Shenzhen Xinboda Industrial Co. v. United States

976 F. Supp. 2d 1333, 2014 CIT 45, 36 I.T.R.D. (BNA) 267, 2014 Ct. Intl. Trade LEXIS 44

• Court: United States Court of International Trade
• Decided: April 16, 2014
• Docket: Slip Op. 14-45; Court 11-00267
• Status: Published

Opinion

OPINION

RIDGWAY, Judge:

In this action, Plaintiff Shenzhen Xinboda Industrial Co., Ltd. (“Xinboda”) — -an exporter of fresh garlic — contests the final results of the U.S. Department of Commerce’s fifteenth administrative review of the antidumping duty order covering fresh garlic from the People’s Republic of China (“PRC”). See Fresh Garlic from the People’s Republic of China: Final Results and Final Rescission, in Part, of the 2008-2009 Antidumping Duty Administrative Review, 76 Fed.Reg. 37,321 (Dep’t Commerce June 27, 2011) (“Final Results”); Issues and Decision Memorandum for the Final Results of the 15th Administrative Review of Fresh Garlic from the People’s Republic of China (June 20, 2011) (Pub.Doc. No. 176) (“Issues & Decision Memorandum”). 1

Pending before the Court is Xnboda’s Motion for Judgment on the Agency Record, contesting Commerce’s determinations as to the surrogate values for whole raw garlic bulbs, financial ratios, and labor, as well as the agency’s application of the “zeroing” methodology in calculating Xnboda’s dumping margin. See generally Memorandum in Support of Motion for Judgment on the Agency Record (“Pl.’s Brief’); Plaintiffs Reply Brief (“Pl.’s Reply Brief’).

The Government opposes Xnboda’s motion and maintains that the Final Results should be sustained in all respects, save one. See Defendant’s Memorandum in Response to Plaintiffs Rule 56.2 Motion for Judgment on the Agency Record at 1, 29-30 (“Def.’s Response Brief’). Specifically, the Government requests that the “zeroing” issue be remanded, to permit Commerce to reconsider and further explain its position. See id.

The DefendanWmtervenors — the Fresh Garlic Producers Association, Christopher Ranch, L.L.C., The Garlic Company, Valley Garlic, and Vessey and Company, Inc. (collectively, the “Domestic Producers”)— also oppose Xinboda’s motion and support the Government as to all four of Xinboda’s claims. See generally Defendant-Intervenors’ Response to Plaintiffs Rule 56.2 Motion for Judgment on the Agency Record at 2 (“Def.-Ints.’ Response Brief’).

Jurisdiction lies under 28 U.S.C. § 1581(c) (2006). 2 For the reasons set forth below, Xinboda’s Motion for Judgment on the Agency Record must be granted.

I. Background

Dumping occurs when goods are imported into the United States and sold at a price lower than their “normal value,” resulting in material injury (or the threat of material injury) to the U.S. industry. See 19 U.S.C. §§ 1673, 1677(34), 1677b(a); see also Union Steel v. United States, 713 F.3d 1101, 1103 (Fed.Cir.2013). The difference between the normal value of the goods and the U.S. price is the “dumping margin.” See 19 U.S.C. § 1677(35). When normal value is compared to the U.S. price and dumping is found, anti-dumping duties equal to the dumping margin are imposed to offset the dumping. See 19 U.S.C. § 1673; see also Union Steel, 713 F.3d at 1103.

When the exporting country is a market economy country, normal value is typically calculated using either the price in the exporting market {i.e., the price in the “market” where the goods are produced) or the cost of production of the goods. See 19 U.S.C. § 1677b. 3 However, where — as here — the exporting country has a non-market economy, there is often concern that the factors of production used to produce the goods at issue are under state control and that market data therefore may not be reliable indicators of normal value. See 19 U.S.C. § 1677(18)(A). In such eases, where the subject merchandise is exported from a non-market economy country and Commerce concludes that concerns about the sufficiency or reliability of the available data do not permit the normal value of the goods to be determined in the typical manner, Commerce “determinéis] the normal value of the subject merchandise on the basis of the value of the factors of production,” 4 including “an amount for general expenses and profit plus the cost of containers, coverings, and other expenses.” See 19 U.S.C. § 1677b(e)(l); see generally Ningbo Dafa Chem. Fiber Co. v. United States, 580 F.3d 1247, 1250-51 (Fed.Cir.2009) (briefly summarizing “factors of production” methodology).

In certain circumstances, where Commerce finds that the available information on the value of factors of production is not adequate for purposes of determining the normal value of the subject merchandise pursuant to the agency’s standard surrogate “factors of production” methodology (described above), Commerce determines the surrogate value of an “intermediate input” instead. See 19 U.S.C. § 1677b(c)(2). Under Commerce’s so-called “intermediate input methodology,” rather than valuing the various individual “upstream” factors of production that are used to produce an intermediate input, Commerce directly values the “downstream” intermediate input itself. 5

The antidumping statute requires Commerce to value factors of production “based on the best available information regarding the values of such factors” in one or more appropriate surrogate market economy countries. See 19 U.S.C. § 1677b(c)(l). The statute further requires that all data must, “to the extent possible,” come from market economy countries that are both (1) “at a level of economic development comparable to that of the nonmarket economy country” at issue and (2) “significant producers of comparable merchandise.” See 19 U.S.C. § 1677b(c)(4). In determining which data constitute the “best available information” for purposes of calculating surrogate values, Commerce seeks data that are “publicly available, product-specific, representative of a broad market average, tax-exclusive and contemporaneous with the [period of review].” Fresh Garlic from the People’s Republic of China: Preliminary Results of, Partial Rescission of, and Intent to Rescind, in Part, the 15th Antidumping Duty Administrative Review, 75 Fed.Reg. 80,458, 80,463 (Dep’t Commerce Dec. 22, 2010) (“Preliminary Results”); see also Issues & Decision Memorandum at 11-12. 6

The underlying antidumping order in this case, which dates back to 1994, covers imports of fresh garlic from the PRC, including whole garlic bulbs and peeled garlic cloves (the products exported by Xinboda). See Antidumping Duty Order: Fresh Garlic From the People’s Republic of China, 59 Fed.Reg. 59,209, 59,209-10 (Dep’t Commerce Nov. 16, 1994). This action involves the fifteenth administrative review of that antidumping order, initiated in December 2009. See Initiation of Antidumping and Countervailing Duty Administrative Reviews and Request for Revocation in Part, 74 Fed.Reg. 68,229, 68,230-31 (Dep’t Commerce Dec. 23, 2009). 7 The period of review is November 1, 2008 through October 31, 2009. See Final Results, 76 Fed.Reg. at 37,321. Commerce selected India as the primary surrogate country for purposes of this review (as it has in prior reviews), and used data from that country to calculate the surrogate values for all factors of production, with the sole exception of labor. See Preliminary Results, 75 Fed.Reg. at 80,462; Issues & Decision Memorandum at 12 (calculating surrogate values for whole raw garlic bulbs based on Indian data); id. at 17 (same as to water); id. at 20-21 (same as to surrogate financial ratios); id. at 28 (calculating surrogate wage rate based on data from eight countries, not including India).

In the course of the administrative review, Commerce compiled voluminous information concerning Xinboda and its operations, particularly the company’s exports of whole garlic bulbs and peeled garlic cloves from the PRC. Commerce similarly compiled detailed information on Zhenzhou Dadi Garlic Industry Co., Ltd. (“Dadi”), the affiliated processor/producer that supplied Xinboda with garlic products produced from raw garlic bulbs that Dadi purchased from local Chinese farmers. See Pl.’s Brief at 8, 22; Issues & Decision Memorandum at 19 (describing Dadi as “Xinboda’s supplier,” and “a non-integrated pro cessor that purchases its raw garlic input (rather than growing it from seed)”). 8 Dadi processed the whole raw garlic bulbs that it purchased — which had diameters of between 50 mm and 65 mm — into whole garlic bulbs and peeled garlic cloves for Xinboda, using relatively simple procedures involving principally manual labor. See generally Pl.’s Brief at 8-9, 22-23; Def.’s Response Brief at 3.

To produce whole fresh garlic, farmers deliver whole raw garlic bulbs to Dadi in large mesh bags sorted by the size of the garlic, as specified in the particular order(s) that Xinboda needs to fill. Pl.’s Brief at 8. Workers sitting at tables in a simple warehouse then rub off the outer skins of the whole raw garlic bulbs (to give the garlic a clean white appearance), cut or trim the roots and stems, place the bulbs into small mesh begs (typically holding three to five bulbs, depending on the customer), and affix the customer’s labels to seal the bags. Id. at 8, 22. Bags are then packed into cartons, ready for shipping. Id.

Peeled garlic cloves are similarly produced from whole garlic bulbs, which are also delivered in large mesh bags. PL’s Brief at 8, 22. The whole bulbs are run in bulk through a machine, to break the bulbs into individual whole cloves and to remove the skins. Id. Workers then hand sort the cloves at nearby tables in the warehouse, separating out cloves that are too small or that may be damaged or blemished. Id. at 9, 22. All cloves suitable for sale as “peeled garlic” are moved into a sterile environment, where they are washed, sorted, and dried, then packed directly into plastic jars, which are injected with a preservative gas and vacuum sealed. Id. at 9, 22-23. The jars are packed into cartons and remain in cold storage until they are transported by container truck to the port for export. Id. at 23. Despite the additional steps, the production of peeled garlic — like the production of whole garlic bulbs — is ultimately relatively simple and involves mostly manual labor. Id.

Xinboda’s administrative operations are modest as well, consisting of a small suite of rooms in a high-rise building, in addition to an administrative office in the provinces. See PL’s Brief at 23; PL’s Reply Brief at 6 n. 4. Xinboda’s sales process is similarly basic and straightforward. Xinboda does not develop or market its own brands and sells only a handful of products (i.e., garlic, onion shoots, and ginger) to its established customer base. PL’s Brief at 23. Its advertising and selling expenses are minimal. PL’s Reply Brief at 8.

Early in the course of the instant administrative review, Commerce concluded (as it had since the tenth review) that “garlic industry producers in the PRC do not generally track actual labor hours incurred for growing, tending, and harvesting activities and, thus, do not maintain appropriate records” which would allow Commerce to verify “the completeness and accuracy” of data reported for the numerous expenses incurred in growing and harvesting whole raw garlic bulbs. See Preliminary Results, 75 Fed.Reg. at 80,462. 9 Lacking the docu mentation necessary to ascertain the costs of growing, tending, and harvesting the subject whole raw garlic, Commerce used its intermediate input methodology to value “growing” and “harvesting” factors of production, as it had since the tenth review. See Issues & Decision Memorandum at 11; see also id. at 11-28; Jinan Yipin Corp. v. United States, 35 CIT -, - & n. 9, 800 F.Supp.2d 1226, 1236 & n. 9 (2011) (“Jinan Yipin II ”) (summarizing agency determination to use “intermediate input” methodology to value raw garlic bulbs for first time, in tenth administrative review). 10 Thus, in lieu of separately valuing each of the various individual growing and harvesting factors of production that are consumed in growing and harvesting a whole raw garlic bulb (ie., the leased land, garlic seed, water, pesticides, herbicides, fertilizer, plastic film, labor, and other “inputs” or commodities), Commerce instead sought to determine the value of the “intermediate input” — ie., the whole raw garlic bulb — at the “farm gate” (ie., before any “post-harvest” processing, and excluding operations such as sales, packing, and transportation). See Preliminary Results, 75 Fed.Reg. at 80,462-63; Jinan Yipin II, 35 CIT at - n. 38, 800 F.Supp.2d at 1257 n. 38 (quoting remand determination in which agency defined “farmgate” prices as prices for produce that goes “straight from the farm to the customer, without intermediary distributors”). 11

To value the whole raw garlic bulb input (ie., the “intermediate input”) at the “farm gate,” Commerce based its calculations in the Final Results on size-specific prices for garlic at the Azadpur APMC Market (located near Delhi and operated by the Azadpur Agricultural Produce Marketing Committee (“APMC”)), as published in the Azadpur APMC’s Market Information Bulletin. See Issues & Decision Memoran dum at 12. Commerce rejected the other potential sources of data on the record— including garlic pricing information included in the financial statements of Garlico Industries Limited (“Garlico”), an Indian purchaser, processor, and trader of garlic, onions, and other vegetables and related products — because those other sources of data did not specify the physical characteristics of the garlic that was priced. See id. at 12-13.

Specifically, to value the whole raw garlic bulbs purchased by Dadi that had a diameter of greater than 55 mm, the Final Results relied on non-contemporaneous Azadpur APMC price data for garlic classified as “grade S.A.” (or “Super-A”), which Commerce then inflated to be contemporaneous with the dates of the period of review. See Preliminary Surrogate Value Memorandum at 4 (Pub.Doc. No. 121); Issues & Decision Memorandum at 12 (explaining Commerce decision to use full year of data for “S.A.”-grade garlic prices in Final Results, rather than three months of data used in the Preliminary Results); PL’s Brief at 9-10; Def.’s Response Brief at 4, 10. Notwithstanding the agency’s established and longstanding policy favoring contemporaneous data (ie., data from within the period of review), Commerce used non-contemporaneous data to value Xinboda’s larger-bulbed garlic because the Azadpur APMC Market ceased use of the “S.A.”-grade classification in February 2008. See Issues & Decision Memorandum at 13 (noting that “grade super-A prices have not been reported since February 2008”); id. at 11-12 (highlighting, in two different places, agency preference for contemporaneous data). To value the whole raw garlic bulbs purchased by Dadi that were somewhat smaller (with a diameter of between 50 mm and 55 mm), the Final Results averaged the non-contemporaneous but indexed Azadpur APMC data for “S.A.”-grade garlic (described above) together with contemporaneous Azadpur APMC data for “A”-grade garlic (ie., data for “A”-grade garlic from within the period of review). See Final Surrogate Value Memorandum at 1 (Pub.Doc. No. 177); PL’s Brief at 9-10; Def.’s Response Brief at 4,10. 12

To calculate the surrogate value for post-harvest labor costs, Commerce averaged industry-specific data on wages and earnings from a group of eight countries that Commerce deemed to be both “significant producers” of comparable merchandise and “economically comparable” to the PRC, and which had also reported data under one particular revision of an international standard. See Issues & Decision Memorandum at 25, 28. However, that group of eight countries did not include India, because — although India reported contemporaneous data under the prior revision of the international standard — India’s reporting had not used the particular revision on which Commerce relied. See id. at 27. Citing “concerns that the industry definitions may lack consistency between different ... revisions” of the standard, Commerce declined to include the Indian data in its calculations in the Final Results. See id.

Because valuing the various direct inputs that are used in producing subject merchandise does not capture certain items that must also be factored into prices — specifically, manufaeturing/factory overhead, selling, general and administrative expenses (“SG & A”), and profit— Commerce calculates surrogate values for those three items using ratios that it derives from the financial statements of one or more companies that produce “comparable merchandise” in the surrogate market economy country. See 19 U.S.C. § 1677b(c)(l); 19 C.F.R. § 351.408(c)(4); see generally Dorbest Ltd. v. United States, 604 F.3d 1363, 1368, 1373-74 (Fed. Cir.2010) (“Dorbest IV”). In the administrative review at issue here, Xinboda’s surrogate financial ratios were drawn from the unconsolidated financial statements of Tata Global Beverages Limited, an Indian company that grows, processes, and sells its own trademarked and heavily-branded and -marketed coffee and tea products, including individually-packaged “Tetley” tea bags as well as “Tata Teas” (a so-called “Super Brand” in India). See Issues & Decision Memorandum at 20-22; Pl.’s Brief at 31-33, 35-36; PL’s Reply Brief at 7-8.

Commerce cited two reasons for selecting the financial statements of Tata Global over the five other sets of financial statements on the record. First, Commerce concluded, based on its review of the other companies’ financial statements, that all but one had received subsidies that the agency had previously determined to be countervailable. See Issues & Decision Memorandum at 21 (stating that Limtex, REI Agro, and LT Foods received subsidies under programs found to be countervailable); id. at 22 (same as to ADF). In light of Commerce’s practice of “disregarding] financial statements where [the agency has] reason to suspect that the company has received actionable subsidies,” Commerce therefore disregarded the financial statements of four of the five companies. See id. at 20. As to the remaining company, Garlico (the only one of the six companies that actually purchased and processed garlic), Commerce concluded that its operations were not comparable, based on the agency’s determination that “the majority of Garlico’s products are described as ‘dehydrated’ or ‘powder,’ ” as well as the determination that Garlico “act[ed] as a trading company (rather than a food processor) on nearly one quarter of its sales.” Id. at 22; see also id. at 12-13 (noting that Garlico purchased raw garlic bulbs).

Finally, in calculating Xinboda’s weighted-average dumping margin, Commerce applied its controversial “zeroing” methodology, which the agency has since abandoned. See Issues & Decision Memorandum at 31-33; Antidumping Proceedings: Calculation of the Weighted-Average Dumping Margin and Assessment Rate in Certain Antidumping Duty Proceedings; Final Modification, 77 Fed.Reg. 8101 (Dep’t Commerce Feb. 14, 2012); see generally Union Steel, 713 F.3d at 1103-04 (summarizing practice of “zeroing”). Thus, in Commerce’s calculations in the Final Results, negative dumping margins (i e., margins of sales of merchandise found to have been sold at non-dumped prices) were given a value of zero, and only positive dumping margins (ie., margins for sales of merchandise sold at dumped prices) were aggregated. In other words, sales that were not found to have involved dumping were not used to offset sales that were found to have involved dumping. See Issues & Decision Memorandum at 33 (explaining that, where the price in an export transaction at issue in the review exceeded normal value, the amount by which the price exceeded normal value did not offset dumping found in other transactions).

Based on the methodologies, analyses, calculations, and data summarized above, Commerce assigned Xinboda a weighed- average dumping margin of $0.06 per kilogram in the Final Results. See Final Results, 76 Fed.Reg. at 37,326. This action ensued.

II. Standard of Review

In an action reviewing an anti-dumping determination by Commerce, the agency’s determination must be upheld except to the extent that it is found to be “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i); see also NMB Singapore Ltd. v. United States, 557 F.3d 1316, 1319 (Fed.Cir.2009). Substantial evidence is “more than a mere scintilla”; rather, it is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Universal Camera Corp. v. Nat’l Labor Relations Bd., 340 U.S. 474, 477, 71 S.Ct. 456, 95 L.Ed. 456 (1951) (quoting Consol. Edison Co. v. Nat’l Labor Relations Bd., 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938)); see also Mittal Steel Point Lisas Ltd. v. United States, 548 F.3d 1375, 1380 (Fed.Cir.2008) (same).

Moreover, any evaluation of the substantiality of evidence “must take into account whatever in the record fairly detracts from its weight,” including “contradictory evidence or evidence from which conflicting inferences could be drawn.” Suramerica de Aleaciones Laminadas, C.A. v. United States, 44 F.3d 978, 985 (Fed.Cir.1994) (quoting Universal Camera Corp., 340 U.S. at 487-88, 71 S.Ct. 456); see also Mittal Steel, 548 F.3d at 1380-81 (same). That said, the mere fact that it may be possible to draw two inconsistent conclusions from the record does not prevent Commerce’s determination from being supported by substantial evidence. Am. Silicon Techs. v. United States, 261 F.3d 1371, 1376 (Fed.Cir.2001); see also Consolo v. Federal Maritime Comm’n, 383 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966).

Finally, while Commerce must explain the bases for its decisions, “its explanations do not have to be perfect.” NMB Singapore, 557 F.3d at 1319-20. Nevertheless, “the path of Commerce’s decision must be reasonably discernable,” to support judicial review. Id. (citing Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)); see generally 19 U.S.C. § 1677f(i)(3)(A) (requiring Commerce to “include in a final determination ... an explanation of the basis for its determination”).

III. Analysis

Xinboda challenges four aspects of the Final Results of the fifteenth administrative review. Xinboda first disputes Commerce’s calculation of the surrogate value for whole raw garlic bulbs. Xinboda contends that the data on which Commerce relied do not reflect “farm gate” prices for the “intermediate input” — whole raw garlic bulbs — that Commerce was supposed to value. Xinboda also protests Commerce’s use of non-contemporaneous data for “S.A.”-grade garlic. See generally PL’s Brief at 1-2, 8-22, 40; PL’s Reply Brief at 1-6. But see Dei’s Response Brief at 1, 6, 10-21; Def.-Ints.’ Response Brief at 2. Xinboda next challenges Commerce’s calculation of the surrogate wage rate. Specifically, Xinboda asserts that Commerce erred in using labor data from multiple countries, and should have relied on Indian data alone. Xinboda further argues that— even if it was permissible for Commerce to use data from multiple countries — Commerce failed to limit its “basket” of countries to those that were “significant producers” of comparable merchandise and also improperly excluded India based on the manner in which the country reported its data. See generally Pl.’s Brief at 2, 37-40; PL’s Reply Brief at 10-14. But see Def.’s Response Brief at 1, 6, 26-28; Def.Ints.’ Response Brief at 2. Xinboda similarly challenges Commerce’s calculation of surrogate financial ratios. According to Xinboda, Commerce’s justification for the financial statements that it selected is fundamentally flawed, and Commerce’s rejection of the financial statements that Xinboda favored was groundless. See generally PL’s Brief at 2, 22-36, 40; PL’s Reply Brief at 6-10. But see Def.’s Response Brief at 1, 6, 21-26; Def.-Ints.’ Response Brief at 2. As its fourth and final challenge to the Final Results, Xinboda contests Commerce’s application of the agency’s “zeroing” methodology in calculating Xinboda’s weighted-average dumping margin. See generally PL’s Brief at 1, 2-7, 40; PL’s Reply Brief at 14-15. But see Def.’s Response Brief at 1, 6, 29-30; Def.-Ints.’ Response Brief at 2.

