Elkay Manufacturing Co. v. United States

OPINION AND ORDER

In this consolidated action, plaintiffs El-kay Manufacturing Company (“Elkay”) and Guangdong Dongyuan Kitchenware Industrial Company, Ltd. (“Dongyuan”) contest a determination (“Final Determination”) that the International Trade Administration, U.S. Department of Commerce (“Commerce” or the “Department”) issued upon the affirmative conclusion of an antidumping duty investigation of drawn stainless steel sinks (“subject merchandise”) from the People’s Republic of China (“China” or the “PRC”). Both plaintiffs challenge aspects of the Department’s calculation of the normal value of subject merchandise. 1

Dongyuan, a Chinese manufacturer and exporter of stainless steel sinks and one of the two producer/exporters that Commerce investigated individually, challenges the Department’s use of import data from Thailand to determine a surrogate value for the cold-rolled stainless steel coil that Dongyuan used as the primary material in producing subject merchandise. Compl. ¶ 11-15 (June 12, 2013), ECF No. 9 (Court No. 13-00199) (“Dongyuan Compl.”). El-kay, a U.S. producer and the petitioner in the investigation, challenges the Department’s method of accounting for selling, general, and administrative (“SG & A”) expenses in the normal value determinations for the two individually-investigated producer/exporters. 2 Compl. ¶ 11-14 *1371 (June 5, 2013), ECF No. 14 (“Elkay Compl.”).

Before the court are Dongyuan’s and Elkay’s motions for judgment on the agency record pursuant to USCIT Rule 56.2. Elkay’s Mot. for J. on the Agency R. (Oct. 21, 2013), ECF No. 26 (“Elkay Mot.”); Dongyuan’s Rule 56.2 Mot. for J. Upon the Agency R. (Oct. 21, 2013), ECF No. 27 (“Dongyuan Mot.”). Also before the court is a request by defendant United States for a partial voluntary remand to allow Commerce to reconsider the use of the Thai import data for determining a surrogate value for cold-rolled stainless steel coil. Def.’s Resp. to Pl.’s Rule 56.2 Mot. for J. on the Agency R. 1 (Feb. 28, 2014), ECF No. 40 (“Def. Opp’n”). The court grants the Department’s partial voluntary remand request and in addition directs Commerce to reconsider its method of accounting for SG & A expenses in the normal value calculations.

J. Background

The decision contested in this case concluded an antidumping duty less-than-fair-value investigation. See Drawn Stainless Steel Sinks From the People’s Republic of China: Investigation, Final Determination, 78 Fed.Reg. 13,019 (Int’l Trade Admin. Feb. 26, 2013) {“Final Determination”), as amended, 78 Fed.Reg. 21,592 (Int’l Trade Admin. Apr. 11, 2013). In response to a petition by Elkay, Commerce initiated the investigation, which examined imports of subject merchandise made during the period of July 1, 2011 through December 31, 2011 (“period of investigation” or “POI”). Drawn Stainless Steel Sinks From the People’s Repub-lie of China: Initiation of Antidumping Duty Investigation, 77 Fed.Reg. 18,207 (Int’l Trade Admin. Mar. 27, 2012). On May 14, 2012, Commerce selected for individual investigation as “mandatory respondents” two Chinese producer/exporters— Dongyuan and a combined entity Commerce identified as consisting of Zhong-shan Superte Kitchenware Co., Ltd. (“Su-perte”) and a related invoicing company, Foshan Zhaoshun Trade Co., Ltd. (“Zhao-shun”) (collectively identified as “Su-perte/Zhaoshun”). Final Determination, 78 Fed.Reg. at 13,019 n. 2.

On October 4, 2012, Commerce issued a preliminary determination (“Preliminary Determination”) that imports of subject merchandise from China are being, or are likely to be, sold in the United States at less than fair value. Drawn Stainless Steel Sinks From the People’s Republic of China: Antidumping Duty Investigation, 77 Fed.Reg. 60,673, 60,673 (Int’l Trade Admin. Oct. 4, 2012) {“Prelim. Determination ”). Commerce calculated preliminary weighted-average dumping margins of 54.25% for Dongyuan, 63.87% for Su-perte/Zhaoshun, and a simple average of the two rates, 59.06%, for “separate-rate” respondents, i.e., those producer/exporters not selected for individual examination that had demonstrated independence from the government of China. Id. at 60,674.

Commerce published its Final Determination on February 26, 2013 and issued an accompanying Issues and Decision Memorandum (“Decision Memorandum”). 3 See Final Determination, 78 Fed.Reg. at 13,-019; Issues & Decision Mem. for the Final Determination in the Antidumping Duty Investigation of Drawn Stainless *1372 Steel Sinks from the People’s Republic of China, A-570-983 (Feb. 19, 2013) (Admin.R.Doc. No. 417) (“Decision Mem.”), available at http://enforcement.trade.gov/ frn/summary/PRC/2013-04379-1.pdf (last visited Dec. 16, 2014). In the Final Determination, Commerce determined that imports of subject merchandise from China are being, or are likely to be, sold in the United States at less than fair value. Final Determination, 78 Fed.Reg. at 13,019. Making certain changes to the analysis it used for the Preliminary Determination, Commerce assigned weighted-average dumping margins of 27.14% to Dongyuan, 39.87% to Superte/Zhaoshun, and a simple average of the two rates, 33.51%, to the separate-rate respondents. Final Determination, 78 Fed.Reg. at 13,020, 13,023.

