s Post-Prelim. Surrogate Values
at Ex. 1. By valuing Fufeng’s corn using the “animal feed” AUV, Commerce matched corn of
the lowest quality with the lowest available AUV. This approach was reasonable, such that
Kelco’s first argument fails.
Kelco’s second argument—that Commerce’s stated reasons for rejecting the Thai Corn
Standard were inadequate—also fails. As already noted, Commerce rejected the Thai Corn
Standard because (1) it was not clear to Commerce that there was any correlation between the
Thai Corn Standard and the Thai tariff headings, (2) the Thai Corn Standard is voluntary, and (3)
there was no record evidence of the type of corn used to produce xanthan gum in Thailand.
Fufeng Final Determination Analysis Mem. at 9–10. As to the first of these reasons, Kelco
argues that the correlation between the Thai Corn Standard and the Thai tariff headings should
have been clear to Commerce because the standard offered specific corn criteria where the
headings did not. Pl. Kelco’s Br. 27. As to the second, Kelco argues that the standard’s
voluntariness did not prevent it from being the “most relevant standard available” for choosing
between tariff-heading AUVs. Id. at 26. Finally, as to Commerce’s third reason, Kelco argues
that information on Thai corn was irrelevant to Commerce’s inquiry: Commerce’s task was to
determine the surrogate value of Fufeng’s Chinese corn. Id. at 27–28.
Kelco’s second argument fails because even the first of Commerce’s reasons for rejecting
the Thai Corn Standard, standing alone, was adequate to support Commerce’s decision: There
was no clear correlation between the Thai Corn Standard and the Thai tariff headings, and
therefore no reason to use the standard to decide which tariff-heading AUV applied to Fufeng’s
corn. According to Kelco, the Thai Corn Standard and the Thai tariff headings dovetailed in
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 31 of 36

Consol. Court No. 13-00288 	Page 32
such a way that the “other corn” heading was reserved for lowest quality corn like the [[
]]. Pl. Kelco’s Br. 23–25; Kelco’s Case Br. at 22–23. Yet the “other corn”
AUV was $0.726/kg, nearly six-and-a-half times the “animal feed” AUV that Commerce
ultimately used to value Fufeng’s corn. Fufeng’s Post-Prelim. Surrogate Values at Ex. 1. If the
Thai Corn Standard in fact correlated with the tariff headings in the way that Fufeng claimed—
such that the “other corn” heading was reserved for lowest quality corn—then it is unclear why
the “other corn” AUV would be so high. As such, there was no clear correlation between the
Thai Corn Standard and the Thai tariff headings. Kelco’s second substantial-evidence argument
therefore fails, and Commerce’s decision to value Fufeng’s corn using the “other corn” AUV
survives this court’s review. 13
VI. 	Commerce’s Requests for Remand to Reconsider its Methodology for Allocating
Energy at Fufeng’s Neimenggu Facility Are Granted
Both Fufeng and Kelco take issue with Commerce’s methodology for allocating energy at
Fufeng’s Neimenggu facility. Fufeng’s Neimenggu facility self-produced energy, such that
Commerce chose to value Neimenggu’s energy inputs as factors of production to the extent that
Neimenggu used its energy for xanthan gum. See Section D Resp. at 6. Valuing the inputs was
complicated by the fact that Neimenggu self-produced two different kinds of energy, steam and
electric. Commerce tried one methodology in its preliminary determination, but eventually
chose a different path in the final determination: The agency (1) calculated the rate at which
Neimenggu consumed energy inputs in the production of all steam and electric energy (whether
used for xanthan gum or for other production), (2) calculated the rate at which Neimenggu
13 Kelco also raises additional claims, in its reply brief only: (1) Past practice and substantial evidence
required Commerce to mimic the tariff-classification process when choosing which heading to use for valuing
Fufeng’s corn, and (2) past practice required Commerce to value Fufeng’s corn input using a simple (not weighted)
average of the available tariff headings. Pl. Kelco’s Reply Br. 16–17, 19–20. Kelco failed to raise either claim any
time before reply, so both are waived, neither exhausted. 28 U.S.C. § 2637(d); USCIT R. 81(l).
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 32 of 36

Consol. Court No. 13-00288 	Page 33
consumed steam and electric energy in the production of xanthan gum, and (3) multiplied the
rates from (1) and (2) together to calculate the rate at which Neimenggu consumed energy inputs
in the production of xanthan gum. Fufeng Final Determination Analysis Mem. at 2–3.
Commerce valued the energy inputs according to the rate from (3). Id.
Both Fufeng and Kelco take issue with Commerce’s calculation of the xanthan-gum
energy-consumption rate. Pl. Fufeng’s Br. 6–13; Pl. Kelco’s Br. 8–10. To calculate the rate,
Commerce divided the amount of energy used to produce Neimenggu’s xanthan gum by the
amount of unfinished xanthan gum that Neimenggu produced. In the numerator (energy used to
produce xanthan gum), Commerce included all of the energy consumed at Neimenggu’s
cornstarch workshop, even though only some of Neimenggu’s cornstarch was actually used in
xanthan-gum production. This, according to Fufeng, was inaccurate.
For its part, Kelco points out that Fufeng’s Neimenggu facility manufactured both
finished and unfinished xanthan gum: That is, Neimenggu did not finish all of the gum that it
started producing, but instead sent some xanthan gum to another facility (Shandong) for
finishing. Kelco argues that Commerce therefore should have used the finished-gum amount as
the denominator for its xanthan-gum energy-consumption rate. Pl. Kelco’s Br. 8–10.
In the alternative, Kelco raises a procedural claim. Kelco argues that Commerce never
had the procedural chance to address its substantive grievance. Id. at 10–11. This is because
Commerce changed its methodology in the final determination, leaving Kelco without a chance
to brief the merits. When Kelco tried to raise concerns through a 19 C.F.R. § 351.224(g)
ministerial-error submission, Commerce rejected the claim as a substantive grievance. Thus,
according to Kelco, its first opportunity to fully debate Commerce’s methodology was before
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 33 of 36