Each of Xinboda’s arguments is analyzed in turn below.

A. Surrogate Value for Whole Raw Garlic Bulbs

As previously explained, in the administrative review at issue, rather than separately valuing each of the various individual “growing” and “harvesting” factors of production that are consumed in growing and harvesting whole raw garlic bulbs, Commerce instead employed its “intermediate input” methodology and sought to determine the value of the intermediate input itself — i.e., the whole raw garlic bulb — at the “farm gate.” See Issues & Decision Memorandum at 11; Def.’s Response Brief at 3; see generally section I, supra (explaining, inter alia, Commerce’s “intermediate input” methodology). Similarly, as it has in numerous other reviews, Commerce here relied upon price data from the Azadpur APMC Market to value the whole raw garlic bulbs. See Issues & Decision Memorandum at 11-12; see generally section I, supra. 13

In choosing the Azadpur APMC data over the other potential sources of surrogate value data on the record of this review, 14 Commerce emphasized the fact that the Azadpur APMC data that it used in the Final Results included prices for various grades of garlic, including grades super-A (“S.A.”) and “A.” According to Commerce, “garlic bulb sizes that range from 55 mm and above are Grade Super-A, and garlic bulb sizes that range between 40 mm and 55 mm are Grade A and Grade Super-A.” Preliminary Surrogate Value Memorandum at 4; Issues & Decision Memorandum at 12; Preliminary Results, 75 Fed.Reg. at 80,463 (noting that definitions of grade “A” and grade “S.A.” garlic used by Commerce in this review were “Consistent with [Commerce’s] findings in the twelfth [administrative review]”).

Commerce found the Azadpur APMC data to be the “best available information” notwithstanding the fact that data on “S A.’’-grade garlic have not been reported since early February 2008 (approximately nine months before the beginning of the period of review in this case). See Issues & Decision Memorandum at 12-13. 15 Commerce nevertheless concluded that, compared to the other data sources on the record, the Azadpur APMC prices that it chose to use were “much more similar to the inputs being valued and more accurately represented] a range of pricing during the [period of review] by providing size-specific pricing information.” Id. at 12-13.

The whole raw garlic bulbs that Dadi (Xinboda’s processor/producer) purchased for whole garlic production ranged from 50 to 65 mm, and from 50 to 55 mm for the production of peeled garlic. See PL’s Brief at 8; see also Def.’s Response Brief at 3. To value bulbs with a diameter of 55 mm or more, the Final Results relied on the Azadpur APMC data for “S.A.”-grade garlic for the period February 2007 through January 2008 (which Commerce inflated to the dates of the period of review using a garlic-specific wholesale price index). See Preliminary Surrogate Value Memorandum at 4; Issues & Decision Memorandum at 12 (explaining Commerce decision to use full year of data for “S.A.”-grade garlic prices in Final Results, rather than three months of data used in the Preliminary Results); PL’s Brief at 9-10; Def.’s Response Brief at 4, 10. To value garlic bulbs with a diameter of between 40 mm and 55 mm, the Final Results averaged the Azadpur APMC data for “S.A.”-grade garlic that is described above together with contemporaneous Azadpur APMC data for “A”-grade garlic (i.e., data for “A”-grade garlic from within the period of review). See Final Surrogate Value Memorandum at 1; PL’s Brief at 9-10; Def.’s Response Brief at 4, 10. The Final Results reflect values of 33.18 rupees per kilogram for “S.A.”-grade garlic, 23.52 rupees per kilogram for “A”-grade garlic, and 27.58 ru pees per kilogram for “A”-and “S.A.”grades combined. See PL’s Brief at 10.

Xinboda challenges the Final Results’ reliance on the Azadpur APMC data principally on two grounds. See generally PL’s Brief at 1-2, 8-22, 40; PL’s Reply Brief at 4-6. Xinboda first argues that the Azadpur APMC price data reflect a product that is much closer to a finished “retail” product than the “intermediate input” that Commerce assertedly seeks to value in this review. According to Xinboda, the Azadpur APMC data thus are not even close to “farm gate” prices. See PL’s Brief at 9-10, 11-19, 21-22; PL’s Reply Brief at 5-6; see also id at 1-4. In addition, Xinboda contests Commerce’s use of Azadpur APMC data for “S.A.”-grade garlic from outside the period of review, arguing that garlic of the size and quality previously designated as “S.A.”-grade was subsumed into “A”grade garlic as of early February 2008. In other words, Xinboda maintains that the Azadpur APMC price data for “A”-grade garlic that is contemporaneous with the period of review include prices for garlic bulbs that previously would have been classified as grade “S.A.”. See PL’s Brief at 10,19-21; PL’s Reply Brief at 4-5.

Xinboda argues that, in light of its challenges to the Azadpur APMC data, “[t]he most accurate approach” would be to calculate the surrogate value for the intermediate input — whole raw garlic bulbs — using averaged garlic price data from the 2009-2010 financial statements of the Indian garlic processor and trader Garlico, which Xinboda placed on the administrative record. PL’s Brief at 15, 21; Xinboda Surrogate Value Submission at Exh. 40, Schedules 1.(5) & I.(e) (Pub.Doc. No. 133) (Garlico financial statements for 2009-2010); see generally PL’s Brief at 9-10, 15, 16, 18-19; PL’s Reply Brief at 3. Xinboda contends that “Garlico’s experience more nearly matches Xinboda’s experience in the purchase of garlic at farm gate prices.” PL’s Brief at 16; see also id at 18-19; PL’s Reply Brief at 3, 6. 16 Alternatively, if Commerce continues to rely on Azadpur APMC data, Xinboda argues that — to calculate the surrogate value for the intermediate input, whole raw garlic bulbs, at the farm gate stage — the agency must use only fully-contemporaneous data, “commenc[ing] its valuation ... with the average of Grade A garlic,” 17 then “deductfing] 70%” from that figure “[t]o account for transportation and handling costs, interstate fees, commissions, and other markups” (to render the figure representative of an intermediate input at a “farm gate” price). PL’s Brief at 21-22; see also id. at 15; PL’s Reply Brief at 6.

The Government maintains that “Commerce reasonably relied upon the [Azadpur] APMC Bulletin data for grades A and Super-A garlic, because they were the only data on the record specific to the size of Xinboda’s raw garlic bulb inputs,” and that Commerce properly rejected “the less specific [Garlico] data advocated by Xinboda.” Def.’s Response Brief at 6, 10; see generally id at 10-21. According to the Government, Xinboda’s challenges to Commerce’s use of the Azadpur APMC data are lacking in merit. See generally id. at 6, 10-21. Both the Government and the Domestic Producers thus contend that, as to the surrogate value for whole raw garlic bulbs, Commerce’s Final Results should be sustained. See id. at 6, 21; Def.-Ints.’ Response Brief at 2.

1. Whether the Final Results Reflect “Farm Gate” Prices for an “Intermediate Input”

Xinboda claims that, in the Final Results, Commerce mis-applied the agency’s intermediate input methodology. See Pl.’s Brief at 11; see generally id. at 9-10, 11-19; Pl.’s Reply Brief at 1-4, 5-6. Xinboda emphasizes that Commerce here was supposed to determine a surrogate value for raw garlic bulbs (as an “intermediate input,” at the “farm gate” stage), reflecting only the “growing, tending, and harvesting costs” associated with the garlic that Xinboda exported. See Pl.’s Brief at 11; see also Pl.’s Reply Brief at 6. Instead, according to Xinboda, the Azadpur APMC prices inherently reflect certain expenses that Xinboda also separately reported, resulting in double-counting. See PL’s Brief at 12. Similarly, Xinboda argues that the Azadpur APMC prices inherently reflect a wide range of significant expenses incurred beyond the farm gate. See id. at 9-10, 11, 12-16; PL’s Reply Brief at 1-3, 5-6. Xinboda thus charges that the Azadpur APMC price data used in Commerce’s Final Results are “laden with additional costs that Xinboda [did] not pay,” that the data do not “in any way represent ] the growing, tending, and harvesting costs of Xinboda’s suppliers,” and that the data therefore are not representative of the value of the intermediate input at issue — i.e., whole raw garlic bulbs — at the farm gate. PL’s Brief at 11, 13, 16; see also id. at 14, 18-19, 21; PL’s Reply Brief at 3, 5-6.

Xinboda argues, for example, that undisputed record evidence establishes that garlic arriving at the Azadpur APMC market is essentially already “fully processed for retail consumption,” with “a fresh white appearance” and with the outside layers of the garlic already removed and the long stems already cut. See PL’s Brief at 12 (citing Declaration of Xinboda Research Consultant, Exh. 1 ¶ 9 (Survey of Garlic Offerings — Azadpur Market, New Delhi) (Pub.Doc. No. 138)); see also PL’s Brief at 14. The Azadpur APMC prices thus must incorporate the costs of such processing. According to Xinboda, these same processes — peeling away the outside layers of the garlic and cutting the long stems — are processes that Dadi performed for Xinboda at Dadi’s own facility. See id. at 12. Xinboda further asserts that it (Xinboda) was required to report to Commerce the labor hours and electricity usage for those tasks, and that Commerce ultimately added the value of that labor and electricity — together with a proportional figure for overhead (“SG & A”) costs — to Commerce’s calculated surrogate value for raw garlic bulbs, i.e., the Azadpur APMC prices. See id. Xinboda thus concludes that, because it separately reported to Commerce the costs of removing the outer layers and long stems of the garlic, and because the expense of such processes are already effectively “embedded” in the Azadpur APMC price, such expenses are, in essence, being double-counted and the Azadpur APMC data do not truly reflect prices at the farm gate. See id.

Commerce’s Final Results largely ignore Xinboda’s point, stating only that “information on the record speaks to the similarities of garlic entering the Azadpur market and the garlic entering Xinboda’s processing facilities.” Issues & Decision Memorandum at 15; see also Defi’s Response Brief at 18. But it is no answer to say — as the Issues & Decision Memorandum does — that both the garlic delivered to Dadi and the garlic arriving at the Azadpur APMC market are “pre-sorted by grade” and packaged in “large mesh sacks.” See Issues & Decision Memorandum at 15. Those facts are not in dispute, 18 but have no relevance to whether in fact Commerce’s use of the Azadpur APMC data effectively results in the double-counting of costs for processes such as the removal of garlic’s outer layers and long stems.

Xinboda similarly asserts that it was required to separately report to Commerce the distances and modes of transportation for transporting garlic from its suppliers to its own facility, and that Commerce ultimately added the value of that transportation to the agency’s surrogate value for whole raw garlic bulbs, ie., the Azadpur APMC prices. See Pl.’s Brief at 12. Xinboda argues, however, that the expenses of comparable transportation logically must already be reflected in the Azadpur APMC prices. See id. Xinboda thus contends that — much like the costs associated with processes such as the removal of garlic’s outer layers and long stems — these transportation costs too are basically being double-counted. See id. According to Xinboda, “[t]hese are ... costs that [Commerce] is well aware of and must subtract from the Azadpur market price.” Id. The Issues & Decision Memorandum is silent on this point.

More generally, Xinboda emphasizes that garlic arriving at the Azadpur APMC Market has been shipped for substantial distances, and that the expense of that transportation obviously must be embedded in the prices charged at the market. See PL’s Brief at 12-13,14,15,16, 21; PL’s Reply Brief at 5-6. 19 Commerce itself acknowledges that “the Azadpur data ... used [by Commerce] in this proceeding represent millions of kilograms of garlic sold from at least eight Indian states.” Issues & Decision Memorandum at 12 (emphasis added). By definition, the transportation of garlic to market takes place beyond the farm gate, belying the assertion that the Azadpur APMC data reflect farm gate prices. See PL’s Brief at 13; PL’s Reply Brief at 5-6; see generally Jinan Yipin II, 35 CIT at -, 800 F.Supp.2d at 1268-71 (discussing evidence concerning costs associated with transportation of garlic to Azadpur APMC Market, noting that Azadpur APMC data there indicated that “at least some of the garlic reflected in [the] data was transported for substantial distances” (as much as 100 miles or more), and concluding that fact of such transportation “undermines Commerce’s claims that the Azadpur APMC data reflect the price of large-bulb Indian garlic at the ‘farm gate’ and that those data ... are representative of the ‘intermediate input’ at issue”).

Xinboda also highlights evidence indicating that the Azadpur APMC prices used in the Final Results include significant “downstream expenses,” such as sums paid to “middlemen” or “intermediaries” including “commission agents, wholesalers and retailers to cover transportation, loading, unloading, storage, overheads, profits, etc.” that are associated with sales at such markets — still more expenses that logically should not be included in a surrogate value for an intermediate input at the farm gate stage. See PL’s Brief at 13-14; see gener ally id. at 13-16, 21-22; Pl.’s Reply Brief at 3, 5-6. 20 The Issues & Decision Memorandum inexplicably states that Xinboda failed to “placet] information on the record” regarding these “extra transportation and handling expenses.” Issues & Decision Memorandum at 15; see also Def.’s Response Brief at 18-19. Quite to the contrary, Xinboda mustered significant documentation to substantiate its claims.

As Xinboda notes, the administrative record in this matter includes the 2009-2010 Annual Report of the Indian Ministry of Agriculture’s Department of Agriculture & Cooperation (“AgriCoop”), 21 which advises that the country’s market system has become increasingly “restrictive and monopolistic” over time, such that “produce is required to be channeled through regulated markets and licensed traders” (Le., the “intermediaries” to which Xinboda refers), resulting in “an enormous increase in the cost of marketing.” See PL’s Brief at 13-14 (quoting Xinboda Surrogate Value Submission at Exh. 35); see also id. at 10; PL’s Reply Brief at 3. Xinboda reasons that the expenses associated with the use of such intermediaries and such “enormous” marketing costs must, as a matter of logic, be included in the Azadpur APMC market prices. See id. at 3, 5-6; see also PL’s Brief at 16. But such expenses have no place in the price that Commerce uses to value raw garlic bulbs as an intermediate input at the farm gate stage for purposes of this administrative review.

Other record evidence to the same general effect includes an article authored by the Director of India’s National Academy of Agricultural Research Management (“NAARM”) stating that the supply chains for agricultural products such as onions, tomatoes, and garlic are “inefficient, dominated by intermediaries.” See PL’s Brief at 14 (citing Xinboda Surrogate Value Submission at Exh. 34); see also PL’s Reply Brief at 3. The Director of NAARM further explained that “[s]tudies have shown that nearly 60-80% of the price consumers pay goes to commission agents, wholesalers and retailers to cover transportation, loading, unloading, storage, overheads, profits, etc.” See PL’s Brief at 14 (citing Xinboda Surrogate Value Submission at Exh. 34); see also id. at 15, 21; PL’s Reply Brief at 3. Another similar article, by a senior agricultural economist from Credit Rating and Information Services of India (“Crisil”), underscores “[t]he difference between the farm gate and retail prices” of onions and other similar vegetables in India and attributes that mark-up to “exploitation] by intermediaries.” See PL’s Brief at 14 (citing Xinboda Surrogate Value Submission at Exh. 34).

Xinboda points out that the involvement of intermediaries in sales at facilities such as the Azadpur APMC Market and the existence of associated additional fees and expenses are borne out by the Domestic Producers’ own Market Research Report, on which both Commerce and the Domestic Producers so heavily rely. See PL’s Brief at 15-16. As Xinboda notes, the Domestic Producers’ initial (2003) Market Research Report clearly distinguishes among “farm level,” wholesale, and retail sales. See id. at 15; Market Research on Fresh Whole Garlic in India (June 2003) at 20-21 (Pub.Doc. No. 131) (“2003 Market Research Report”). 22 Similarly, the October 2006 update to the Market Research Report states that an individual transporting produce out of a local APMC jurisdiction to a market such as the Azadpur APMC Market must pay a market fee to the local market at the local district’s exit checkpoint. See Pl.’s Brief at 15-16; Clarifications on Garlic Study (Oct. 2006) at 6 (Pub.Doc. No. 133) (“2006 Market Research Report Update”).

Xinboda cites the Garlico pricing data on the record as corroboration of Xinboda’s claims that the Azadpur APMC market prices reflect substantial expenses that Xinboda did not incur and which, moreover, should not be reflected in a surrogate value for whole raw garlic bulbs as an intermediate input at the farm gate. See Pl.’s Brief at 9-10, 15, 16, 18-19, 21; PL’s Reply Brief at 3, 6. According to Xinboda, “[djeducting the average 70% markup reported by [India’s National Academy of Agricultural Research Management] from the Grade A prices of garlic sold on the Azadpur [APMC] market during the [period of review] amounts to a farm gate price of 7.055 Rs/kg.” PL’s Brief at 15; see also id. at 21-22. Xinboda argues that this figure “comes very close to the average prices for raw garlic that Garlico paid” during the period of review, and thus constitutes further proof that the Azadpur APMC price data used in Commerce’s Final Results do not reflect farm gate prices for whole raw garlic bulbs, the intermediate input that Commerce sought to value. See id. at 15.

The Final Results fail even to acknowledge the evidence adduced by Xinboda (and outlined above), much less address it; and the Final Results brush off Xinboda’s concerns about additional fees and expenses embedded in the Azadpur APMC prices with the general proposition that Commerce “is not required to duplicate the exact experience of an exporter when calculating surrogate values.” Issues & Decision Memorandum at 14; see also Def.’s Response Brief at 17-18. As the Issues & Decision Memorandum further recognizes, however, Commerce is re quired to “select from among the available surrogate values those that permit the [agency] to calculate the most accurate dumping margin possible.” Issues & Decision Memorandum at 14. Further, as Jinan Yipin I emphasized, “when valuing an intermediate [input] in [a non-market economy] country case,” Commerce must “be mindful that ... it must find a surrogate representative of that intermediate product.” Zhengzhou Harmoni Spice Co. v. United States, 38 CIT 453, 472-73, 617 F.Supp.2d 1281, 1300 (2009) {“Jinan Yipin I ”). It is beyond cavil that, in a situation such as this, a surrogate value that reflects a level of trade that is beyond the farm gate and the intermediate input at issue cannot yield “the most accurate dumping margin possible.”

The record evidence that Xinboda has marshaled significantly undermines Commerce’s representation that the Azadpur APMC prices used in the Final Results reflect farm gate prices for whole raw garlic bulbs, the intermediate input in question. This matter therefore must be remanded to Commerce to permit the agency to analyze and respond to Xinboda’s arguments and evidence, and to reconsider the calculation of the surrogate value for whole raw garlic bulbs in light of those arguments and evidence (including, if necessary, making appropriate adjustments to the Azadpur APMC data in order to exclude specific costs or, alternatively, selecting another source of data). 23

2. Whether the Final Results Properly Relied on Non-Contemporaneous Data for Grade “S.A. ” Garlic

As explained above, to value the garlic that Dadi purchased that had a bulb diameter of 55 mm or more, the Final Results relied on Azadpur APMC data for “S.A.”grade garlic for the period February 2007 through January 2008 (which was then inflated to the dates of the period of review). See section I, supra. Similarly, to value the garlic that Dadi purchased that had a bulb diameter of between 50 mm and 55 mm, the Final Results combined the Azadpur APMC data for “S.A.”-grade garlic (described above) with contemporaneous Azadpur APMC data for “A”-grade garlic {ie., data for “A”-grade garlic from within the period of review). See id.

It is undisputed that the Azadpur APMC data for “S.A.”-grade garlic that Commerce used in the Final Results are not contemporaneous with the period of review. See Issues & Decision Memorandum at 12-13; Pl.’s Brief at 9; Def.’s Response Brief at 4. It is similarly undisputed that the Azadpur APMC Market Bulletin has not published prices for “S.A.” grade garlic since early February 2008. See Issues & Decision Memorandum at 13; PL’s Brief at 19; Def.’s Response Brief at 16.

What is squarely in dispute is whether, as Xinboda maintains, the Azadpur APMC price data for grade “A” garlic for the period of review at issue — ie., the contemporaneous data — include not only garlic with a bulb diameter of between 40 mm and 55 mm, but also garlic with a bulb diameter of 55 mm or more. In other words, what is squarely in dispute is whether grade “S.A.” garlic was, as Xinboda puts it, “subsumed” into grade “A” garlic as of February 2008. Thus, what is squarely in dispute is whether the Final Results’ use of Azadpur APMC data for “S.A.”-grade garlic not only was unnecessary, but, in fact, fundamentally distorted Commerce’s calculation of the surrogate value for whole raw garlic bulbs. PI.’s Brief at 10, 20-21; PL’s Reply Brief at 4; see generally Issues & Decision Memorandum at 13; PL’s Brief at 19-21; PL’s Reply Brief at 4-5; Def.’s Response Brief at 16-17.

In the Final Results, Commerce summarily dismissed Xinboda’s concerns, asserting that there is “no evidence ... on the record which clearly explains why grade super-A prices have not been reported since February 2008” and that the agency could not simply “assume that grade super-A prices have been subsumed under grade A prices.” See Issues & Decision Memorandum at 13; see also Def.’s Response Brief at 16-17. As Xinboda observes, however, it is of no real import why the Azadpur APMC Market ceased use of the “grade S.A.” classification. See PL’s Brief at 19-20. 24 Whatever the reason, the fact remains that the “S.A.” classification apparently was no longer in use at the time of the subject period of review. 25 Moreover, as Xinboda points out, uncontroverted record evidence means that Commerce need not “assume” anything concerning whether, in fact, garlic that previously would have been graded “S.A.” is now included within grade “A.” See id. Specifically, Xinboda’s research consultant attested on the record that Azadpur APMC Market vendors now sell garlic with bulb diameters as large as 65 mm under the classification grade “A.” See Declaration of Xinboda Research Consultant, Exh. 1 ¶¶ 5-6 (Survey of Garlic Offerings — Azadpur Market, New Delhi); see also PL’s Brief at 10, 19-20 (explaining, inter alia, that Xinboda research consultant visited the Azadpur APMC Market and surveyed all garlic vendors there); PL’s Reply Brief at 4-5. 26

The Final Results do not mention— much less refute — the report of Xinboda’s consultant. For its part, the Government seeks to minimize Xinboda’s evidence as the “self-serving statements of a researcher that Xinboda hired.” Def.’s Response Brief at 16. But the Government’s attempts to discredit the research consultant are both impermissible post hoc rationalization and ill-conceived. See Pl.’s Reply Brief at 4-5.

It is well-established that an agency determination cannot be sustained on the strength of a rationale supplied after the fact by litigation counsel. See Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168-69, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962). As the Supreme Court has explained, “an agency’s action must be upheld, if at all, on the basis articulated by the agency itself.” Motor Vehicle Mfrs. Ass’n, 463 U.S. at 50, 103 S.Ct. 2856. Moreover, even if Commerce itself had questioned the report of Xinboda’s research consultant as “self-serving,” that objection nevertheless would have been unavailing. The same accusations of self-interest could be leveled (arguably with even greater force) at the Market Research Report, which was commissioned and placed on the record by the Domestic Producers and on which both Commerce and the Domestic Producers have so heavily relied (in this and numerous other administrative reviews). Cf. Jinan Yipin II, 35 CIT at -, 800 F.Supp.2d at 1258-63 (rejecting Chinese producers’ attempts to discount the 2003 Market Research Report as a “private market study commissioned by [the Domestic Producers], which is neither an official nor an objective source”). Here, there is nothing whatsoever on the record to impugn the overall credibility of Xinboda’s research consultant or to cast doubt on the reliability of the specific information that was supplied. 27 Under these circumstances, it ill-behooves the Government to disparage the research consultant’s work as that of a mere “hired gun.”

In short, the existing evidence of record supports only one conclusion — that grade “S.A.” garlic was subsumed into grade “A” garlic as of February 2008. As such, Commerce’s use of the non-eontemporaneous prices for “S.A.”-grade garlic would have been both unnecessary and distortive. Specifically, if (as all record evidence indicates) the data for grade “A” garlic that were contemporaneous with the period of review included garlic with bulb diameters of up to 65 mm, there was no need for Commerce to use indexed non-contemporaneous data for grade “S.A.” garlic to value larger-bulbed garlic; the value of such larger-bulbed garlic would be already accounted for in the contemporaneous data for grade “A” garlic.

But, more importantly, if (as all record evidence indicates) the contemporaneous data for grade “A” garlic include garlic with bulb diameters of up to 65 mm, then it stands to reason that the Final Results must be distorted. For example, by valuing Dadi’s garlic with a bulb diameter of 50 mm to 55 mm using a combination of the indexed, non-contemporaneous data for “S.A.”-grade garlic together with the contemporaneous data for “A”-grade garlic (which, it now appears, already reflected values for garlic with a bulb diameter of up to 65 mm), Commerce presumably skewed the surrogate value toward Iarger-bulb (typically higher-value) garlic.

As noted above, in light of the significant problems with the Azadpur APMC data, Xinboda advocates use of the Garlico prices to calculate the surrogate value for whole raw garlic bulbs. See Pl.’s Brief at 16, 19, 21. 28 In the alternative, if Commerce continues to rely on the Azadpur APMC data, and if (as all existing record evidence indicates) the contemporaneous data for grade “A” garlic include garlic with bulb diameters of up to 65 mm, then (as Xinboda notes) it would seem that Xinboda’s raw garlic bulb input should be valued based entirely on those contemporaneous data — with, of course, any adjustments that might be necessary to exclude expenses that should not be reflected in an “intermediate input” at the “farm gate.” See id. at 20-22; see also section III.A.1, supra (discussing apparent need for adjustments to Azadpur APMC data to ensure that calculated surrogate value re-fleets “intermediate input” at “farm gate” stage).

Accordingly, in the course of the remand to reconsider whether the surrogate value for raw garlic bulbs calculated in the Final Results included expenses that should have been excluded from the valuation of an intermediate input at the farm gate (discussed in section III.A.1 above), Commerce also shall consider the record evidence indicating that the contemporaneous Azadpur APMC data for grade “A” garlic include garlic with a bulb diameter of up to 65 mm and shall make any necessary revisions to its surrogate value calculations, including the use of another source of data, if appropriate.

B. Surrogate Wage Rate

As section I above explained, in non-market economy cases such as this, all data used to calculate surrogate values for factors of production must satisfy the two prongs of 19 U.S.C. § 1677b(c)(4); see gen erally section I, swpra. Specifically, the data must, to the extent possible, come from one or more market economy countries that are “at a level of economic development comparable to that of the non-market economy country” at issue — here, the PRC. See 19 U.S.C. § 1677b(c)(4)(A) (the “economic comparability” requirement). In addition, the data must, to the extent possible, come from one or more market economy countries that are “significant producers of comparable merchandise.” See 19 U.S.C. § 1677b(c)(4)(B) (the “significant producer” requirement).

For factors of production other than labor, Commerce typically uses values from a single market economy country — known as the primary surrogate country (here, India) — that Commerce has determined to be both (a) economically comparable to the non-market economy country in question and (b) a significant producer of merchandise that is comparable to the merchandise at issue. See 19 C.F.R. § 351.408(c)(2). However, in the past, Commerce treated the valuation of labor quite differently than other factors of production. See generally Dorbest IV, 604 F.3d at 1367-68.

Historically, Commerce valued the cost of labor in non-market economy country cases using “regression-based wage rates reflective of the observed relationship between wages and national income in market economy countries,” a methodology codified in the agency’s regulations at 19 C.F.R. § 351.408(c)(3). See 19 C.F.R. § 351.408(c)(3) (invalidated in Dorbest IV); see also Dorbest IV, 604 F.3d at 1368. Thus, between 1997 and 2010, in contrast to its treatment of other factors of production in non-market economy cases, Commerce calculated surrogate values for labor based on data from a broad “basket” of market economy countries, and did not limit itself to countries at a level of eeonomic development comparable to the non-market economy country in question.

The Court of Appeals’ 2010 decision in Dorbest TV fundamentally altered the landscape. See generally Dorbest IV, 604 F.3d at 1370-73. In that case, the Court of Appeals invalidated Commerce’s regulation providing for the valuation of labor using a multi-country regression analysis (i.e., 19 C.F.R. § 351.408(c)(3)), holding that the regulation did not comply with the provisions of 19 U.S.C. § 1677(c)(4) requiring the agency to use data from economically comparable market economy countries that are significant producers of comparable merchandise. Id., 604 F.3d at 1366, 1370-73, 1377.

In response to the Court of Appeals’ ruling in Dorbest TV, Commerce abandoned the surrogate labor calculation methodology codified in its regulations and implemented an interim methodology. Under its interim methodology, Commerce calculated a surrogate value for labor by averaging wage data from countries that the agency found to be both “economically comparable” to the non-market economy country in question and “significant producers” of merchandise comparable to the merchandise at issue in a case. See Anti-dumping Methodologies in Proceedings Involving Non-Market Economies: Valuing the Factor of Production: Labor; Request for Comment, 76 Fed.Reg. 9544, 9544, 9546 (Dep’t Commerce Feb. 18, 2011) (“Interim Labor Methodology”). The agency later modified its interim methodology to limit averaging to industry-specific data. See id. at 9544. Under both the original and modified interim methodology, Commerce deemed all countries that exported comparable merchandise within a three-year period to be “significant producers.” See id. at 9546. 29

Commerce characterized its Interim Labor Methodology as the agency’s “attempt! ] to balance its desire for multiple data points with the statutory requirements that [factors of production] data be from countries that are both economically comparable and significant producers” of subject merchandise. See Antidumping Methodologies in Proceedings Involving Non-Market Economies: Valuing the Factor of Production: Labor, 76 Fed.Reg. 36,-092, 36,093 (Dep’t Commerce June 21, 2011) (“Revised Labor Methodology”). Although “the amount of available data was more constrained as a result of the [Court of Appeals’] Dorbest decision,” Commerce reasoned that the agency’s “industry-specific interim methodology still provided the best available wage rate because it allowed for multiple data points.” Id.

After the administrative review at issue here was initiated but before the Final Results issued, Shandong Rongxin ruled that Commerce’s inference that all countries exporting subject merchandise were “significant producers” was “wholly unsupported” and constituted an impermissible construction of 19 U.S.C. § 1677b(e)(4). See Shandong Rongxin Import and Export Co. v. United States, 35 CIT -, -, 774 F.Supp.2d 1307, 1315-16 (2011). Based on its review of the matter, Commerce determined that any alternative definition for “significant producer” which would comply with Shandong Rongxin would greatly restrict the number of countries from which the agency could source labor data. See Revised Labor Methodology, 76 Fed.Reg. at 36,093. According to Commerce, “the base for an average wage calculation would be so limited that there would be little, if any, benefit to relying on an average of wages from multiple countries for purposes of minimizing the variability that occurs in wages across countries.” Id.

Ultimately, Commerce concluded — given both the Court of Appeals’ decision in Dorbest IV and the decision of this court in Shandong Rongxin — that “relying on multiple countries to calculate the wage rate [was] no longer the best approach” for determining the surrogate value of labor. See Revised Labor Methodology, 76 Fed.Reg. at 36,093. Under Commerce’s Revised Labor Methodology, announced in June 2011, Commerce now relies exclusively on industry-specific labor cost data from the primary surrogate country, which the Revised Labor Methodology describes as “the best approach for valuing the labor input in [non-market economy] antidumping duty proceedings.” See id. In releasing the Revised Labor Methodology, Commerce advised that “[in] ongoing [non-market economy] proceedings,” the agency would “consider on a case-by-case basis whether it is feasible to implement the new labor methodology within statutory deadlines.” See id. at 36,093; see also id. at 36,094 (stating that Revised Labor Methodology “will be applied to ongoing administrative [non-market economy] proceedings where the statutory deadlines permit”).

The Issues & Decision Memorandum in the instant review — dated June 20, 2011 (i.e., 10 days after the Revised Labor Methodology’s issuance on June 10, 2011, and one day before the revised methodology’s publication in the Federal Register)— reflects the use of Commerce’s modified interim methodology in calculating the surrogate value for labor. See Issues & Decision Memorandum (dated June 20, 2011); Final Results, 76 Fed.Reg. 37,321 (June 27, 2011); Revised Labor Methodology, 76 Fed.Reg. at 36,092 (June 21, 2011); see generally Issues & Decision Memorandum at 23-28 (discussing agency’s calculation of surrogate value for labor).

To identify which countries were at a level of development comparable to the PRC for purposes of calculating the surrogate value for labor, Commerce began with the list of countries in the agency’s so-called Surrogate Country Letter, in which Commerce identified a number of countries as potential surrogates from which the agency would be able to derive values for all factors of production other than labor. See Issues & Decision Memorandum at 23-24. From those countries, Commerce selected a pair of “bookend” countries — specifically, the country with the highest gross national income (i.e., Peru) and the country with the lowest gross national income (ie., India, the primary surrogate country). See id. at 24. Commerce then identified all market economy countries in the World Bank’s World Development Report with per capita gross national incomes for 2008 that fell between those of the two “bookend” countries. See id. That produced a list of 43 countries. See id.

For purposes of its surrogate labor calculations, Commerce considered those 43 countries to be “economically comparable” to the PRC. Issues & Decision Memorandum at 24. Next, Commerce determined which of the 43 countries were also “significant producers” of comparable merchandise, by identifying every country that exported any quantity of comparable merchandise (defined as exports under the seven ten-digit tariff provisions identified in the scope of the underlying antidumping duty order). See id. 30 That narrowed the list of 43 countries to 27 countries that Commerce viewed as satisfying both the “economic comparability” prong and the “significant producer” prong of the statute. See id.; 19 U.S.C. § 1677b(c)(4).

The next step in Commerce’s labor analysis was to assess which of the 27 countries had made available adequate data that could be averaged together to produce the surrogate value for labor in this review. See Issues & Decision Memorandum at 25. Commerce decided to rely on earnings or wages data reported to the International Labor Organization (“ILO”) from 2003 to 2008. See id. at 25, 27. When countries report industry-specific data to the ILO (including the wages and earnings data used in the review here at issue), those data are reported in accordance with a uniform code known as the International Standard Industrial Classification of All Economic Activities (“ISIC”), maintained by the United Nations Statistical Division. See id. at 27. 31 However, the ISIC code is updated from time to time, and not all countries report industry-specific data under the same revision. In this instance, Commerce determined that none of the countries that it had found to be both “economically comparable” and “significant producers” had reported data under the most recent revision of ISIC, Revision 4. See id. 32 Commerce therefore reviewed ISIC Revision 3 to identify the appropriate classification for garlic production. Under Revision 3, Commerce concluded that the most specific provision was “Sub-Classification 15, ... described as ‘Manufacture of Food Products and Beverages.’ ” See id. at 27-28. Commerce was left with a total of eight countries — specifically, Ecuador, Egypt, Indonesia, Jordan, Peru, the Philippines, Thailand, and Ukraine — that the agency viewed as satisfying both the “economic comparability” and “significant producer” prongs of the statute and that the agency also considered to have reported adequate and reliable industry-specific wage and earnings data for use in calculating a surrogate value for labor in the instant review. See id. at 25, 28. Commerce’s final step was to filter and adjust the data from the eight remaining countries (for example, indexing data using a consumer price index, where necessary, to be contemporaneous with the period of review here). See id. at 27.

Commerce thus calculated the surrogate value for labor in the administrative review at issue “using a simple average of the data provided to the ILO under Sub-Classification 15 of the ISIC-Revision 3 standard by countries determined to be economically comparable to the [People’s Republic of China] and significant producers of comparable merchandise.” Issues & Decision Memorandum at 28; see also id. at 27. Conspicuously absent from Commerce’s final group of eight countries was India. India had reported industry-specific data under ISIC Revision 2, but not ISIC Revision 3, and — voicing concern that “the industry definitions may lack consistency between different ISIC revisions” — Commerce declined to combine data from ISIC Revision 2 (including data from India) with the data from ISIC Revision 3 on which it based its calculations in this case. See id. 33

Xinboda attacks Commerce’s calculation of the surrogate value for labor on multiple grounds. See generally Pl.’s Brief at 37-JO; PL’s Reply Brief at 10-14. Speeifical ly, Xinboda argues that it was “arbitrary and capricious” for Commerce to base its calculations on labor data from multiple countries. See Pl.’s Brief at 37. Xinboda further claims that the only reasonable data for use in valuing labor were data from the primary surrogate country, India — an argument based in part on Commerce’s adoption of its Revised Labor Methodology. See id. at 37-39. Citing Shandong Rongxin, Xinboda also contends that Commerce’s labor calculations were tainted by data from countries that were not “significant producers” of comparable merchandise. See id. at 39-40; Pl.’s Reply Brief at 10-12. In addition, Xinboda contests Commerce’s decision to rely solely on labor data reported under ISIC Revision 3, which had the effect of excluding data from India. See PL’s Brief at 38-39; PL’s Reply Brief at 13.

1. Whether the Final Results Should Have Relied on Indian Data Alone

Xinboda first contends that, in determining the surrogate value for labor, Commerce erred in using data from a “basket of countries.” See generally PL’s Brief at 37-38; PL’s Reply Brief at 10; Issues & Decision Memorandum at 25. According to Xinboda, Commerce should have relied on labor data from India alone. See PL’s Brief at 37-38.

Xinboda claims that, because Commerce valued all other factors of production using only Indian data, it was “arbitrary and capricious” for the agency to value labor using data from multiple countries. PL’s Brief at 37. As Shandong Rongxin explained, however, there is nothing inherently improper or unlawful in Commerce’s use of data from more than one country in calculating a surrogate value for labor. See Shandong Rongxin, 35 CIT at-, 774 F.Supp.2d at 1314. 34 Both the statute and relevant case law explicitly authorize the agency’s use of data from multiple countries in circumstances such as these. Bee 19 U.S.C. § 1677b(e)(4) (instructing Commerce to “utilize, to the extent possible, the prices or costs of factors of production in one or more market economy countries”) (emphasis added); 19 U.S.C. § 1677b(c)(l)(B) (requiring Commerce to use “the best available information regarding the values of ... factors [of production] in a market economy country or countries considered to be appropriate”) (emphasis added); Dorbest IV, 604 F.3d at 1372 (contemplating, with approval, agency use of data from a “subset of ... countries” to value labor). Shandong Rongxin further concluded that averaging data from multiple countries to produce a single surrogate value for labor was “well justified” as a matter of policy, in light of “the existence of variation in wages among sun ilarly economically developed countries, and the reasons for [that variation].” Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1314. Xinboda’s claim that Commerce acted arbitrarily and capriciously in using labor data from multiple countries is therefore unavailing. See generally Def.’s Response Brief at 26-27.

Xinboda fares better on its claim that— as the primary surrogate country and (next to the PRC) the world’s second-ranking producer of garlic (with more than twice the production of the next highest country), and with an economy at a level of development comparable to that of the PRC — India is the only reasonable source for surrogate data on labor in this review. See Pl.’s Brief at 38-39; id. at 39 (arguing that “India’s labor rate represents the best available information in this review” due to “the relative comparability of its garlic production”). In addition, according to Xinboda, “the vast size of the Chinese labor population” (which Xinboda characterizes as “arguably the most important factor[] in labor market comparability”) renders India “the only country remotely comparable” to the PRC for purposes of calculating the surrogate value for labor in this review. Id. at 38; see also id. at 39 (arguing that “India’s labor rate represents the best available information in this review” due to “its unique population”).

As further proof in support of its claim that Commerce should have relied solely on labor data from the primary surrogate country, India, Xinboda invokes Commerce’s Revised Labor Methodology, asserting that even the agency has now “concluded [that] it should value labor using data from the primary surrogate country only.” PL’s Brief at 37 (citing Revised Labor Methodology). According to Xinboda, in adopting the revised methodology, Commerce “effectively determined that ... labor [data] from a single surrogate country is the ‘best available information’ ” for use in calculating a surrogate value for labor. PL’s Brief at 37-38. Xinboda argues that, given the new methodology, Commerce “cannot ... maintain that multiple countries are ‘best’ in this proceeding,” because — Xinboda reasons — “[b]y definition, there can only be one ‘best’ method of valuing labor.” Id. at 38.

The Government emphasizes that, in adopting the Revised Labor Methodology, Commerce stated that its application to ongoing proceedings would be evaluated “on a case-by-case basis.” See Def.’s Response Brief at 27; Revised Labor Methodology, 76 Fed.Reg. at 36,093. The Government adds that, in any event, “the application of a later developed methodology in subsequent proceedings does not render a prior determination unlawful, or unsupported by the record evidence.” Def.’s Response Brief at 27 (citing Home Products Int’l, Inc. v. United States, 36 CIT -, -, 810 F.Supp.2d 1373, 1378 (2012) (“Home Products I ”), opinion following remand, 36 CIT -, 837 F.Supp.2d 1294 (2012) (“Home Products II”), affd, 501 Fed.Appx. 981 (Fed.Cir.2013)).

The Issues & Decision Memorandum reflects no consideration of the feasibility of applying the new methodology here, although — notably—the Revised Labor Methodology is dated June 10, 2011 (10 days before the date of the Issues & Decision Memorandum), and the Revised Labor Methodology appeared in the Federal Register a mere one day after Commerce’s final determination in this review. See Revised Labor Methodology (dated June 10, 2011), 76 Fed.Reg. at 36,092 (June 21, 2011); Issues & Decision Memorandum (dated June 20, 2011); Final Re- suits, 76 Fed.Reg. 37,321 (June 27, 2011). 35 Moreover, quite apart from the timing of Commerce’s issuance of its Revised Labor Methodology and the agency’s final determination in this review, the line of authority that the Government cites concerning the retroactive application of new agency methodologies does not neatly dispose of Xinboda’s claim. 36 The heart of Xinboda’s argument is not simply that Commerce’s use of data from multiple countries to value labor is inconsistent with Commerce’s new methodology. Xinboda’s argument is much more basic: That the only reasonable source of data for use in calculating the surrogate value for labor is India, the primary surrogate country. In other words, Commerce’s change in methodology may provide support for Xinboda’s argument; but the change in methodology is not itself the basis for that argument. 37

This issue therefore must be remanded to Commerce, to allow the agency to address Xinboda’s claims that the use of data from countries other than India was unreasonable and to allow the agency to evaluate and explain the feasibility of applying the Revised Labor Methodology here, in a manner consistent with the agency’s actions in other similar proceedings. See generally, e.g., Final Remand Results of Redetermination Pursuant to Second Remand at 2, 25-26 (March 29, 2012), filed in Jinan Yipin Corp. v. United States, Court No. 06-00189 (voluntarily applying Revised Labor Methodology in second remand determination in tenth administrative review of same antidumping order at issue here); Taian Ziyang Food Co. v. United States, 37 CIT -, -, 918 F.Supp.2d 1345, 1357-58 (2013) (“Taian Ziyang III ”) (sustaining remand results in ninth administrative review of same antidumping order at issue here, where agency applied Revised Labor Methodology); Jinan Yipin Corp. v. United States, 36 CIT -, -, 2012 WL 2001755 * 1-2 (2012) (sustaining results of remand in eighth administrative review of same antidumping order at issue here, where agency applied Revised Labor Methodology); Xiamen Int’l Trade and Indus. Co. v. United States, 37 CIT -, -, 953 F.Supp.2d 1307, 1321 (2013) (granting request for voluntary remand to permit agency to apply Revised Labor Methodology in administrative review initiated in late March 2010); Camau Frozen Seafood Processing Import Export Corp. v. United States, 86 CIT -, -, 880 F.Supp.2d 1348, 1352 (2012) (“Camau I”) (noting that, where administrative review (initiated in April 2010) was pending when revised methodology was adopted, Commerce applied Revised Labor Methodology in final determination issued approximately three months later).