Consol. Court No. 13-00288 	Page 34
this court. Kelco ventriloquizes concern for Commerce, arguing that the agency should have a
chance to address its substantive argument before this court does.
Commerce requests voluntary remand to consider both claims, without presently
admitting error as to either. Gov’t Resp. Br. 32–33; Def.’s Mot. for Partial Voluntary Remand,
ECF No. 66. Neither Kelco nor Fufeng object to the substantive grounds of the other’s claim.
Pl. Fufeng’s Reply Br. 1–2 (“Kelco did not address [our remand request] at all in its
[r]esponse.”); Def.-Intervenor Fufeng’s Resp. Br. 5–6. When Commerce requests remand to
“reconsider its previous position” but does not admit error, whether to remand is in the court’s
discretion. 	SKF USA Inc. v. United States, 254 F.3d 1022, 1029 (Fed. Cir. 2001). However,
remand will generally be granted so long as Commerce has a “substantial and legitimate”
concern. The court holds that Commerce’s concerns as to both claims are substantial and
legitimate for the reasons set forth in Kelco’s and Fufeng’s briefing, respectively, and so
remands each to Commerce for reconsideration.
On remand, Commerce should consider whether or not it was appropriate to include all of
the energy consumed at Neimenggu’s cornstarch workshop in the numerator of Neimenggu’s
xanthan-gum energy-consumption rate. If so, Commerce should explain why. Commerce
should also consider whether or not it was appropriate to use the full amount of unfinished
xanthan gum produced at Neimenggu as the denominator of Neimenggu’s xanthan-gum energy-
consumption rate and, if so, explain why.
CONCLUSION AND ORDER
After carefully reviewing the parties’ briefs and the administrative record, the court
remands Commerce’s selection of the Thai Ajinomoto financial statements over the Thai
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 34 of 36

Consol. Court No. 13-00288 	Page 35
Fermentation statements for further explanation by Commerce. The court also remands
Commerce’s energy allocation. The court sustains Commerce’s reasoning in all other respects.
Upon consideration of all papers and proceedings herein, it is hereby:
ORDERED that the final determination of the International Trade Administration,
United States Department of Commerce (“Commerce”), published as Xanthan Gum from the
People’s Republic of China, 78 Fed. Reg. 33351 (Dep’t Commerce June 4, 2013) (final determ.),
as amended by Xanthan Gum from the People’s Republic of China, 78 Fed. Reg. 43,143 (Dep’t
Commerce July 19, 2013), be, and hereby is, REMANDED to Commerce for redetermination; it
is further
ORDERED that Plaintiffs’ Rule 56.2 Motions for Judgment on the Agency Record be,
and hereby are, GRANTED as provided in this Opinion and Order; it is further
ORDERED that Commerce must issue a redetermination (“Remand Redetermination”)
in accordance with this Opinion and Order that is in all respects supported by substantial
evidence, in accordance with law, and supported by adequate reasoning; it is further
ORDERED that Commerce shall reevaluate whether the Thai Ajinomoto or Thai
Fermentation financial statements constitute the better source for surrogate financial ratios,
explicitly comparing the imperfection in the Thai Ajinomoto statements (evidence of subsidies)
with that in the Thai Fermentation statements (incompleteness), and shall recalculate the
surrogate financial ratios consistent with this decision; it is further
ORDERED that Commerce shall reevaluate whether or not it was appropriate to include
all of the energy consumed at the Neimenggu facility’s cornstarch workshop in the numerator of
Neimenggu’s xanthan-gum energy-consumption rate, and shall recalculate Neimenggu’s energy-
input factor-of-production values consistent with this decision; it is further
ORDERED that Commerce must reevaluate whether or not it was appropriate to use the
full amount of unfinished xanthan gum produced at the Neimenggu facility as the denominator of
Neimenggu’s xanthan-gum energy-consumption rate, and shall recalculate Neimenggu’s energy-
input factor-of-production values consistent with this decision; it is further
ORDERED that Commerce shall recalculate Fufeng’s weighted-average dumping
margins consistent with any recalculation of the surrogate financial ratios, and of Neimenggu’s
energy-input factor-of-production values; it is further
ORDERED that Commerce shall have ninety (90) days from the date of this Opinion and
Order in which to file its Remand Redetermination, which shall comply with all directives in this
Opinion and Order; that the Plaintiff and Defendant-Intervenor shall have thirty (30) days from
the filing of the Remand Redetermination in which to file comments thereon; and that the
Defendant shall have thirty (30) days from the filing of Plaintiff and Defendant-Intervenor’s
comments to file comments.
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 35 of 36

Consol. Court No. 13-00288 	Page 36
/s/ Richard W. Goldberg
Richard W. Goldberg
Senior Judge
Dated: March , 2015
New York, New York
Case 1:13-cv-00288-RWG Document 78 Filed 04/08/15 Page 36 of 36