2. Whether Commerce’s “Significant Producer” Analysis Was Contrary to Law

Xinboda also claims that the Final Results’ valuation of labor ran afoul of 19 U.S.C. § 1677b(c)(4)(B), the provision of the antidumping statute that generally limits the data on which Commerce may rely in valuing factors of production to data from countries that are “significant producers” of merchandise that is “comparable” to the merchandise at issue in a proceeding. See generally PL’s Brief at 38-39; PL’s Reply Brief at 13; 19 U.S.C. § 1677b(c)(4). 38 Specifically, Xinboda argues that Commerce erred in “defining a ‘significant producer’ as a country that ... exported comparable merchandise” during the period of review. See PL’s Brief at 39; PL’s Reply Brief at 10-12; Issues & Decision Memorandum at 24-25. Xinboda disputes Commerce’s “inference that the volume of a country’s exports implies significant domestic production.” PL’s Reply Brief at 11; see also PL’s Brief at 39; PL’s Reply Brief at 11-12; Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1316. Xinboda thus contends that Commerce’s interpretation of the term “significant producers” to include any and all countries that exported any volume of garlic during the relevant time period was an unreasonable reading of the statute. See PL’s Brief at 39; PL’s Reply Brief at 10-12.

For example, according to Xinboda, “eight of the countries used [in Commerce’s labor calculations] exported no garlic in 2009 and a ninth exported $52 [worth of garlic].” See PL’s Reply Brief at 11; see also id. at 11-12 (arguing, inter alia, that “[t]he use of [data from] countries [that] exported ... $52 worth of garlic is not better, much less the best, evidence when weighed against the labor rate of the country that ranks as by far and away the second largest garlic producer in the world next to China — India”); PL’s Brief at 39 (asserting that “[t]he countries used to value labor in this case included countries with minuscule exports that in no rational manner can be construed as significant or [as] evidence of ... meaningful domestic production”). Xinboda concludes that, to the extent that Commerce continues to rely on data from multiple countries in calculating the surrogate value for labor, Commerce “must be limited to the countries that can be reasonably defined as significant producers.” PL’s Brief at 40.

Shandong Rongxin is precisely on point. See Shandong Rongxin, 35 CIT at -, -, 774 F.Supp.2d at 1314, 1315-16. Here, as there, neither Commerce nor the Government “explained] how treating a country with any quantity of exports” is a permissible construction of the statute’s reference to “significant producers.” Id., 35 CIT at -, 774 F.Supp.2d at 1314. And here, as there, the effect of Commerce’s definition is to treat as “significant producers” countries “which almost certainly have no domestic production — at least not any meaningful production, capable of having influence or effect.” Id., 35 CIT at -, 774 F.Supp.2d at 1316. As such, here — as in Shandong Rongxin— Commerce’s interpretation of “significant producers” cannot be sustained. Id., 35 CIT at -, 774 F.Supp.2d at 1316. The Final Results’ surrogate value for labor therefore must be remanded to Commerce for further consideration on this point as well. 39

The directions of the court in Shandong Rongxin are equally applicable in this case:

On remand, if Commerce wishes to continue to rely on export statistics as a proxy for significant production, it must include some additional mechanism to ensure that it does not propagate the fiction that countries with a few dollars of exports are engaged in significant production. Alternatively, Commerce is free to adopt an altogether different approach to identifying significant production.

Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1316. Commerce, of course, took a truly “altogether different approach” in adopting its Revised Labor Methodology in response to both the Court of Appeals’ decision in Dorbest and the decision in Shandong Rongxin. See Revised Labor Methodology, 76 Fed.Reg. at 36,093 (in response to court decisions, providing for use of “data on industry-specific wages from the primary surrogate country” to value labor). To the extent that Commerce decides on remand to apply its Revised Labor Methodology to value labor in this review, Xinboda’s concerns will be resolved. Otherwise, Commerce shall modify the way in which it determines whether a country is a “significant producer” of comparable merchandise in accordance with the remand instructions in Shandong Rongxin (outlined above), and shall recalculate the surrogate value of labor accordingly.

3. Whether Commerce Erred in Excluding Data Not From ISIC Revision S

In its remaining challenge to the surrogate labor rate set forth in the Final Results, Xinboda claims that Commerce erred in using only industry-specific labor data reported under sub-classification 15 of Revision 3 of the International Labor Organization’s International Standard Industrial Classification of All Economic Activities (“ISIC Revision 3”), which had the effect of excluding data from India — the primary surrogate country and the source of the data used in valuing all other factors of production in the administrative review at issue. See generally Pl.’s Brief at 38, 39; PL’s Reply Brief at 10, 13; Issues & Decision Memorandum at 23, 27-28. Although India reported industry-specific data under the parallel provision of “ISIC Revision 2,” India did not report such data under the subsequent revision of ISIC on which Commerce relied in the Final Results. See Issues & Decision Memorandum at 27-28; PL’s Brief at 38. Xinboda contends that, to the extent that Commerce relies on data from multiple countries to calculate the surrogate value for labor here, Commerce should be required to include “ISIC-Rev 2” data (and, in par ticular, Indian data) in its calculations. See id. at 39; PL’s Reply Brief at 13; see also Issues & Decision Memorandum at 23.

This same issue — ie., Commerce’s decision to rely exclusively on labor data from ISIC Revision 3, and thus to exclude data from the primary surrogate country (India) which was reported under ISIC Revision 2 — was presented in Shandong Rongxin, and again in Home Products and Since Hardware. See Shandong Rongxin, 35 CIT at -, -, 774 F.Supp.2d at 1311, 1312-15; Home Products I, 36 CIT at -, -, 810 F.Supp.2d at 1378, 1379-80; Since Hardware (Guangzhou) Co. v. United States, 37 CIT -, -, 911 F.Supp.2d 1362, 1381-82 (2013) (discussing August 14, 2012 order remanding, inter alia, issue of surrogate value for labor to permit agency to conform determination to that in Home Products). In the final results in each of those cases, Commerce took the position that — in calculating a surrogate value for labor — the use of data from the broadest possible basket of countries was the agency’s paramount concern. See Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1315; Home Products I, 36 CIT at -, 810 F.Supp.2d at 1379-80; Issues and Decision Memorandum for the Final Results in the Administrative Review of Floor-Standing, Metal-Top Ironing Tables and Certain Parts Thereof from the People’s Republic of China, A-570-888 (March 22, 2011) (comment 2) (addressing 2008-2009 administrative review, at issue in Since Hardware). Nevertheless, in the final results in each of the three cases, Commerce rejected data from the primary surrogate country, India — even though the Indian data satisfied all other agency criteria — due to Commerce’s preference for data from a single ISIC revision. Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1315; Home Products I, 36 CIT at -, 810 F.Supp.2d at 1379; Since Hardware, 37 CIT at -, 911 F.Supp.2d at 1381-82; see also Issues and Decision Memorandum for the Final Results in the Administrative Review of Floor-Standing, Metal-Top Ironing Tables and Certain Parts Thereof from the People’s Republic of China, A-570-888 (comment 2).

In all three cases, the court questioned the reasonableness of Commerce’s emphasis on the use of labor data from a single ISIC revision when the result is, among other things, to exclude data from the primary surrogate country (in each case, India). See Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1314-15; Home Products I, 36 CIT at -, 810 F.Supp.2d at 1379-80; Since Hardware, 37 CIT at -, 911 F.Supp.2d at 1381-82 (citing Home Products). In Shandong Rongxin and Home Products, the issue was remanded to Commerce to explain “why the need for consistency across ISIC revisions predominate[d] over the need for a broad basket of countries to value labor.” Shandong Rongxin, 35 CIT at -, 774 F.Supp.2d at 1315; Home Products I, 36 CIT at -, 810 F.Supp.2d at 1380. Similarly, in Since Hardware, the issue was remanded to permit Commerce to conform its determination to the remand determination in Home Products, “to include Indian data under ISIC Revision 2, as well as any other appropriate country.” See Since Hardware, 37 CIT at -, 911 F.Supp.2d at 1381-82. On remand in Home Products and Since Hardware, Commerce recalculated the respective surrogate values for labor using not only data reported under ISIC Revision 3, but also data reported under Revision 2, including data from, inter alia, India. See Home Products II, 36 CIT at -, 837 F.Supp.2d at 1296; Since Hardware, 37 CIT at -, 911 F.Supp.2d at 1382. And, on remand in Shandong Rongxin, Commerce recalculated the surrogate labor value using exclusively data from India, applying the agency’s Revised Labor Methodology. See Shandong Rongxin Import & Export v. United States, 35 CIT -, -, 2011 WL 3841578 *1 (2011) (sustaining Final Results of Redetermination Pursuant to Remand (Aug. 4, 2011), which, inter alia, applied Revised Labor Methodology, valuing labor using solely data from India).

As in Shandong Rongxin, Home Products, and Since Hardware, the Final Results in this case asserted Commerce’s “long-standing preference” for labor data from “a robust basket” of countries. See Issues & Decision Memorandum at 26; see also id. at 27 (stating that Commerce “prefers to use data from multiple sources” and favors “more data points”); Def.’s Response Brief at 27 (arguing that “using more data is preferable to using less data because it allows for a more accurate calculation of the labor wage rate”). Yet — as in Shandong Rongxin, Home Products, and Since Hardware — Commerce here declined to use labor data from the surrogate country, India, “[d]ue to concerns that the industry definitions may lack consistency between different ISIC revisions.” Issues & Decision Memorandum at 27; see also id. (asserting that “averaging wage rates within the same ISIC revision (ie., not mixing revisions) constitutes the best available information”); Def.’s Response Brief at 28 (same). A remand is therefore warranted, to permit Commerce to reconsider its exclusion of labor data from India, in light of Shandong Rongxin, Home Products, and Since Hardware.

To the extent that Commerce decides on remand to apply its Revised Labor Methodology in this review, and thus bases the surrogate value for labor exclusively on Indian data, Xinboda’s concerns will be resolved. See Revised Labor Methodology, 76 Fed.Reg. at 36,093-94 (providing for surrogate valuation of labor based on “data on industry-specific wages from the primary surrogate country,” and noting that agency “make[s] every attempt to identify and review relevant industry-specific wages in the primary surrogate country that are as contemporaneous as possible with the period of investigation”). In the alternative, to the extent that Commerce continues to rely on labor data from multiple countries, Commerce shall revise its calculations to include ISIC Revision 2 data from India (and any other countries, as necessary), or Commerce shall explain in detail why such action would not be appropriate. 40

C. Surrogate Financial Ratios

In constructing normal value for a foreign producer in a non-market economy country, Commerce bases its determination on “the value of the factors of production utilized in producing the merchandise.” 19 U.S.C. § 1677b(c)(l). However, as section I above explains, valuing product-specific factors of production does not capture certain overall “general expenses and profit.” Id. Hence, Commerce must separately reflect in the agency’s calculation of normal value (1) manufacturing overhead, (2) selling, general and administrative expenses (“SG & A”), and (3) profit. Id. As with other factors of production, Commerce uses surrogate values to determine a respondent’s financial ratios, relying on the financial statements of one or more producers of identical or comparable merchandise which serve as surrogates for that purpose. See generally id.; 19 C.F.R. § 351.408(c)(4); Ad Hoc Shrimp Trade Action Comm. v. United States, 618 F.3d 1316, 1319-20 (Fed.Cir.2010) (providing overview of use of financial statements in determining surrogate financial ratios).

Commerce’s regulations require that surrogate financial ratios be derived from “non-proprietary information gathered from producers of identical or comparable merchandise in the surrogate country.” 19 C.F.R. § 351.408(c)(4) (emphasis added); see also Issues & Decision Memorandum at 22. The Final Results at issue here add that Commerce “favors the financial statements of surrogates that [1] produce the identical merchandise, [2] consume the identical raw material, and [3] have identical or comparable production experience.” Issues & Decision Memorandum at 19 n. 60. And, in determining whether a potential surrogate produces merchandise that is comparable, Commerce typically considers (1) physical characteristics, (2) end uses, and (3) production processes. See Issues and Decision Memorandum for the Antidumping Duty Investigation of Certain Oil Country Tubular Goods from the People’s Republic of China (April 8, 2010) (comment 13) (“OCTG Issues & Decision Memorandum”) (citing Issues and Decision Memorandum for the Administrative Review of Certain Cased Pencils from the People’s Republic of China; Final Results (July 16, 2002) (comment 5)) (noting that, in selecting financial statements, agency “may consider how closely the surrogate producers approximate the [non-market economy] producers’ experience”).

In addition, financial ratios that are derived from the financial statement of a potential surrogate that has benefited from countervailable subsidies may be less representative of the relevant industry than ratios derived from the financial statements of potential surrogates that did not benefit from such subsidies. Accordingly, where Commerce has “reason to believe or suspect” that a potential surrogate may have benefited from subsidies that the agency has previously found to be countervailable, it is Commerce’s practice to disregard the financial statements of that company if there is other useable data on the record. See Issues & Decision Memorandum at 20 (citing Certain New Pneumatic Off-the-Road Tires from the People’s Republic of China: Final Affirmative Determination of Sales at Less Than Fair Value and Partial Affirmative Determination of Critical Circumstances, 73 Fed.Reg. 40,-485, 40,486 (Dep’t Commerce July 15, 2008) and accompanying Issues and Decision Memorandum (July 7, 2008) (comment 17.A)). In determining whether there is “reason to believe or suspect” that a potential surrogate has received actionable subsidies, the Final Results emphasized that Commerce “reifies] on information in financial statements on an ‘as is’ basis,” and does not look beyond a company’s financial statements to consider other information on the administrative record. See Issues & Decision Memorandum at 20.

As Commerce explained in the Final Results, the agency’s “preference” is “to use financial data from more than one surrogate producer.” Issues & Decision Memorandum at 20; see also Dorbest IV, 604 F.3d at 1374 (noting Commerce’s preference “ ‘to use multiple financial statements in order to eliminate potential distortions that may arise from using those of a single producer,’ as long as those financial statements ‘are not distortive or otherwise unreliable’ ”). In the Final Results at issue here, however, Commerce calculated Xinboda’s surrogate financial ratios based solely on the 2009-2010 unconsolidated financial statement of Tata Global Beverages Limited (“Tata Global”), an Indian company that grows, processes, and sells its own coffee and tea products. See Issues & Decision Memorandum at 20, 22. Of the five other financial statements on the administrative record, four were rejected because Commerce concluded that the companies had benefited from subsidies under programs that the agency had previously found to be countervailable. Id. at 20-22. 41 Commerce also disregarded the financial statement of Garlico — the only one of the six potential surrogates that actually purchased and processed gar-lie — based on the agency’s findings that Garlico acted as a “trading company” on roughly a quarter of its sales, that the company’s primary production was of “downstream food products,” and that “all of the [company’s] raw garlic and raw onion sales were traded goods.” Id. at 22.

Xinboda takes strong exception to Commerce’s decision to rely on Tata Global’s financial statement, and protests the agency’s rejection of the financial statement of Garlico. See generally Pl.’s Brief at 2, 22-36, 40; PL’s Reply Brief at 6-10. The Government maintains that Commerce properly relied on Tata Global’s financial statement, citing two reasons. See Def.’s Response Brief at 4-5 (citing Issues & Decision Memorandum at 20-22); see generally Def.’s Response Brief at 6, 21-26. First, the Government asserts that Commerce has long relied on the financial statements of tea companies in calculating surrogate financial ratios in reviews involving the antidumping order on fresh garlic from the PRC. See id. at 21-22. In addition, the Government argues that Tata Global was the only company with production processes similar to those of Xinboda that had not been the beneficiary of a subsidy that the agency had found to be countervailable. See id. at 21.

1. Whether Commerce’s Use of Financial Statements of Tea Producers — and Tata Global, in Particular — Is Uniform Agency Practice

The Government claims that Commerce has found tea production to be comparable to garlic production “since the 2001-2002 period of review” (ie., since the eighth administrative review of the antidumping order on fresh garlic from the PRC). See Def.’s Response Brief at 21-22 (citing Fresh Garlic from the People’s Republic of China: Final Results of Antidumping Duty Administrative Review and New Shipper Reviews, 69 Fed.Reg. 33,626 (Dep’t Commerce June 16, 2004), and accompanying Issues and Decision Memorandum (June 16, 2004) (comment 7)); see also Issues & Decision Memorandum at 20 (emphasizing that agency “used financial statements from non-integrated tea processors” in calculating surrogate financial ratios “[i]n the most recent segment of this proceeding”).

However, the Government overstates the facts concerning Commerce’s use of the financial statements of tea producers such as Tata Global as surrogates for producers of garlic. See Pl.’s Reply Brief at 7. Contrary to the Government’s implication, Commerce has no uniform practice. Xinboda points out that Commerce rejected Tata Global’s financial statements in the ninth administrative review, 42 after evidence came to light indicating that the company was focused not on the bulk tea market, but rather on individually-packaged teas and other such products (suggesting that the company’s processing had become significantly more complex and refined than it had previously been). Id. And, in many of the administrative reviews and new shipper reviews since that time, Commerce has chosen to use financial statements other than those of Tata Global to calculate surrogate financial ratios. 43 In fact, the ongoing and most recent reviews not only do not rely on the financial statements of Tata Global, they do not even use the financial statements of other tea producers; instead, they rely on the financial statements of AgriPure Holdings Public Company Limited (“AgriPure”), a producer and exporter of sweet corn and other fresh vegetables. See n. 43, supra (surveying financial statements used in various reviews of antidumping order on fresh garlic from the PRC).

Further, as Xinboda points out, there is no standard Commerce practice of treating tea processing as comparable to the production of whole garlic bulbs. See Pl.’s Reply Brief at 7; see also Issues & Decision Memorandum at 20 (stating that tea processing is “sufficiently similar to garlic processing in that neither product is highly processed or preserved prior to sale”); Preliminary Surrogate Value Memorandum at 7 (stating that “tea is comparable to subject merchandise (i.e., whole and peeled garlic)”). Thus, for example, Commerce in the past has taken the position that tea production is comparable to the production of peeled garlic, while — as surrogates for the producers of whole garlic bulbs — Commerce generally has favored rice producers. See Pl.’s Brief at 23; Pl.’s Reply Brief at 7; see also, e.g., Fresh Garlic from the People’s Republic of China: Issues and Decision Memorandum for the Final Results of the New Shipper Reviews and Rescission, In Part, of the New Shipper Reviews (Sept. 24, 2009) (comment 7) (concluding that production processes of Tata and another producer “are more comparable to that of peeled garlic,” and that “rice production ... is comparable to that of less-processed whole garlic bulbs”).

Accordingly, while it is true that financial statements of tea processors often were relied on in past reviews of the anti-dumping order on fresh garlic from the PRC, there has been no uniform agency practice. Further, although Commerce often relied on Tata Global’s financial statements, Commerce also often used the financial statements of other companies. And Commerce’s current trend is to eschew the financial statements of tea processors entirely.

2. Whether the Final Results Properly Relied on Financial Statement of Tata Global and Properly Rejected Financial Statement of Garlico

Much as it disputes the Government’s intimation that Commerce routinely relies on the financial statements of tea producers, Xinboda similarly challenges the key determination underpinning Commerce’s reliance on the financial statement of Tata Global here — specifically, Commerce’s determination that Tata Global was the only company with production processes similar to those of Xinboda that had not been the beneficiary of a subsidy that the agency had found to be countervailable. See Issues & Decision Memorandum at 20-22; see also Def.’s Response Brief at 21. Xinboda maintains that, in fact, Tata Global did benefit from such subsidies. See Pl.’s Brief at 27-30; PL’s Reply Brief at 8-9. Further, Xinboda contests Commerce’s determinations concerning the comparability of the respective operations of Xinboda, Tata Global, and Garlico. See PL’s Brief at 27, 30-36; PL’s Reply Brief at 7-8, 9-10. 44

a. Allegations of Subsidies Received by Tata Global

As proof of subsidies allegedly received by Tata Global, Xinboda placed on the record of the administrative review copies of loan agreements that were publicly available, on file with the Indian Ministry of Corporate Affairs. See PL’s Brief at 27; see also Issues & Decision Memorandum at 20. Xinboda asserts that these documents constitute evidence that Tata Global “secured loans based on its possession of goods ready and packed for export” — i e., so-called “packing credits,” or “loans collateralized (ie., hypothecated) by packed finished goods.” PL’s Brief at 27-28. According to Xinboda, “[t]hese very ‘packing credits’ were cited [as the basis for Commerce’s] decision to reject the ... financial statements of rice processor REI Agro.” Id. at 28; see also Issues & Decision Memorandum at 21 n. 70 (noting that packing credit program under which REI Agro received benefits “has been found countervailable,” citing Polyethylene Terephthalate Film, Sheet, and Strip from India: Final Results of Countervailing Duty Administrative Review, 73 Fed.Reg. 7708 (Dep’t Commerce Feb. 11, 2008) and accompanying Issues and Decision Memorandum (Feb. 4, 2008)). Commerce infers that such loans are procured at below-market interest rates. See PL’s Brief at 28.

In its Final Results, Commerce acknowledges the loan agreements that Xinboda placed on the record. See Issues & Decision Memorandum at 20. However, there is no indication that Commerce ever reviewed the loan agreements themselves or otherwise gave them any consideration. The Final Results state simply that “it is [Commerce’s] practice to rely on information in financial statements on an ‘as is’ basis when calculating surrogate financial ratios,” and that — in reviewing Tata Global’s financial statement — Commerce found no evidence of such loans. Id. (citing Certain Frozen Warmwater Shrimp From the People’s Republic of China: Notice of Final Results and Rescission, in Part, of 2004/2006 Antidumping Duty Administrative and New Shipper Reviews, 72 Fed. Reg. 52,049 (Dep’t Commerce Sept. 12, 2007) and accompanying Issues and Decision Memorandum (Sept. 5, 2007) (comment 2)); see also Def.’s Response Brief at 23-24.

The Government argues that “Xinboda has not demonstrated that it was improper for Commerce to give more weight to information in [Tata Global’s] financial statements than information contained in other documents, such as the loan agreements proffered by Xinboda.” Def.’s Response Brief at 24. But those are not the facts here. This is not a situation where Commerce considered conflicting evidence and then exercised its expertise and discretion in deciding which evidence to credit. Contrary to the Government’s assertion, Commerce here made no determination as to the relative weight to accord Tata Global’s financial statement versus the loan agreements. Commerce confined its review to the face of Tata Global’s financial statement only, resting on the agency’s practice of “relypng] on information in financial statements on an ‘as is’ basis.” Issues & Decision Memorandum at 20 (asserting that agency found no evidence of loans in question “in [the agency’s] analysis of Tata Tea’s 09/10 financial statement”).

Significantly, neither Commerce nor the Government has offered any explanation or justification for the agency’s stated practice of relying on financial statements “on an ‘as is’ basis” and ignoring all other record evidence. See PL’s Brief at 28-29, 30; PL’s Reply Brief at 8. 45 Xinboda maintains that, like any other administrative agency, Commerce is required to make its determination on the entire administrative record. See PL’s Brief at 28-29; PL’s Reply Brief at 8 (citing 19 U.S.C. § 1516a(b)(l)(B)(i) (requiring “substantial evidence on the record”); 19 U.S.C. § 1516a(b)(2)(A) (defining “record” for purposes of § 1516a(b)(l)(B)(i) to include “all information presented to or obtained by” Commerce)).

Moreover, Xinboda argues, the legislative history of Commerce’s “reason to believe or suspect” standard militates against the position that the agency has staked out. See PL’s Brief at 28-29, 30; PL’s Reply Brief at 8. Specifically, Congress has instructed Commerce to “avoid using any prices which it has reason to believe or suspect may be dumped or subsidized prices” in calculating normal value using the factors of production methodology. Omnibus Trade and Competitiveness Act of 1988, Conference Report to Accompany H.R. 3, H.R.Rep. No. 100-576 at 590 (1988) (Conf. Rep.), reprinted in 1988 U.S.C.C.A.N. 1547, 1623. In doing so, Congress emphasized that Commerce is not expected “to conduct a formal investigation to ensure that such prices are not dumped or subsidized,” but, instead, is to “base its decision [as to whether there is “reason to believe or suspect”] on information generally available to it at that time.” Id., H.R.Rep. No. 100-576 at 590-91, reprinted in 1988 U.S.C.C.A.N. 1547, 1623-24. 46

Neither Commerce nor the Government has pointed to anything in Congress’ text that suggests that legislators intended Commerce to blind itself to evidence of potential “dumped or subsidized prices.” Nor has Commerce or the Government cited anything in the legislative history that indicates that Congress intended Commerce to disregard a company’s financial statement only if a subsidy is explicitly evidenced on the face of that statement. Indeed, Congress’ directive to Commerce to base its determinations broadly on “[all] information generally available to [Commerce] at the time” would seem to cut against limiting the agency’s review solely to the face of companies’ financial statements alone.

A.S noted above, the legislative history states that Commerce is not required to investigate each claim of dumped or subsidized prices; to the contrary, Commerce is to bar a company’s data as non-representative even if the agency merely “suspect[s]” that the company has been the beneficiary of subsidies. By authorizing Commerce to disregard financial statements if the agency merely has “reason to ... suspect” a countervailable subsidy, the legislative history reflects a clear policy of erring on the side of rejecting financial statements — as opposed to establishing an irrebuttable presumption that a company did not benefit from an actionable subsidy if no such subsidy is evident on the face of the company’s financial statement (which, based on the existing record, appears to be Commerce’s practice). Cf. Zhejiang Mach. Import & Export Corp. v. United States, 31 CIT 159, 169, 473 F.Supp.2d 1365, 1374 (2007) (emphasizing that “[t]he ‘reason to believe or suspect’ standard is ... a relatively ‘low threshold,’ ” in rejecting challenge to agency’s use of surrogate prices for material inputs where actual market prices paid were suspected of reflecting subsidies). Ignoring evidence that a company has been the beneficiary of a subsidy program previously found to be countervailable unless the proof appears on the face of the company’s financial statement arguably guts Congress’ “reason to believe or suspect” standard, by reading the phrase “or suspect” right out of it. 47

In sum, Commerce is required to provide a reasonably discernable path to its decision, so as to support judicial review. See NMB Singapore, 557 F.3d at 1319. No such path is discernable here. It may be possible for Commerce to articulate and adequately support a justification for its practice of “relyfing] on information in financial statements on an ‘as is’ basis,” and ignoring all other record evidence. 48 But Commerce has not even attempted to do so yet. This matter therefore must be remanded to Commerce to permit the agency to reconsider and fully explain its stated practice (in light of, inter alia, the legislative history of the “reason to believe or suspect” standard, as well as the agency’s fundamental obligation to base its determinations on substantial evidence on the record as a whole), 49 to review and consider the evidence of alleged subsidies placed on the record by Xinboda (if appropriate), 50 and to explain in detail the rationale for the agency’s determination on remand.

b. Comparability of Xinboda, Tata Global, and Garlico

Just as it contests Commerce’s determination that Tata Global did not benefit from countervailable subsidies, so too Xinboda argues that Commerce erred in its determinations concerning the relative comparability of Xinboda, Tata Global, and Garlico. See Issues & Decision Memorandum at 20 (stating that Commerce “consideres] tea processing to be sufficiently similar to garlic processing”); see generally Pl.’s Brief at 30-32; Pl.’s Reply Brief at 7- 8, 9. Xinboda states its case bluntly, asserting that — for use in calculating Xinboda’s financial ratios — Commerce selected the financial statement of “a fully integrated plantation company manufacturing individually packaged teas and related beverages [i.e., Tata Global Beverages] over a company [ie., Garlico] that actually purchases, processes, and trades raw garlic and other raw vegetables comparable to raw garlic.” Pl.’s Brief at 27.

Raw Materials Consumed. Highlighting Commerce’s established practice of according substantial weight to the financial statements of potential surrogates that consume the same raw material as the respondent, Xinboda argues that Tata Global is not representative of Xinboda’s production experience because — unlike Xinboda and unlike Garlico — Tata Global does not consume or process garlic (or, for that matter, onions or any other raw vegetables). See Issues & Decision Memorandum at 19 n. 60 (stating that Commerce “favors the financial statements of surrogates that ... consume the identical raw material”); Pl.’s Brief at 34 (citing, inter alia, First Administrative Review of Steel Wire Hangers From the People’s Republic of China: Final Results and Final Partial Rescission of Antidumping Duty Administrative Review, 76 Fed.Reg. 27,994 (Dep’t Commerce May 13, 2011) and accompanying Issues and Decision Memorandum (May 9, 2011) (comment 2) (characterizing certain financial statements as “the best available information to calculate the surrogate financial ratios” because they reflect key relevant raw material input, and thus “most accurately reflect the production experience of the respondents”)) 51 ; see also Issues & Decision Memorandum at 20 (describing Tata Global as a “producer and seller of tea”); Pl.’s Brief at 34 (stating that Tata Global “consumes no garlic or comparable vegetables”); id. at 30, 35 (same); Pl.’s Reply Brief at 10 (same).

In contrast, as its name suggests, Garlico is a garlic processor. See Issues & Decision Memorandum at 22 (acknowledging Garlico’s production of garlic products) 52 ; see also id. at 11 (referring to Garlico as “an Indian garlic producer that purchases raw garlic bulbs for further preparation”); PL’s Brief at 27 (describing Garlico as “a company that ... purchases, processes, and trades raw garlic and other raw vegetables comparable to raw garlic”); id. at 9-10, 15, 18-19, 21, 23-24, 33, 34, 35, 36 (same); Pl.’s Reply Brief at 9-10 (same). Garlico’s financial statement documents the fact that the company consumed and processed large quantities of garlic, as well as onions and other similar vegetables. See Pl.’s Brief at 33; see generally id. at 33-35. Indeed, according to Garlico’s schedule of “Raw Materials Consumed,” Garlico’s consumption of garlic by volume outstrips its consumption of all other raw materials, with onion consumption a close second. See id. at 34.

Neither Commerce nor the Government has responded to Xinboda’s observation that the volume of garlic and onions that Garlico consumed in production dwarfs Garlico’s “trading sales” of those items, calling into question the significance that was attached to such sales in Commerce’s decision to disregard Garlico’s financial statement. See Pl.’s Brief at 36. 53 In any event, according to Xinboda, Commerce has an established methodology for handling traded goods so as not to skew financial ratios. Id. 54 Even more to the point, notwithstanding the importance of the factor in Commerce’s selection of financial statements, neither Commerce nor the Government has addressed Xinboda’s point that Xinboda and Garlico consumed the exact same raw material (ie., whole raw garlic bulbs), and Tata Global did not.

Production Processes (Including Extent of Integration). As discussed above, another key factor that Commerce considers in selecting financial statements is the comparability of “production processes” or “production experience.” See Issues & Decision Memorandum at 19 n. 60 (referring to “production experience”); OCTG Issues & Decision Memorandum (comment 13) (referring to “production processes”). Xinboda contends that Tata Global is not representative of Xinboda’s production experience because Tata Global is “fully integrated.” See PL’s Brief at 27, 30; see generally id. at 30-32. In contrast, both Xinboda and Garlico are non-integrated. See id. at 23 (explaining that integrated garlic producer both grows and processes garlic for export, and that Xinboda is non-integrated); id. at 23-24 (explaining that Garlico is a purchaser of raw vegetables such as garlic and onions, and that Garlico processed and sold some and traded some).

Xinboda emphasizes, as a threshold matter, that the level and extent of integration of a company’s operations are important considerations in evaluating the company’s production process — particularly given the facts of this case, which involve the application of Commerce’s “intermediate input” methodology. See Pl.’s Brief at 23, 30-31, 32. Under the intermediate input methodology, the respondent (Xinboda) does not report growing and harvesting factors of production, but, instead, reports the cost of the “intermediate input” — ie., the grown, harvested whole raw garlic bulb. See generally sections I & III.A.1, supra (summarizing and discussing Commerce’s intermediate input methodology). Embedded in the cost of that grown, harvested whole raw garlic bulb are all the growing and harvesting costs, “including the financial overheads, SG & A, and profit.” See Pl.’s Brief at 30. Xinboda argues that it was therefore “absolutely critical” for Commerce to avoid selecting the financial statements of “a grower” for use in calculating surrogate financial ratios in this case, in order to prevent “double-counting” of Xinboda’s “financial overheads.” Id. at 30-31, 32.

Xinboda’s admonition notwithstanding, Commerce selected the financial statement of Tata Global — a company that is, among other things, a major grower of coffee and tea. See PL’s Brief at 31. Xinboda points to the record evidence of Tata Global’s “extensive ownership” of “tea and coffee estates and other plantations across India” as support for the legitimacy of Xinboda’s concerns. See id. Previously, Commerce has given significant weight to Tata’s status as a grower in rejecting the company’s financial statement — even where the agency was not using its intermediate input methodology. See, e.g., Antidumping Duty Order on Fresh Garlic from the People’s Republic of China: Issues and Decision Memorandum for the Twelfth New Shipper Reviews (Sept. 19, 2008) (comment 3) (“Twelfth New Shipper Reviews Issues & Decision Memorandum”) (rejecting Tata financial statements where, inter alia, evidence indicated that company grew “over 70 percent of the tea it processe[d]”). Neither Commerce nor the Government has responded to this issue.

Other evidence similarly supports Xinboda’s more general concerns vis-a-vis the overall extent of the integration of Tata Global’s business, as well as the scope of its global operations. See, e.g.; Issues & Decision Memorandum at 20-21 (referring to “global aspects of Tata Tea’s business” and company “movement towards becoming a beverage business”); PL’s Brief at 27 (describing Tata Global as “a fully integrated plantation company manufacturing individually packaged teas and related beverages”); id. at 31 (referring to nature and extent of Tata Global’s “vertical integration”). 55

In the Final Results, Commerce tacitly acknowledges that Tata Global “is now an integrated producer in the global branded beverage business, with offices all over the world and employing thousands of employees.” See Issues & Decision Memorandum at 20. In recognition of the integrated nature of Tata Global’s operations, Commerce decided to use the company’s unconsolidated financial statement for purposes of the Final Results, rather than the con-' solidated financial statement that the agency used in the Preliminary Results. See id. at 20-21; see also Bet’s Response Brief at 22-28. However, neither Commerce nor the Government has offered any explanation as to how the change from Tata Global’s consolidated financial statement to its unconsolidated statement effectively mooted Xinboda’s concerns.

The Final Results minimize the extent of Tata Global’s vertical integration, citing information from notes to the company’s financial statement. Issues & Decision Memorandum at 21. According to Commerce, those notes indicate that Tata Global’s tea sales comprised more than 99% of total sales, and that integrated production and traded goods represent, respectively, 8.26% and 0.21% of the company’s sales. See id. But Xinboda contends that Commerce misinterpreted the notes, which— according to Xinboda — actually indicate that certain relevant costs were “loaded into numerator expenses that drive up ratios rather than included in the raw materials denominator that drives down surrogate factory overhead and SG & A.” See Pl.’s Brief at 31-32. 56 Neither Commerce nor the Government has addressed this point.

Xinboda concludes that the implications of the highly-integrated nature of Tata Global’s operations for Tata Global’s financial ratios are so pervasive and so profound that they cannot be remedied by the use of Tata Global’s unconsolidated financial statement, and that Commerce should have rejected Tata Global as a surrogate. See Pl.’s Brief at 31-32, 40. The qualified nature of Commerce’s own language in the Final Results — specifically, Commerce’s statement that the agency’s use of Tata Global’s unconsolidated financial statement addresses “[m ]any of Xinboda’s concerns” — lends further credence to Xinboda’s position, and underscores the need for an explanation by Commerce of the concerns that were left unaddressed. See Issues & Decision Memorandum at 20-21 (emphasis added); see also Pl.’s Brief at 33.

Xinboda also emphasizes differing production processes as support for its claim that Commerce improperly rejected the financial statement of Garlico and improperly relied on that of Tata Global. Xinboda contrasts the details of Tata Global’s tea production processes with the garlic production processes of Xinboda and Garlico. See, e.g., PL’s Brief at 27 (arguing that Commerce “draws inaccurately and incompletely from the record both in support of Tata Global Beverages and in its rejection of Garlico[ ]” in selecting “a fully integrated plantation company manufacturing individually packaged teas and related beverages over a company that actu ally purchases, processes, and trades raw garlic and other [comparable] raw vegetables”); id. at 32 (referring to “the many specific ways that [Tata Global’s] production experience was not representative of the garlic industry (some of which [Commerce] itself recognized in previous reviews wherein it rejected [Tata Global]”). 57

For example, Xinboda argues that, because — in recent years — Tata Global “has been focused on individually packaged teas and other beverages,” the company’s “processing is ... far more complicated than Xinboda’s.” Pl.’s Reply Brief at 7; id. at 9 (dismissing “[Commerce’s] conclusion that Garlico’s products are more highly processed than individually packaged tea products sold under several major Super Brands” as “incorrect on its face”). Xinboda emphasizes that Tata Global sells 90% of its tea in individually-packaged tea bags, which (Xinboda asserts) logically must drive up packing expenses (as well as marketing and branding expenses, and would also impact profit). See PL’s Brief at 35; PL’s Reply Brief at 8 (asserting that Tata Global’s “packaging operations” are “obviously far more vast and complex than those of Xinboda, which just has to stick a handful of customer labels on some jars or mesh bags of fresh garlic”). In prior reviews, Commerce has pointed to Tata Global’s focus on the production of individually-packaged teas, instant tea, and other non-bulk, “value-added” tea products as a basis for rejecting Tata Global’s financial statement. See, e.g., Issues and Decision Memorandum for the Administrative Review of the Antidumping Duty Order on Fresh Garlic from the People’s Republic of China (June 13, 2005) (comment 5) (“Ninth Administrative Review Issues & Decision Memorandum”); Twelfth New Shipper Reviews Issues & Decision Memorandum (comment 3). 58 But neither Commerce nor the Government has addressed these facts here.

Xinboda details the relatively modest production processes of Xinboda, its affiliate processor/producer Dadi, and Garlico. See PL’s Brief at 8-9, 11-12, 22-23, 35; PL’s Reply Brief at 7, 8; see also section I, supra (providing detailed summary of production processes of Xinboda and its affiliated processor/producer, Dadi). Xinboda points out that Commerce rejected Garlico’s financial statement based largely on Commerce’s belief that the “downstream products” produced by Garlico — ie., dehydrated and powdered garlic products — “require further production than that required for peeled or whole, fresh garlic.” See Issues & Decision Memorandum at 22. But Xinboda maintains that the similarities between the processes of Xinboda’s affiliate and Garlico greatly outweigh any differences, and that the production processes of the two have much more in common with one another than they do with processes for the production of tea (particularly non-bulk tea, which is principally what Tata Global produces). See Pl.’s Brief at 36 (concluding that, “[o]verall, the record reveals a remarkable comparability between Garlico and Xinboda and many serious points of dissimilarity between Tata Global Beverages and Xinboda”); id. at 35 (concluding that Tata Global “is not representative of Xinboda overall”); see also id. (stating that “Tata sells 90% of its tea in individually packed bags”).

Xinboda notes, for example, that virtually the entirety (ie., 98%) of Garlico’s installed capacity is dedicated to garlic processing (in contrast to Tata Global, which has none). See Pl.’s Brief at 33-34. Xinboda similarly points to the schedules to Garlico’s financial statement, which identify two significant operations that Garlico shares with Xinboda’s production of peeled garlic: “water expenses” and “sterilization expenses.” See id. at 35. In other words, as Xinboda explains, “for Garlico to make garlic flakes and powder or onion flakes or powder, it must buy the raw materials, peel the skins, and sterilize the vegetables — just like Xinboda would do for peeled garlic cloves. The remaining steps, [ie.,] further drying and flaking or pulverizing[,] can be done with simple machines.” Id. Xinboda concludes that, if Commerce were correct that Garlico’s production processes were so much more sophisticated and/or extensive (as Commerce suggests), then “Garlico’s overheads for factory overhead and SG & A would be overstated [ie., higher].” Id. But, according to Xinboda, “Garlico’s combined factory overhead and SG & A are merely two-thirds as high as those of [Tata Global].” Id.

The bases for Commerce’s conclusion that “tea processing ... [is] sufficiently similar to garlic processing” were the agency’s findings that “neither product is highly processed or preserved prior to sale.” Issues & Decision Memorandum at 20. To the contrary, however, it appears that Xinboda’s peeled garlic is preserved. See Pl.’s Brief at 22-23 (explaining that jars of peeled garlic “are injected with a preservative gas and vacuum sealed”). Xinboda similarly takes issue with the second basis for Commerce’s conclusion — ie., Commerce’s finding that neither Xinboda’s garlic products nor Tata Global’s tea products are “highly processed.” See Issues & Decision Memorandum at 20. In particular, Xinboda argues that — relatively speaking — tea “undergoes significantly more processing” than either whole or peeled garlic. Pi’s Brief at 35 (emphasis added); Pl.’s Reply Brief at 7 (explaining that Tata Global’s processing is “indeed far more complicated than Xinboda’s”); see also PL’s Brief at 8-9, 22-23 (describing production processes for whole and peeled garlic).

Moreover, some stages of the processing that tea undergoes — specifically, the drying and crumbling of tea leaves — essentially mirror the processing that Commerce cited as a basis for rejecting the financial statement of Garlico here. See Pl.’s Brief at 35; Issues & Decision Memorandum at 22 (explaining agency’s rejection of Garlico financial statement, based in part on list of products described as “dehydrated” and “powder”). As such, Xinboda posits, if Commerce had cause to reject the financial statement of Garlico (a producer of, inter alia, dehydrated and powdered garlic products), then Commerce had even greater cause to reject the financial statement of Tata Global (which produces dried and crushed tea leaves, and no garlic products of any kind whatsoever). See Pl.’s Brief at 35.

In prior reviews, Commerce has rejected use of Tata Global’s financial statements based on the agency’s determinations that the company’s products are “more highly processed or preserved prior to sale” than fresh garlic products, that the company’s processing operations are “more advanced,” and that the company’s “production experience” did not match that of garlic producers. See, e.g., Twelfth New Shipper Reviews Issues & Decision Memorandum (comment 3); Ninth Administrative Review Issues & Decision Memorandum (comment 5). Yet neither Commerce nor the Government has sought to distinguish this review from the other reviews in which the agency has disregarded Tata Global’s financial statements. Nor has Commerce or the Government otherwise responded to the points that Xinboda has raised.

As a final aspect of production processes and experiences, Xinboda turns to the nature and extent of Tata Global’s “vastly different global selling model.” PL’s Reply Brief at 7. Specifically, Xinboda contends that Tata Global’s production experience is not comparable to the production experiences of Xinboda and Garlico, due to Tata Global’s selling practices, which involve higher, unrepresentative packaging, marketing, and branding expenses, and profit. See generally PL’s Brief at 23, 32-33, 35; PL’s Reply Brief at 7-8. 59

Xinboda analogizes its sales practices to those of Garlico, noting that — unlike Xinboda and Garlico — Tata Global develops, markets, and sells its own branded products, including such major international labels as “Tetley” teas, in addition to “Tata” teas (which are accorded “Super Brand” status in India), as well as many other trademarked tea and coffee products. PL’s Brief at 32-33; PL’s Reply Brief at 7 (same); see also PL’s Brief at 35 (highlighting impact on marketing and branding expenses of fact that 90% of Tata Global’s tea sales are in individually-packaged bags); PL’s Reply Brief at 7 (same). In sharp contrast, Xinboda’s sales process is quite modest, involving an established customer base and virtually zero marketing. See PL’s Brief at 23. Xinboda sells very few products (i.e., garlic, onion shoots, and ginger) and does not develop or market its own brands. PL’s Brief at 23. Xinboda merely labels its garlic products as instructed by its customers, which design and market their own brands and provide them label artwork to Xinboda. See PL’s Brief at 23, 32; PL’s Reply Brief at 8. It is thus Xinboda’s customers that incur the SG & A associated with — and reap the profits that flow from — developing, marketing, and selling branded products. Id. 60

According to Xinboda, Tata Global’s extensive branding and marketing, summarized above, contribute to “aberrantly high” surrogate values for SG & A and profit. See Pl.’s Brief at 33 (stating that Tata Global’s SG & A is “nearly three times higher than that of Garlico” and that Tata Global’s profit is “approximately nine times higher than [that of] Garlico”); see also PL’s Reply Brief at 7-8 (arguing that Tata Global is “much more highly integrated on the sales side of its business”). 61 Xinboda faults Commerce for “ignoring” Xinboda’s claim that the nature and extent of Tata Global’s branding, marketing, and sales render the company unrepresentative as a potential surrogate. See PL’s Brief at 33. Commerce has previously taken note of the extent of Tata Global’s “branding” in explaining the agency’s rejection of the company’s financial statement. See, e.g., Ninth Administrative Review Issues & Decision Memorandum (comment 5) (rejecting Tata financial statement because, inter alia, “eighty-six percent of [the company’s] turnover is a result of its branded tea products”). However, neither Commerce nor the Government has addressed the relevant facts here.

Nature and End Uses of Merchandise. One final factor in selecting financial statements for use in calculating surrogate financial ratios is the nature of the merchandise that a potential surrogate produces (identical, comparable, or otherwise); and the factors that Commerce considers in evaluating whether merchandise is comparable include both “physical characteristics” and “end use.” See 19 C.F.R. § 351.408(c)(4) (referring to “identical or comparable merchandise”); Issues & Decision Memorandum at 19 n. 60 (referring to “identical merchandise”); OCTG Issues & Decision Memorandum (comment 13) (referring to “physical characteristics” and “end uses” of merchandise).

Xinboda argues that Commerce “ignores” the end uses of the companies’ products. See PL’s Brief at 35-36 (citing OCTG Issues & Decision Memorandum (comment 13)); see Issues & Decision Memorandum at 18-22. Nor does Commerce analyze the products’ respective physical characteristics. See id. The Final Results are similarly silent on the ultimate comparability of Tata Global’s products and the whole fresh garlic and peeled garlic that Xinboda exported. See id.

It is, however, undisputed that Tata Global produced no garlic products of any kind. It thus seems clear, at least at first blush, that Garlico’s garlic products are more “comparable” to Xinboda’s products than are Tata Global’s products, even assuming for the sake of argument that Garlico’s garlic products involved more processing than Xinboda’s products. As Xinboda observes, garlic and onions are spice commodities in the same botanical family, whereas tea is a beverage — and, in the case of Tata Global, “a heavily paekaged[,] heavily branded, [and heavily] marketed product” at that. See PL’s Brief at 35-36.

Summary. The law requires Commerce to make a reasoned decision as to the surrogate financial statement(s) on which it chooses to rely, and to both adequately explain its rationale and support its decision with substantial evidence. As outlined above, Commerce’s determination concerning the relative comparability of Xinboda, Tata Global, and Garlico is not sufficient and therefore must be remanded for further consideration.

Accordingly, in the course of the remand to review and reconsider, inter alia, Commerce’s interpretation and application of the “reason to believe or suspect” standard and (if appropriate) Xinboda’s evidence of subsidies allegedly received by Tata Global (as discussed in section III.C.2.a above), Commerce also shall review and reconsider its findings and conclusions concerning the relative comparability of Xinboda, Tata Global, and Garlico, in light of the agency’s preference for the use of multiple financial statements, as well as the criteria set forth in 19 C.F.R. § 351.408(c)(4), and the agency’s policy favoring the financial statements of surrogates that produce identical merchandise, consume the identical raw material, and have identical or comparable production experience. In considering whether merchandise is comparable (if not identical), Commerce shall consider the physical characteristics and end uses of the merchandise, in addition to the production process. Commerce also shall explain the weight accorded to each of the factors that the agency considers in evaluating potential surrogates, including, for example, whether Commerce assigns greater weight to the factor of identical raw material inputs. In addition, Commerce shall identify, explain, and apply any other factors, criteria, or standards that it typically considers in evaluating surrogate financial statements for potential use in a case such as this.

Applying the relevant factors, criteria, and standards, Commerce shall reconsider its determination on the relative comparability of Xinboda, Tata Global, and Garlico, in light of Xinboda’s arguments, all record evidence, and the analysis set forth above (including Commerce’s determinations in other reviews of the antidumping order on fresh garlic from the PRC). And Commerce shall explain in detail the bases for its redetermination on remand, supporting all findings and conclusions by reference to substantial evidence in the record.

D. Zeroing

Xinboda’s final claim challenges Commerce’s use of its controversial — and now abandoned — “zeroing” methodology in calculating Xinboda’s weighted-average dumping margin in the underlying administrative review. See Issues & Decision Memorandum at 31-33; Pl.’s Brief at 1, 2-7, 40; PL’s Reply Brief at 14-15. 62 Xinboda maintains that, rather than assigning a value of zero to sales where merchandise was sold at prices above the calculated normal value, Commerce instead should have used such sales to offset other sales where merchandise was sold at prices below normal value. See PL’s Brief at 2-3. Xinboda claims — both generally and with specific reference to the facts of this case in particular — that zeroing in the subject administrative review was “inconsistent with [Commerce’s] policy of not zeroing negative margins in [original, underlying antidumping] investigations” and, further, was otherwise “improper.” See PL’s Brief at 2-3 (citing Dongbu Steel Co. v. United States, 635 F.3d 1363 (Fed.Cir.2011)); see also PL’s Brief at 4-7; PL’s Reply Brief at 14-15.

Xinboda concedes that the court in Union Steel sustained Commerce’s interpretation of 19 U.S.C. § 1677(35) as permitting the agency to continue the use of zeroing in administrative reviews (involving average-to-transaction comparisons) while discontinuing the practice in original antidumping investigations (involving average-to-average comparisons) — a decision that the Court of Appeals upheld as reasonable. See Pl.’s Brief at 3 (citing Union Steel v. United States, 36 CIT -, 823 F.Supp.2d 1346 (2012), aff'd, 713 F.3d 1101 (Fed.Cir.2013)). However, Xinboda emphasizes that Commerce more recently has stated that it is generally abandoning the practice of zeroing even in administrative reviews — at least as to those reviews with preliminary results published on or after April 16, 2012. See PL’s Brief at 4 (citing Antidumping Proceedings: Calculation of the Weighted-Average Dumping Margin and Assessment Rate in Certain Anti-dumping Duty Proceedings; Final Modification, 77 Fed.Reg. 8101 (Dep’t Commerce Feb. 14, 2012)). 63

Even more to the point, Xinboda contends that the Union Steel rationale “does not apply to [non-market economy] cases” and “especially does not apply to the Garlic from China reviews.” PL’s Brief at 6; see generally id. at 3, 6-7; PL’s Reply Brief at 14. As Xinboda frames its argument:

[Commerce] does not apply average monthly costs to individual U.S. sales in [non-market economy] cases. Rather, the Department determines a single average cost for the twelve-month [period of review] (18 months in first reviews). That is, the Department normally collects twelve months of import values from all over the world into India per raw material input and arrives at one single average unit value for each. Similarly, the Department finds one single [period of review] surrogate cost for energy and labor, etc. As for the key surrogate costs of factory overhead, SG & A and profit, they are almost never co-extensive with [non-market economy periods of review] because the Indian fiscal year is April 1 through March 31. Thus, in this case, [Commerce] again took a single twelve-month cost for April 1, 2009 through March 31, 2010, even though the [period of review] runs from November 2008 through October 2009. In this case, the disconnect from “monthly averages” is even more severe than most cases due to the Department’s selection of the major surrogate values for bulb inputs and financial ratios. In this review ... the Department inflated a pre-[period of review] value for “S.A.” grade garlic to the [period of review] and selected a top-branded beverage company over a garlic processing and trading company for the surrogate financial ratios. Furthermore, it is important to note that the information (including inflators and financial statements) is never available by the time the [non-market economy] exporter must price U.S. sales — even if the respondent were to correctly guess which sources the Department would ultimately select.

PL’s Brief at 6-7.

Xinboda seeks to illustrate its point with “a simple example”:

[A]ssume the cost of garlic inputs or production rises one penny per month of the [period of review]. Assume further that the respondent did raise U.S. prices each month to compensate for this. However, the [non-market economy] respondent is stuck with the mid-point cost based on [Commerce’s] methodology. This drives some sales above and some sales below normal value by the very methodology in a way that is basically indistinguishable from average-to-average comparisons in investigations.

Pl.’s Brief at 7. Xinboda concludes that, “[t]aken in combination, the unique facts applicable to [non-market economy] normal value ... compel the conclusion that zeroing negative margins is unreasonable.” Id. According to Xinboda, “offsetting (i.e., abandoning zeroing) is even more important to the accurate calculation of margins in the [non-market economy] review context than it might be in the context of a market economy investigation.” Id.; see also id. at 3 (characterizing claim as “an issue of first impression”). 64

The Government acknowledges that, in the Final Results, “Commerce did not provide a full explanation about its interpretation of the statute permitting offsetting of margins in certain proceedings but not in others.” Def.’s Response Brief at 29; see also Issues & Decision Memorandum at 31-33 (comment 10). The Government therefore requests a voluntary remand, “so that Commerce may revisit the zeroing issue and, as necessary, provide the Court with the additional explanation it provided [in] Union Steel.” Del’s Response Brief at 29-30; see also id. at 2, 6. The Domestic Producers support a voluntary remand. See Def.-Ints.’ Response Brief at 2 (endorsing Government’s request for voluntary remand “to address judicial determinations addressing the issue of zeroing” that post-dated publication of Final Results in this case). Xinboda, however, objects. See generally PL’s Reply Brief at 14-15.

Xinboda argues that it “does not perceive any benefit to a remand for the purpose stated by [the Government].” PL’s Reply Brief at 14. According to Xinboda, a remand for the sole purpose of allowing Commerce to provide the explanation that the agency provided in Union Steel would be futile, because Union Steel did not present the specific question that Xinboda presses in the instant case — that is, the reasonableness of zeroing in administrative reviews such as the one at issue here, involving imports into the United States from a non-market economy country. See id.

The Government correctly observes, however, that SKF counsels in favor of granting an agency’s request for a voluntary remand where — as here — the agency seeks to reconsider intervening legal developments that “may affect the validity of the agency action.” See Def.’s Response Brief at 30 (quoting SKF USA, Inc. v. United States, 254 F.3d 1022, 1028-29 (Fed.Cir.2001)). Moreover, Xinboda’s reservations notwithstanding, it cannot be said that the requested remand would be futile. See generally Nippon Steel Corp. v. United States, 458 F.3d 1345 (Fed.Cir.2006). 65 All parties agree that the expla nation of zeroing set forth in the Final Results is insufficient. It would be inefficient to deprive Commerce of the opportunity to review its position in light of the numerous intervening legal developments and to clearly set forth the bases for its determination. A remand will allow Commerce to consider Xinboda’s arguments, apply the agency’s expertise, articulate its interpretation of the pertinent statute, and explain the basis for its action in this case; and, conceivably, remand may obviate entirely the need for further judicial review. See generally 2 K. Davis & R. Pierce, Jr., Administrative Law Treatise §§ 15.2 (citing McKart v. United States, 395 U.S. 185, 193-95, 89 S.Ct. 1657, 23 L.Ed.2d 194 (1969)), 15.12 (3d ed. 1994) (discussing doctrine of exhaustion of administrative remedies).

The Government’s request for a voluntary remand on this issue therefore must be granted. On remand, Commerce shall review its application of the zeroing methodology in this case and shall set forth in full the rationale for its redetermination, in light of Xinboda’s arguments concerning administrative reviews in non-market economy cases generally and in garlic cases in particular, as well as all relevant intervening legal developments (including, but not limited to, the decisions in Union Steel and the agency’s decision to discontinue the practice of zeroing in administrative reviews in general).

IV. Conclusion

For all the reasons set forth above, this matter must be remanded to Commerce for further consideration of the surrogate value for whole raw garlic bulbs, the surrogate wage rate, and the surrogate financial ratios, as well as to permit the agency to review the application of its “zeroing” methodology in the administrative review at issue in light of Xinboda’s arguments and all relevant intervening legal developments.

A separate order will enter accordingly.

1 . Because the administrative record in this action includes confidential information, two versions of the record were filed with the Court. The public version of the administrative record consists of copies of all documents in the record, with all confidential information redacted. The confidential version consists of complete, un-redacted copies of only those documents that include confidential information. Citations to documents in the public record are noted as “Pub. Doc. No.

2 . All citations to statutes herein are to the 2006 edition of the United States Code. Similarly, all references to regulations are to the 2008 edition of the Code of Federal Regulations. The pertinent text of the statutes and regulations cited remained the same at all times relevant herein.

3 . In addition, in certain market economy cases, Commerce may calculate normal value üsing the price in a third country {i.e., a country other than the exporting country or the United States). See, e.g., RHP Bearings Ltd. v. United States, 288 F.3d 1334, 1338 (Fed.Cir.2002) (discussing 19 U.S.C. §§ 1677b(a)(l)(B)(ii), 1677b(a)(l)(C)); see also Ningbo Dafa Chem. Fiber Co. v. United States, 580 F.3d 1247, 1251 n. 1 (Fed.Cir.2009) (explaining exception).

4 .Factors of production "include, but are not limited to ... hours of labor required, ... quantities of raw materials employed, ... amounts of energy and other utilities consumed, and ... representative capital cost, including depreciation.” See 19 U.S.C. § 1677b(c)(3); see also Dorbest Ltd. v. United States, 604 F.3d 1363, 1367 (Fed.Cir.2010) ("Dorbest IV") (discussing factors of production).

5 . For a summaty overview of Commerce's intermediate input methodology and the history of its application to garlic from the PRC, see Jining Yongjia Trade Co. v. United States, 34 CIT -, - & n. 6, 2010 WL 5121964 *2-5 & n. 6 (2010) (explaining, inter alia, that, when Commerce employs its intermediate input methodology, "the cost (or value) of the whole garlic bulb [is] used as a substitute for the costs of growing and harvesting [factors of production] ('upstream [factors of production]') actually reported by [the foreign producer at issue]”).

Circumstances where Commerce has used its intermediate input methodology include cases where the intermediate input "accounts for an insignificant share of total output” and "the potential increase in accuracy to the overall calculation that [would] result[ ] from valuing each of the [factors of production] is outweighed by the resources, time, and burden such an analysis would place on all parties to the proceeding.” Fresh Garlic from the People’s Republic of China: Preliminary Results of, Partial Rescission of, and Intent to Rescind, in Part, the 15 th Antidumping Duty Administrative Review, 75 Fed.Reg. 80,458, 80,462 (Dep’t Commerce Dec. 22, 2010) ("Preliminaiy Results”) (citing Notice of Final Determination of Sales at Less Than Fair Value: Polyvinyl Alcohol from the People's Republic of China, 68 Fed.Reg. 47,538 (Dep’t Commerce Aug. 11, 2003) and accompanying Issues and Decision Memorandum (Aug. 4, 2003)).

6 . In the Issues & Decision Memorandum, Commerce states that it prefers, "where possible and in no particular order, a publicly available value which is: (1) an average non-export value; (2) representative of a range of prices within the period of investigation/POR; (3) product-specific; and, (4) duty and tax-exclusive.” Issues & Decision Memorandum at 11-12; see also id. at 12 (emphasizing that "[Commerce] has consistently stated that it prefers to select surrogate values that are: (1) for products as similar as possible to the input being valued; (2) contemporaneous with, or closest in time to the [period of review]; and (3) representative of a range of prices in effect during the [period of review]”); Import Administration Policy Bulletin 04.1, Non-Market Economy Surrogate Country Selection Process, at "Data Considerations" (March 1, 2004) (stating that, "[i]n assessing data and data sources, it is [Commerce's] stated practice to use investigation or review period-wide price averages, prices specific to the input in question, prices that are net of taxes and import duties, prices that are contemporaneous with the period of investigation or review, and publicly available data”); see generally Jinan Yipin Corp. v. United States, 35 CIT -, - n. 7, 800 F.Supp.2d 1226, 1234 n. 7 (2011) ("Jinan Yipin II”) (explaining that five criteria set forth in Policy Bulletin originally were promulgated for Commerce’s use in identifying a surrogate country, but that agency nevertheless has frequently cited Policy Bulletin as establishing criteria that guide agency's selection from among alternative data sources after surrogate country has been identified).

7 . As the Court of Appeals has explained:

Every year after an [antidumping order] issues, an interested party can request that Commerce conduct an administrative review of the order. In this review, Commerce analyzes the actual merchandise imported throughout the previous year that is subject to the order. (In some administrative reviews, Commerce analyzes the merchandise imported over the previous year and a half.) This system, often described as a retroactive system, enables Commerce to calculate a final duty rate based on the actual imports themselves as opposed to information obtained before importation even began.

Sioux Honey Ass’n v. Hartford Fire Ins. Co., 672 F.3d 1041, 1047 (Fed.Cir.2012); see also Union Steel, 713 F.3d at 1103 (stating that “[a]ny exporter of ... goods subject to [an] antidumping order may annually request an administrative review to determine the exact amount by which the foreign market value exceeds the U.S. price and assess the precise amount of duties owed for [its] exports”).

8 . Xinboda explains that "[m]ost of the whole garlic bulb that Xinboda purchases as the major input for both [whole garlic and peeled garlic] is delivered directly from the farms that surround Dadi’s production facilities.” PL's Brief at 8. Specifically, ‘‘[flor Dadi’s peeled garlic production during the [period of review], 93% of the raw garlic was delivered from farms no further than 15 km away from Dadi’s facility, and for the whole garlic product, 70.65% was delivered from a distance no further than 12.3 km.” Id.

9 . See Jining Yongjia Trade Co., 34 CIT at -, 2010 WL 5121964 *3-5 (explaining in detail the various reasons for Commerce’s decision to use agency's intermediate input methodology to value whole raw garlic bulbs, beginning with tenth administrative review).

10 . The Issues & Decision Memorandum mistakenly states that Commerce has used its intermediate input methodology in valuing garlic since the 2004-2005 administrative review. See Issues & Decision Memorandum at 11. In fact, Commerce first used the methodology in the tenth administrative review, which covered November 1, 2003 through October 31, 2004. See Fresh Garlic from the People's Republic of China: Final Results and Partial Rescission of Antidumping Duty Administrative Review and Final Results of New Shipper Reviews, 71 Fed.Reg. 26,329, 26,329 (Dep't Commerce May 4, 2006) (specifying period of review for tenth administrative review); id. at 26,330-31 (noting application of intermediate input methodology).

11 . As lining Yongjia points out, Commerce briefly summarized the difference between the "intermediate input” — i.e., the whole raw garlic bulb — and the finished product, fresh garlic, in the agency's Issues and Decision Memorandum in the tenth administrative review:

[T]he raw garlic bulb that is harvested from the ground ... is not immediately shipped to the United States. Rather, the garlic that PRC exporters ship to the United States requires at least a minimum amount of processing and packing prior to export.... [T]he garlic harvested from the ground is, at a minimum, cleaned to remove the outer skins in order to give the garlic bulb its characteristic white, fresh appearance. This whole [garlic bulb] is then typically packed in mesh bags and cartons for shipment to the United States. In the case of peeled garlic, the processing is more extensive and typically involves additional labor, energy, and several packing inputs (including the use of an antiseptic solution and nitrogen gas).

Issues and Decision Memorandum for the Administrative Review and New Shipper Reviews of the Antidumping Duty Order on Fresh Garlic from the People's Republic of China (April 26, 2006) (tenth administrative review) (comment 1) (quoted in Jining Yongjia Trade Co., 34 CIT at - n. 6, 2010 WL 5121964 *2 n. 6).

12 . To value garlic with a bulb diameter of between 40 and 55 mm, Commerce used a combination of Azadpur APMC data for grades "A” and "S.A.’' garlic, due to the seeming overlap in the physical characteristics of the two grades — i.e., because, depending on traits other than bulb size, garlic with a diameter of between 40 and 55 mm could be classified as either grade "A” or grade "S.A.” (at least during the period from May 2006 to February 2008, when the Azadpur APMC Market was using both of those grades). See Preliminary Surrogate Value Memorandum at 4.

13 . In the Final Results, Commerce makes much of the fact that the Azadpur APMC data were used "[i]n the past three reviews” (i.e., the twelfth, thirteenth, and fourteenth reviews) of the underlying antidumping order here, as well as the fact that, in Jining Yongjia (involving the twelfth review), the court upheld Commerce’s decision to rely on Azadpur APMC data, rather than data from other potential data sources — specifically, data from the World Trade Atlas ("WTA") and data from the Indian Agricultural Marketing Information Network ("AGMARKNET”), a database maintained by India’s Ministry of Agriculture. See Issues & Decision Memorandum at 11, 13 & n. 37; Jining Yongjia Trade Co., 34 CIT at -, 2010 WL 5121964 *8-13.

To place Commerce’s observations in a more complete context, it is worth noting that Commerce first sought to use the Azadpur APMC data in the tenth administrative review. However, Commerce ultimately abandoned the Azadpur APMC data for purposes of that review, and instead relied on AGMARKNET data, after concerns were raised about, inter alia, whether the Azadpur APMC prices were, in fact, prices for an "intermediate input” at the "farm gate.” See Final Remand Results of Redetermination Pursuant to Second Remand at 7, 23 (March 29, 2012), filed in Jinan Yipin Corp. v. United States, Court No. 06-00189; Jinan Yipin II, 35 CIT at -, -, -, 800 F.Supp.2d at 1257-58, 1264, 1268-73 (discussing "record evidence suggest[ing] that [Commerce’s calculations based on the Azadpur APMC data] may not have valued the intermediate input at all, and — instead—may have valued a final product”). The referenced remand results in Jinan Yipin post-date the court’s decision in Jining Yongjia that Commerce relied on in the Final Results here.

Finally, it goes without saying that each administrative review involves different arguments and a different administrative record (and, not infrequently, different data sets and different parties) and therefore must be considered on its own merits.

14 . In addition to the Azadpur APMC data, other potential sources of data for use in calculating a surrogate value for whole raw garlic bulbs include Indian World Trade Atlas ("WTA”) import statistics; Indian export statistics; Indian domestic market data from government sources, including data from India's National Horticultural Board and data from the Indian Spices Board, as well as data from the Indian Agricultural Marketing Information Network ("AGMARKNET”), a database maintained by India’s Ministry of Agriculture; and garlic pricing information included in the 2009-2010 financial statements of the Indian garlic processor and trader Garlico. See Issues & Decision Memorandum at 11.

15 . The Azadpur APMC Market did not begin classifying garlic as grade "S.A.” until May 1, 2006; and the classification has not been used since early February 2008. See Issues & Decision Memorandum at 13; Pl.'s Brief at 19; Def.'s Response Brief at 16; Jinan Yipin II, 35 CIT at - n. 41, 800 F.Supp.2d at 1258 n. 41.

16 . According to Xinboda, “Garlico reports average prices of 2.338 Rs/kg in 2009-10 and 3.257 Rs/kg in 2008-09 for the garlic it purchased for processing^] and 5.34 Rs/kg in 2009-10 and 3.00 Rs/kg in 2008-09 for the garlic it traded.” Pl.’s Brief at 15; see also id. at 9-10; PL’s Reply Brief at 3.

17 . According to Xinboda, based on the contemporaneous Azadpur APMC data, the average price for “A”-grade garlic is 23.517 rupees per kilogram. See PL’s Brief at 21.

18 . See Pl.’s Brief at 11-12 (acknowledging that garlic received by Dadi was "in large mesh bags” and was "pre-sorted according to size,” and that garlic arriving at Azadpur APMC Market is similarly "packed in large mesh bags and sorted according to size”); see also PL’s Reply Brief at 6 (same).

19 . Xinboda wryly observes that its affiliated processor/producer, Dadi, "does not travel to Beijing to visit the urban farmer's market to purchase its garlic. Rather, local farmers deliver garlic [to Dadi] from surrounding farms.” PL’s Brief at 18.

20 . See also Jinan Yipin II, 35 CIT at -, 800 F.Supp.2d at 1270-71 (noting that “the apparent involvement of intermediaries” in sales at the Azadpur APMC Market “substantiates the Chinese Producers' concerns that the prices included in the Azadpur APMC data may include costs, fees, and commissions that hike up the prices” and "undermines Commerce’s claims that the Azadpur APMC data reflect the price of large-bulb Indian garlic at the ‘farm gate’ and that those data ... are representative of the value of the 'intermediate input’ at issue”).

21 . Xinboda explains that "AgriCoop is an Indian governmental entity and, according to [Commerce’s] long-standing policy, a reliable source of information on the status of India’s [Agricultural Produce Marketing Committee markets].” PL's Brief at 14.

22 . In the Final Results, Commerce describes the Azadpur APMC data as "wholesale market prices.” See Issues & Decision Memorandum at 14. As the 2003 Market Research Report recognizes, however, “wholesale” prices are not the same as "farm gate” prices. See 2003 Market Research Report at 20-21.

The 2003 Market Research Report and the two subsequent supplements are frequent references in reviews of the antidumping order covering fresh garlic from the PRC. See 2003 Market Research Report; Clarifications on Garlic Study (Oct. 2006) (Pub.Doc. No. 133) ("2006 Market Research Report Update”); Clarifications on Surrogate Values Supplemental Questionnaire (July 2007) (Pub.Doc. No. 131) ("2007 Market Research Report Update”); see also, e.g., Qingdao Sea-Line Trading Co. v. United States, 37 CIT -, -, -, 2013 WL 4038618 *4, *6 (2013) (stating that Commerce placed both 2003 Market Research Report and 2007 update on record in course of remand in 2008-2009 new shipper review); Jinan Yipin II, 35 CIT at-& n. 10, passim, 800 F.Supp.2d at 1237 & n. 10, passim (explaining that Final Results in tenth administrative review relied in large part on 2003 Market Research Report, discussing report in context of remand results in tenth review, and noting that report was first placed on record of eighth administrative review); Taian Ziyang Food Co. v. United States, 35 CIT -, - n. 11, - & nn. 12-13, 783 F.Supp.2d 1292, 1303 n. 11, 1304 & nn. 12-13 (2011) ("Taian Ziyang II") (noting, inter alia, that Commerce placed 2003 Market Research Report on record during second remand proceeding in ninth administrative review).

23 . Although no party has briefed the point, it appears that Commerce already has "subtracted a 7% fee (6% commission fee plus 1% market fee) charged on transactions at the Azadpur APMC [Market]” from the Azadpur APMC data used in the Final Results. See Preliminary Surrogate Value Memorandum at 4.

24 .Although Xinboda correctly notes that the rationale for discontinuing the usage of the "S.A." classification has no bearing on the issue here, Xinboda advises that the administrative case brief that it filed with Commerce following issuance of the Preliminary Results "discussed at length possible reasons as to why traders on the Azadpur market discontinued the use of the designation 'grade S.A.1 garlic.” See PL’s Brief at 20.

Further, according to Xinboda, the use of the grade "S.A.” classification appears to have been limited to the Azadpur APMC Market. Xinboda thus explains that “[t]here is no indication on this record or in any other administrative or new shipper review of fresh garlic from China that the appellation 'S.A.1 was used in any ... Indian market [other than the Azadpur APMC Market] or in Indian retail stores during this or earlier periods of review to describe a particular size and/or quality of Indian garlic.” Pl.’s Brief at 19. Xinboda similarly states that its “[e]xtensive searches on the global trade website ‘Alibaba.com' on garlic from India" turned up "no designation 'S.A.' for garlic from India.” Id. at 20.

Xinboda emphasizes that "[n]o party to this case has provided any evidence that [the] designation ["S.A.”] was applied to large-bulb garlic on any Indian market after its use was discontinued on the Azadpur market in February 2008.” Pl.’s Reply Brief at 4; see also PL's Brief at 20 (emphasizing that Domestic Producers “have offered no rebuttal whatsoever to [the] record evidence that no ‘Grade S.A.’ garlic exists in the Indian markets”). "Indeed,” Xinboda concludes, “there is no record evidence at all that the designation 'S.A.’ has been used anywhere in India since February 5, 2008.” Id.

25 . Significantly, neither Commerce nor the Domestic Producers have offered any explanation as to how garlic that previously would have been classified as "S.A.”-grade has been classified since January 2008, if such garlic was not subsumed within the grade "A” classification (as Xinboda contends it was). Certainly Commerce and the Domestic Producers do not contend that such high-quality, large-bulb garlic is no longer offered for sale at the Azadpur APMC Market. Their silence on this point speaks volumes.

26 . To the same effect, Xinboda placed on the record the results of its “[e]xtensive searches on the global trade website 'Alibaba.com' on garlic from India,” which "revealed that 'Grade A’ garlic covers sizes 30-70 mm and is sometimes described as ‘super white garlic.’ ” See Pl.’s Brief at 20.

27 . See Pl.’s Reply Brief at 4-5 (noting that, notwithstanding ample time to do so, neither Commerce nor the Domestic Producers proffered any evidence to rebut the information provided by Xinboda’s research consultant).

28 . The Final Results explained the selection of the Azadpur APMC data over the Garlico price data on the grounds that the Garlico data were not "product-specific or size-specific” and that there was no evidence that "Garlico's raw garlic purchases are in any way representative of a broad market average, ... or are even reflective of Xinboda’s operations.” See Issues & Decision Memorandum at 13; see also Def.'s Response Brief at 15-16. As outlined above, however, the Azadpur APMC data too suffer from a number of serious flaws. And the Final Results are much too quick to reject the Garlico data. On remand, Commerce must carefully reevaluate the relative merits of the various alternative data sources on the record, including the Garlico price data, in light of the numerous infirmities in the Azadpur APMC data that have been identified here. See, e.g., Pl.’s Brief at 18-19 (emphasizing that "Xinboda’s producer, Dadi, does not travel to Beijing to visit the urban farmer’s market to purchase its garlic. Rather, local farmers deliver garlic from surrounding farms. As such, the Indian producer Garlico’s purchase prices most closely approximate ... farm gate prices”); Pl.’s Reply Brief at 3 (contrasting "the wholesale market prices from a large market in New Delhi, India” (i.e., the Azadpur APMC Market) with "the farm gate prices an Indian garlic processor located in one of the largest garlic-growing regions in India [i.e., Garlico] pays for its garlic input”); id. at 6 (arguing that "[the] definition of 'farm gate’ describes an important characteristic of Xinboda’s garlic— and it marks the common characteristic with the garlic that the processor Garlico purchases”).

The Government attempts to go well beyond Commerce's stated reasons for rejecting the Garlico garlic pricing data, arguing that "Commerce also identified other shortcomings and limitations in the Garlico data” that preclude their use to value raw garlic bulbs. See Def.'s Response Brief at 15 & n. 4 (arguing, inter alia, that Garlico "does not produce 'identical’ merchandise”). Again, however, such argument constitutes impermissible post hoc rationalization. See Burlington Truck Lines, 371 U.S. at 168-69, 83 S.Ct. 239; Motor Vehicle Mfrs. Ass'n, 463 U.S. at 50, 103 S.Ct. 2856.

29 . Dongguan succinctly summarizes Commerce’s interim methodology, as modified to reflect the agency’s use of industry-specific data:

First, Commerce creates a list of economically comparable countries based on gross national income. Second, based on this list, Commerce then identifies which countries had exports of comparable merchandise during the period of review. Third, Commerce identifies which of these countries reported wage data during an applicable five-year period. Fourth, Commerce determines which countries reported industry-specific data. Finally, Commerce calculates an average wage rate from those countries found to be economically comparable that have export[s] of comparable merchandise and which reported the appropriate data.

Dongguan Sunrise Furniture Co. v. United States, 36 CIT -, -, n. 29, 865 F.Supp.2d 1216, 1236 n. 29 (2012); see also Interim Labor Methodology, 76 Fed.Reg. at 9546-47.

30 . Commerce's Issues & Decision Memorandum acknowledged the April 2011 decision in Shandong Rongxin, but shrugged it off, stating simply that the agency “ha[d] not yet completed its analysis on remand in that case” and that the court’s decision therefore ''[was] not yet final.” See Issues & Decision Memorandum at 24-25.

31 . In the Issues & Decision Memorandum, Commerce explains that "[t]he ISIC code is maintained by the United Nations Statistical Division and is updated periodically. The ILO [i.e., International Labor Organization], an organization under the auspices of the United Nations, utilizes this classification [i.e., the ISIC code] for reporting purposes.” Issues & Decision Memorandum at 27; see also Shandong Rongxin, 35 CIT at - n. 3, 774 F.Supp.2d. at 1311 n. 3 (describing ISIC as "a uniform, periodically updated system for the classification of economic activity”). The Issues & Decision Memorandum further explains that ”[t]he ISIC code establishes a two-digit breakout for each manufacturing category, and also often provides a three- or four-digit sub-category for each two-digit category.” Issues & Decision Memorandum at 27; see also Shandong Rongxin, 35 CIT at - n. 3, 774 F.Supp.2d at 1311 n. 3 (noting inconsistencies and inaccuracies in Commerce’s use of terminology to describe ISIC taxonomy, and explaining that, officially, two-digit "Divisions” are "further separated into three-digit ‘Groups’ or four-digit 'Classes’ ”).

32 . According to the Issues & Decision Memorandum, at the time of the review in question, labor data were available from the ILO under three revisions: "ISIC-Rev. 2, ISIC-Rev. 3, and ISIC-Rev. 4.” Issues & Decision Memorandum at 27.

33 . According to Commerce, while sub-classification 15 of ISIC Revision 3 covers "Manufacture of Food Products and Beverages,” Revision 2 covers "Manufacture of Food, Beverages and Tobacco.” See Xinboda Preliminary Surrogate Value Memorandum at Exh. 3 (Pub.Doc. No. 102); Issues & Decision Memorandum at 27-28.

34 . See also, e.g., Camau Frozen Seafood Processing Import Export Corp. v. United States, 37 CIT -, -, 929 F.Supp.2d 1352, 1355 (2013) (‘‘Camau II") (emphasizing that "Commerce has the statutory authority to use multiple surrogate countries” in valuing factors of production, including labor); Since Hardware (Guangzhou) Co. v. United States, 37 CIT -, -, 911 F.Supp.2d 1362, 1382 (2013) (summarizing August 14, 2012 order, which rejected claim that Commerce was required to value labor using only data from India, holding that "the statute does not mandate [that] Commerce must, as a matter of law, use Indian data alone”); Dongguan Sun rise Furniture, 36 CIT at -, 865 F.Supp.2d at 1235 (rejecting challenge to use of data from multiple countries in valuing labor, explaining that "the statute permits Commerce to use data from multiple countries”); Home Products Int’l, Inc. v. United States, 36 CIT -, -, 810 F.Supp.2d 1373, 1378 (2012) ("Home Products I”) (rejecting argument that use of Indian data alone was mandated by statute, the Court of Appeals’ decision in Dorbest, and Shandong Rongxin), opinion following remand, 36 CIT -, 837 F.Supp.2d 1294 (2012) ("Home Products II”), aff'd, 501 Fed.Appx. 981 (Fed.Cir.2013).

35 . Compare, e.g., Gold East Paper (Jiangsu) Co. v. United States, 37 CIT -, -, 918 F.Supp.2d 1317, 1333-34 (2013) (declining to require agency to apply Revised Labor Methodology where revised methodology issued more than nine months after final determination in antidumping investigation); Grobest & I-Mei Indus. (Vietnam) Co. v. United States, 36 CIT -, -, 815 F.Supp.2d 1342, 1359-60 (2012) (holding that agency was not required to apply Revised Labor Methodology where revised methodology issued "almost eleven months after the Final Results” in administrative review); Home Products II, 36 CIT at -, 837 F.Supp.2d at 1297 (holding that agency was not required to apply Revised Labor Methodology where revised methodology issued three months after final determination in administrative review); Dorbest Ltd. v. United States, 35 CIT -, - & n. 9, 789 F.Supp.2d 1364, 1369 & n. 9 (2011) ("Dorbest VII”) (declining to require agency to apply Revised Labor Methodology where revised methodology issued roughly six-and-one-half years after January 2005 amended final determination in antidumping investigation, stating that revised methodology is "not retroactive”).

36 . See, e.g., Laizhou Auto Brake Equip. Co. v. United States, 32 CIT 711, 722 (2008) (stating that “[a]t the time [a] new methodology is finalized and effective it becomes the best available information, but until that point, Commerce must be granted some discretion to assess the advantages and disadvantages of applying a work-in-progress methodology in place of an existing one”); SeAH Steel Corp. v. United States, 34 CIT -, - n. 15, 704 F.Supp.2d 1353, 1364 n. 15 (2010) (same).

37 . An agency cannot moot a party’s argument merely by revising its methodology to address the party’s concern and then declaring that the revised methodology will have prospective application only.

38 . See 19 U.S.C. § 1677b(c)(4)(B) (requiring that Commerce "utilize, to the extent possible, the prices or costs of factors of produclion in one or more market economy countries that are ... significant producers of comparable merchandise”).

39 . See generally Downhole Pipe & Equipment LP v. United States, 36 CIT -, -, 887 F.Supp.2d 1311, 1327 (2012) (remanding calculation of surrogate value for labor, where Commerce conceded need to distinguish between "producers” and "significant producers,” in accordance with Shandong Rongxin); cf. Dorbest VII, 35 CIT at -, 789 F.Supp.2d at 1372 (stating that Equatorial Guinea, which had a reported export figure of a mere "USD $308 over the course of three years,” could not be considered "a significant producer” of wooden bedroom furniture, but holding that such claim had been waived).

40 . Xinboda contends that the Indian data reported under ISIC Revision 2 “actually constitute!] the best available information,” because, according to Xinboda, those data were “more contemporaneous than data reported under the later revision.” Pl.’s Brief at 38; see also Xinboda Administrative Case Brief at 82 (Pub.Doc. No. 155) (arguing that, due to Commerce's reliance on ISIC Revision 3 data, the two highest wage rates used in the Final Results were "earnings from 2004 (Ecuador) and 2003 (Philippines),” even though the period of review for the instant review is November 2008 through October 2009). Xinboda further claims that, to the extent that Commerce continues to value labor based on wage data from multiple countries, "several countries reporting under ISIC-Rev. 3 which were used by Commerce should be excluded (or in the case of the Philippines, exchanged for more contemporaneous data ...) because of the resulting over-inflation of the wage rate.” Pl.'s Brief at 38 n. 2 (citing Xinboda Administrative Case Brief at 84).

The Final Results do not directly address the points outlined above. Accordingly, if Commerce does not decide on remand to apply its Revised Labor Methodology in this review (and thus does not rely exclusively on Indian labor data), the remand determination shall address Xinboda’s concerns regarding alleged “over-inflation” as well as the contemporaneity of the data on which Commerce relies.

41 . Because the Indian fiscal year runs from April through the following March, the record includes two sets of financial statements for each company (i.e., statements for April 2008 through March 2009, as well as April 2009 through March 2010), to cover the entire period of review. See Issues & Decision Memorandum at 18 & n. 59. For purposes of the Final Results, Commerce considered only financial statements for 2009-2010. Id. at 20.

The six sets of financial statements on the record include statements for (1) Tata Global, (2) Limtex (India) Limited ("Limtex”), a producer of tea, biscuits, leather products, and information technology services, (3) LT Foods Ltd. ("LT Foods”), a rice processor, (4) ADF Foods Ltd. ("ADF”), a diversified manufacturer of prepared meals, frozen foods, and preserved and pickled items, (5) REI Agro Limited (“REI Agro”), a rice processor, and (6) Garlico Industries Limited (“Garlico”), a purchaser, processor, and trader of garlic, onions, and other vegetables and related products. Issues & Decision Memorandum at 18 & n. 58.

Commerce found that Limtex, LT Foods, ADF, and REI Agro all had benefited from subsidies that the agency had previously found countervailable. Specifically, Limtex’s financial statements indicate that it sold a Duty Entitlement Pass Book ("DEPB”) license; LT Foods’ financial statements indicate that it received Export Promotion Capital Goods Scheme ("EPCGS”) subsidies; and the financial statements of ADF and REI Agro indicate that they received "packing credits.” See Issues & Decision Memorandum at 21-22 & nn. 68, 70, 73; see also Def.’s Response Brief at 21, 23.

42 . The turn of events that Xinboda describes actually occurred in the ninth administrative review, although Xinboda mistakenly refers to the tenth review. See Issues and Decision Memorandum for the Administrative Review of the Antidumping Duty Order on Fresh Garlic from the People's Republic of China (June 13, 2005) (comment 5) ("Ninth Administrative Review Issues & Decision Memorandum”). As in various other reviews, Commerce there "declined to use the financial statements of Tata,” explaining:

[I]n addition to cultivating and manufacturing black tea, Tata is also very heavily engaged in the production of instant tea, packet tea and other value-added forms of bulk tea. While its financials do not specifically break out its sales in terms of bulk, packet or other value-added forms of tea, there are other indications that most of its costs and/or sales reflect the production of packet and other value-added forms of tea. The financial statement notes that eighty-six percent of its consolidated turnover is a result of its branded tea products. Moreover, Tata's energy expenses ..., for example, disproportionately reflect its production of packet tea. The electricity consumed in the production of packet tea is over four times the electricity usage for bulk tea. Similarly, the consumption of furnace oil is nearly seventeen times higher. Furthermore, ... we note that it is our practice to use financial data[,] when available, from a company with a comparable production process rather than data based on production and processing of a product that is more highly processed or preserved prior to sale.

Id. (emphases added) (footnotes omitted); see also, e.g., Antidumping Duty Order on Fresh Garlic from the People's Republic of China: Issues and Decision Memorandum for the Twelfth New Shipper Reviews (Sept. 19, 2008) (comment 3) ("Twelfth New Shipper Reviews Issues & Decision Memorandum”) (rejecting use of Tata Tea financial statements for similar reasons).

43 . See, e.g., Antidumping Duty Order on Fresh Garlic from the People’s Republic of China: Issues and Decision Memorandum for the Eleventh Administrative Review and New Shipper Reviews (June 11, 2007) (comment 3) (choosing Limtex financial statements for 2004-2005 administrative review); Anti-dumping Duty Order on Fresh Garlic from the People’s Republic of China: Issues & Decision Memorandum for the Eleventh New Shipper Reviews (Sept. 20, 2007) (comment 2) (choosing Limtex for 2005-2006 new shipper review); Issues and Decision Memorandum for the Final Results of the Twelfth Administrative Review: Fresh Garlic from the People’s Republic of China ("PRC”) (June 9, 2008) (comment 3) (choosing Limtex for 2005-2006 administrative review); Twelfth New Shipper Reviews Issues & Decision Memorandum (comment 3) (choosing Limtex over Tata Tea, for reasons of comparable production experience, in 2006-2007 new shipper reviews); Issues and Decision Memorandum for the Final Results and Rescission, in Part, of the Antidumping Duty Administrative Review of Fresh Garlic from the People's Republic of China (June 10, 2013) (comment 9) (choosing AgriPure, "a producer and exporter of canned sweet com and fresh vegetables,” for 2010-2011 administrative review); Decision Memorandum for the Preliminary Results of Antidumping Duty New Shipper Review of Fresh Garlic from the People’s Republic of China: Shijiazhuang Goodman Trading Co. Ltd. (Nov. 4, 2013) at 13 (noting agency’s preliminary determination to use AgriPure financial statements for 2011-2012 new shipper review); Decision Memorandum for the Preliminary Results of the 2011-2012 Antidumping Duty Administrative Review: Fresh Garlic from the People’s Republic of China (Dec. 16, 2013) at 19 (reflecting agency's preliminary determination to use AgriPure financial statements for 2011-2012 administrative review).

44 . Xinboda contends that Commerce’s findings concerning Garlico "are based on a serious misreading of Garlico’s financial statement.” Pl.’s Brief at 33.

45 . The administrative determination that Commerce cites in the Final Results does little or nothing to buttress Commerce’s position. See Issues and Decision Memorandum for the Final Results of 2004/2006 Antidumping Duty Administrative and New Shipper Reviews of Certain Frozen Warmwater Shrimp from the People's Republic of China (Sept. 5, 2007) (comment 2). Nothing in that determination supports the notion that Commerce should confine itself solely to information on the face of a company’s financial statements and ignore other record evidence of the company’s receipt of subsidies. Apparently Commerce has grafted onto its subsidization analysis a practice taken from a rather different context. In the referenced Issues & Decision Memorandum, Commerce reasoned that, in examining the financial statements of zero/negative profit surrogate companies used in the calculation of surrogate financial ratios, the agency "must” accept the information in those statements "as is,” noting that Commerce could not simply "adjust the line items of the financial statements of any given surrogate company.” Id. However, the relevance of that rationale in this context — where the issue is whether a company has benefited from an actionable subsidy — is unclear. It is one thing to alter a document; it is a very different matter to review the document in the context of other documents and evidence on the record. Certainly Xinboda here does not propose to "adjust the line items of [Tata Global’s] financial statement ].”

46 . Congress also instructed Commerce "to use, if possible, data based on production of the same general class or kind of merchandise using similar levels of technology and at similar levels of volume as the producers subject to investigation.” H.R.Rep. No. 100-576 at 591, reprinted in 1988 U.S.C.C.A.N. 1547, 1624.

47 . To the extent that Commerce’s practice of requiring evidence of receipt of a countervail-able subsidy on the face of a financial statement is — in essence — a practice that requires evidence of a “reason to believe,” the practice presents a significantly higher bar than "reason to suspect.” See Zhejiang Mach. Import & Export Corp., 31 CIT at 168, 473 F.Supp.2d at 1374 (rejecting notion that “conclusive evidence” is required to satisfy "reason to believe or suspect” standard, and emphasizing that showing needed to meet standard is “clearly less rigorous than required for an actual finding of subsidies in fact”). Commerce has not explained how its stated practice of confining itself to the four corners of a company’s financial statement can be squared with Congress’ expressed intent.

48 .For example, it is at least conceivable that Commerce could establish that — for reasons it has not previously articulated — it has good reason to be highly confident that any coun tervailable subsidies received by a firm will be reflected on the face of that firm's financial statement. But see, e.g,, n. 49, infra (summarizing argument that some types of subsidies “are not normally reported as line items in financial statements”). However, even assuming that Commerce could establish that it has good reason to be highly confident in its practice, it nevertheless is not clear that Commerce could reconcile that practice with the relevant legislative history.

49 . Commerce’s interpretation and application of the "reason to believe or suspect” standard is being contested in a similar context in at least one other pending proceeding. The briefs in that case may be instructive as Commerce revisits its stated practice in the course of the remand here. See, e.g., Supplemental Briefs of all parties (Jan. 6, 2014 & Feb. 28, 2014), filed in Itochu Building Prods. Co. v. United States, Consol. Court No. 12-00065.

In that case, the Itochu plaintiffs contend, inter alia, that Commerce should disregard the financial statement of a particular company due to alleged subsidies, even though there is no indication of the receipt of actionable subsidies on the face of the financial statement. See, e.g., [Itochu] Plaintiffs' Response to the Court’s Request for Supplemental Briefing (Jan. 6, 2014) at 7-9, passim. The Itochu plaintiffs argue that Commerce may not treat "the fact that subsidies are not expressly reported as a line item in a particular financial statement” as "conclusive evidence that a party did not receive subsidies.” Id. at 8. According to the Itochu plaintiffs: "Certain subsidies {e.g., receipt of grants) often are reported as line items in financial statements. Other subsidies {e.g., reductions in costs, such as exemptions from taxes, duties, etc.) are not normally reported as line items in financial statements. Thus, the presence or absence of an express reference to a cost reduction subsidy in a line item of a financial statement does not constitute conclusive evidence that the entity did not receive subsidies, and does not constitute sufficient evidence to support a determination that a financial statement is not distorted by subsidies if there exists substantial evidence on the record that subsidies were received.” Id.

In other words, in that case — as in the instant case — the plaintiff contends that Commerce cannot lawfully confine its review to the face of a financial statement, but, instead, must consider all relevant evidence of record. And it is worth noting that the Government in that case appears to take a position that is diametrically opposed to the position that Commerce and the Government have taken here, to wit: “Commerce does not always rely solely on a financial statement in deciding whether it has a reason to believe or suspect that a surrogate company received countervailable subsidies____Commerce bases its determination on the totality of the circumstances, from information generally available to it at that time.... When Commerce does rely on the financial statement to make its determination, this does not mean Commerce is precluded from reviewing other evidence. If other evidence were to exist that is ... [the ] opposite of or conflicts with a financial statement, Commerce would account for this information in its determination.” See Defendant’s Response to Order for Supplemental Briefing (Jan. 6, 2014) at 10 (emphases added). It is also worth noting that the Government there has now requested a voluntary remand on the surrogate financial statements issue, to permit Commerce to "reexamine the record related to ... the [disputed] financial statement and issue a decision, consistent with its reexamination.” See Defendant's Motion for Voluntary Partial Remand and Consent Motion for Errata (Feb. 26, 2014) at 1.

On remand here, Commerce would be well-advised to undertake a comprehensive review of its interpretation and application of the "reason to believe or suspect” standard, including issues beyond that squarely presented in this case, to ensure the consistency and coherency of its determination on remand and agency practice as a whole. For example, Commerce may wish to consider whether the relevant legislative history can be reconciled with the agency’s position that a financial statement will be disregarded only if the subsidy program in question has been previously determined to be countervailable. Similarly, Commerce may wish to consider whether the legislative history can be reconciled with the agency's position that the prior determination of countervailability must have been made by Commerce itself (and not any other entity). See generally, e.g., [Itochu] Plaintiffs’ Response to the Court’s Request for Supplemental Briefing (Jan. 6, 2014) at 15-16 (and authorities cited there). Further, if Commerce were to determine on remand that it is appropriate for the agency to look beyond the face of a financial statement at least in some circumstances, it would be useful for Commerce to set forth the purpose or purposes for which the agency will undertake such a broader review — for example, to ascertain whether an alleged actionable subsidy was in fact received, to ascertain whether an alleged subsidy was actionable, and/or for other purposes.

Finally, in calculating normal value in a non-market economy proceeding, Commerce applies the "reason to believe or suspect” standard in at least two different (albeit similar) contexts: (1) when determining whether to use a financial statement in calculating surrogate financial ratios (as in this case); and (2) when determining whether to use a particular value for material inputs. See, e.g. Shandong Huarong Mach. Co. v. United States, 31 CIT 30, 37-39, 2007 WL 54069 (2007) (discussing "reason to believe or suspect” standard in context of determining appropriate values for factors of production). On remand, in citing case law and prior administrative determinations concerning the "reason to believe or suspect” standard, Commerce and the parties should be mindful of — and specifically address — whether cases and administrative determinations that arise in a context other than the selection of financial statements are (or are not) relevant here.

50 . Xinboda also contends that Tata Global’s financial statement itself reflects the loans that are documented in the loan agreements that Xinboda placed on the record. Specifically, Xinboda states that Tata Global’s Annual Report ”indicate[s] on its face in Schedule 3, 'Secured Loans,’ that [Tata Global] secured loans from 'Banks ... Secured by way of hypothecation of ... finished products.' ” See Pl.’s Brief at 28 (quoting Tata Global Annual Report).

According to the Government, "Commerce found that a review of the loan agreements listed in Schedule 3 ... did not reflect any loans previously found countervailable.” Def.’s Response Brief at 23 (citing Issues & Decision Memorandum at 20). However, in the Final Results, Commerce actually stated that, "in [the agency’s] analysis of Tata Tea’s 09/10 financial statement, [the agency] did not find evidence ” of the loans at issue. See Issues & Decision Memorandum at 20 (emphasis added). The record is thus ambiguous as to whether in fact Commerce has reviewed the list in Schedule 3. On remand, Commerce shall review that list and shall explain in detail its findings based on that review.

In addition, Xinboda asserts broadly that “the Tata Global Beverages annual report actually ... contain[s] numerous references to many other types of subsidies,” arguing that “[t]he fact that there has not been an actual U.S. countervailing duty investigation on some of these other programs does not reasonably distinguish them based on the ... legislative history” of the "reason to believe or suspect” standard. See Pl.’s Brief at 30. However, as it does in numerous other places in its briefs, Xinboda fails to spell out its argument, and instead directs readers to the administrative case brief that Xinboda filed with the agency. This type of passing allusion and incorporation-by-reference does not suffice to preserve a litigant’s arguments. All such points must be deemed waived.

51 . Xinboda cites the administrative determination in the first administrative review of steel wire hangers from the PRC (noted above) for the proposition that — of the various factors that Commerce considers in evaluating potential surrogates — Commerce's "past practice” has been to "lend[] the greatest weight to similarity of raw material inputs.” Pl.’s Reply Brief at 10; see also Pl.'s Brief at 34-35 (and authorities cited there). That may well be a fair reading of that determination— and, frankly, it is a point that probably warrants greater attention from the parties. Xinboda, however, cites two decisions of the Court of International Trade for the same proposition. See Pl.'s Reply Brief at 10; see also Pl.’s Brief at 34-35 (and authorities cited there). And neither of those cases can possibly be read to support Xinboda’s assertion.

52 . Xinboda criticizes the fact that, in Commerce's analysis of surrogate financial ratios in the Final Results, Garlico is referred to as a "food processor.” See PL’s Brief at 33; Issues & Decision Memorandum at 18 n. 58 (referring to Garlico as "food processor”); id. at 22 (same); see also Def.’s Response Brief at 24 (same). Xinboda argues that this somewhat generic description of Garlico "ignor[es] its very name and the schedules to its financial statements indicating that it processed and consumed large amounts of garlic.” Pl.’s Brief at 33. However, it is undisputed that Garlico's production was not limited to garlic products. See, e.g., id. (noting that, in addition to garlic, Garlico also "processed and consumed large amounts of garlic and comparable vegetables such as onions ”) (emphasis added). Moreover, elsewhere in the Final Results, Commerce describes Garlico as "an Indian garlic producer that purchases raw garlic bulbs for further preparation.” See Issues & Decision Memorandum at 11.

53 . Specifically, Garlico consumed 33,156 quintals of garlic, but traded only 6,331 quintals (roughly one-fifth of its consumption). See Pl.’s Brief at 36.

Further, Xinboda questions Commerce’s assumption that Garlico did not process or repackage the garlic (and the onions and other vegetables) that the company traded. See PL's Reply Brief at 9. Xinboda argues that there is no record evidence to support such an assumption. Id.

54 . According to Xinboda: "[Commerce] does not include the quantity of traded goods in the factory overhead denominator (the higher the denominator, the lower the ratio) but only includes them in the SG & A denominator under the reasonable assumption that SG & A expense (the numerator) must be incurred to sell the traded goods.” PL’s Brief at 36. Xinboda further states that, "[b]ecause there is no machinery involved in processing garlic bulbs for resale, Garlico’s factory overhead ratio is not short of numerator expenses (typically machinery depreciation expense is the largest) that would render it unrepresentative of Xinboda's experience.” Id. The Government has not addressed Xinboda's statements concerning Commerce’s practice.

Xinboda concludes that, in any event, "it was unreasonable for [Commerce] to jettison the only garlic processor of record,” Garlico, merely because the company "made a portion of its revenue from trading — particularly considering the numerous defects in the alternative surrogate [that Commerce] chose.” PL’s Brief at 36.

55 . See also, e.g., PL's Brief at 32-33 (discussing Tata Global’s development, ownership, and marketing of "many branded tea and coffee products”); PL’s Reply Brief at 6 n. 4 (describing Tata Global as "a global giant with a giant corporate apparatus in India alone”); id. at 7-8 (explaining that Tata Global "is much more highly integrated on the sales side of its business” than is Xinboda, with a "vastly different global selling model” and "packaging operations ... far more vast and complex”).

56 . Specifically, Xinboda explains that the notes to Tata Global’s financial statement reflect the cost of "green leaf” produced in Tata Global’s own estates and consumed by the company, and, further, indicate that the cost was — as the notes to Tata Global’s financial statement put it — -"included under various head of expenses” on Tata Global’s books. See Pl.’s Brief at 31; see also Twelfth New Shipper Reviews Issues & Decision Memorandum (comment 3) (rejecting Tata financial statements in light of, inter alia, concerns that they were "unreliable because the cost of green leaf produced and consumed on the company’s own estate was unascertainable which resulted in a substantial understating of the raw materials consumed by the company”). According to Xinboda, "head of expenses” is a term that refers to "various overhead and SG & A numerator categories” for the company’s "financial ratios for factory overhead and SG & A.” Pl.’s Brief at 31. The bottom line, according to Xinboda, is that “the integration and the cost is there; and this note indicates the cost was loaded into numerator expenses that drive up ratios rather than included in the raw materials denominator that drives down surrogate factory overhead and SG & A.” Id. at 31-32.

57 . See also, e.g., Pl.’s Brief at 33-34 (comparing and contrasting installed capacity of Xinboda, Garlico, and Tata Global); id. at 35 (comparing Garlico’s production processes for dehydrated and powdered garlic products with Xinboda’s production process for peeled garlic); id. (explaining that tea "undergoes significantly more processing than whole or peeled garlic — some of which is similar to the very processing that allegedly disqualifies Garlico”); id. (arguing that Tata Global has "much higher packing expenses” than Xinboda, and "is not representative of Xinboda overall”); id. at 36 (noting that processing garlic bulbs for resale requires no machinery); Pl.’s Reply Brief at 8 (summarizing Tata Global’s “packaging operations” as "obviously far more vast and complex than those of Xinboda, which just has to stick a handful of customer labels on some jars or mesh bags of fresh garlic”).

58 . Thus, for example, in the ninth administrative review of the antidumping order on fresh garlic from the PRC, Commerce rejected Tata Global’s financial statement, explaining, inter alia, that "Tata is ... very heavily engaged in the production of instant tea, packet tea and other value-added forms of bulk tea,” and "there are ... indications that most of its costs and/or sales reflect the production of packet and other value-added forms of tea.” Ninth Administrative Review Issues & Decision Memorandum (comment 5). Similarly, in another review, Commerce disregarded Tata Global's financial statement where the evidence indicated, among other things, that "the tea that Tata Tea produces comes in different varieties and value-added types that require higher levels of inputs than bulk tea.” Twelfth New Shipper Reviews Issues & Decision Memorandum (comment 3).

59 . In support of its claim that branded and unbranded products cannot be reliably compared, Xinboda alludes generally to ‘TRS rules” that, according to Xinboda, illustrate and support Xinboda’s claim that branded or trademarked products have a direct impact on prices and profits. See PL’s Brief at 32-33; PL’s Reply Brief at 7-8. However, Xinboda provides no citations to those regulations, referring readers to its administrative case brief filed with the agency. See id. at 8.

60 . Like Xinboda, and unlike Tata Global, Garlico does not develop and market its own branded products. In the underlying administrative review, Xinboda documented Garlico’s modest advertising and selling expenses. See Xinboda Administrative Case Brief at 62. Further, Garlico, like Xinboda and unlike Tata Global, does not sell its own branded garlic. Id. at 65.

61 . Elsewhere, Xinboda states that the "heavily packagedf] heavily branded, [and heavily] marketed” nature of Tata Global's products "directly impacted] SG & A and profit: [Tata Global’s] SG & A ratio was 26.28% whereas Garlico's was merely 9.32%; [Tata Global’s] profit was 17.62%, whereas Garlico’s was merely 2.28%.” See PL’s Brief at 36.

62 . The practice of ''zeroing” is explained in greater detail in Union Steel. See generally Union Steel v. United States, 36 CIT -, -, 823 F.Supp.2d 1346, 1348-50 (2012), aff'd, 713 F.3d 1101, 1103-04 (Fed.Cir.2013); see also JTEKT Corp. v. United States, 642 F.3d 1378, 1383-84 (Fed.Cir.2011); Dongbu Steel Co. v. United States, 635 F.3d 1363, 1365-66 (Fed.Cir.2011).

63 . As noted above, the Preliminary Results in this review were published in December 2010. See Preliminary Results, 75 Fed.Reg. at 80,458.

Xinboda expresses skepticism about Commerce's announced change of practice concerning the use of zeroing in administrative reviews, asserting that "[a]lthough [Commerce] appears to have conceded zeroing in average-to-average transactions in the Final Modification: Reviews [i.e., 77 Fed.Reg. 8101], few respondent practitioners view this as sincere.” See Pl.'s Brief at 6 n. 1. Xinboda predicts that Commerce "will simply find reasons to depart from average-to-average comparisons in more reviews and [will] continue right on zeroing.” Id.

64 . In the course of oral argument, counsel for Xinboda reaffirmed the company’s position that the facts of this case are critically different from those before the Court of Appeals in Union Steel, because the issue here is presented "in the non-market economy context." See Recording of Oral Argument at 8:37-8:57.

65 . Although the Government's brief at one point refers to a remand for the purpose of allowing Commerce to, inter alia, place on the record of this case the explanation that Commerce provided in Union Steel, the Government elsewhere casts the purpose and scope of the proposed voluntary remand rather more broadly. Compare Def.’s Response Brief at 29-30 and id. at 2 (referring to voluntary remand "to provide Commerce an additional opportunity to explain its ‘zeroing’ methodologies in administrative reviews and investigations, and its interpretation of 19 U.S.C. § 1677(35)”); id. at 6 (requesting generally "that Xinboda’s ‘zeroing’ claim be remanded so that Commerce may provide additional explanation”).