Elkay initiated an action challenging the Final Determination by filing a summons on May 6, 2013 and a complaint on June 5, 2013. Summons, ECF No. 1; Elkay Compl. ¶ 1. Dongyuan initiated a separate action challenging the Final Determination by filing a summons on May 13, 2013 and a complaint on June 12, 2013. Summons, ECF No. 1 (Ct. No. 13-00199); Dongyuan Compl. ¶ 1. The court consolidated these cases on August 30, 2013. 4 Order, ECF No. 25.

Elkay and Dongyuan filed their motions for judgment on the agency record and accompanying briefs on October 21, 2013. Elkay’s Mot.; Rule 56.2 Br. of Elkay Mfg. Co. in Supp. of its Mot. for J. on the Agency R., ECF No. 29 (“Elkay Br.”); Dongyuan Mot.; Consol. PI. Guangdong Dongyuan Kitchenware Mem. in Supp. of Mot. for J. on the Agency R., ECF No. 31 (“Dongyuan Br.”). On February 28, 2014, defendant filed a consolidated response in opposition to both motions, Def. Opp’n 1, and Dongyuan filed a response opposing Elkay’s 56.2 motion, Def.-Intervenor Gu-angdong Dongyuan Kitchenware Resp. Br. in Opp’n to Pl.’s Rule 56.2 Mem., ECF No. 39 (“Dongyuan Opp’n.”). Elkay and Don-gyuan each filed replies on April 4, 2014. Consol. PI. Guangdong Dongyuan Kitchenware Reply Br. in Supp. of Mot. for J. on the Agency R., ECF No. 44 (“Dongyuan Reply”); Reply Br. of Elkay Mfg. Co. in Supp. of its Mot. for J. on the Agency R., ECF No. 45 (“Elkay Reply”).

II. Discussion

The court exercises jurisdiction under section 201 of the Customs Courts Act of 1980, 28 U.S.C. § 1581(c), pursuant to which the court reviews actions commenced under section 516A of the Tariff Act of 1930, as amended (the “Tariff Act”), 19 U.S.C. § 1516a, including an action contesting a final determination that Commerce issues to conclude an antidumping investigation. 5 In reviewing a final determination, the court “shall hold unlawful any determination, finding, or conclusion found ... to be unsupported by substantial evidence on the record, or otherwise not in accordance with law .... ” 19 U.S.C. § 1516a(b)(l)(B)(i). Substantial evidence is evidence that “a reasonable mind might accept as adequate to support a conclusion.” Consol. Edison Co. v. NLRB, 305 *1373 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938).

Section 773(c)(1) of the Tariff Act, 19 U.S.C. § 1677b(c)(l), provides for the calculation of the normal value of subject merchandise from a nonmarket economy (“NME”) country “on the basis of the value of the factors of production utilized in producing the merchandise,” plus certain additions. 6 The statute directs that “the valuation of the factors of production shall be based on the best available information regarding the values of such factors in a market economy country or countries considered to be appropriate by the administering authority.” Id.

A. The Court Grants Defendant’s Request for a Voluntary Remand so that Commerce May Reconsider the Use of Thai Import Data to Value Cold-Rolled Stainless Steel Coil

During the antidumping investigation, Commerce identified Colombia, Indonesia, Peru, the Philippines, South Africa, Thailand, and Ukraine as countries comparable to the PRC in terms of economic development and found that all of these countries, excluding the Philippines, were significant producers of merchandise comparable to the merchandise under consideration. Decision Mem. for Prelim. Determination for the Antidumping Duty Investigation of Drawn Stainless Steel Sinks from the People’s Republic of China 6 (Sept. 27, 2012) (Admin.R.Doc. No. 337) (“Prelim. Decision Mem.”), available at http:// enforcement.trade.gov/frn/summary/PRC/ 2012-24549-l.pdf (last visited Dec. 16, 2014). Noting that cold-rolled stainless steel coil is the chief raw material used to produce the subject merchandise, Commerce chose Thailand as the primary surrogate country for the Preliminary Determination, finding that “Thailand provides the most specific information to value each respondent’s most significant input (ie., stainless steel).” Id. at 7. Comparing Thai import data published in the Global Trade Atlas (“GTA”) to available GTA import data from the Philippines and Indonesia, which Superte/Zhaoshun proposed for use but which Commerce considered less specific to the input to be valued, Commerce chose Thai import data. Id. at 7, 16-17. Commerce excluded from the GTA Thai import database the imports from various countries that Commerce found to maintain broadly available subsidies. Id. at 17. Commerce also excluded imports from countries on which Thailand imposes anti-dumping duties. Id. In the Final Determination, Commerce 'continued to select Thailand as.the primary surrogate country with which to value the cold-rolled stainless steel coil inputs and continued to make the exclusions from the import database. 7 See Decision Mem. 8-10.

*1374 In contesting the Department’s valuation of the stainless steel coil factor of production (“FOP”), Dongyuan seeks an order directing Commerce “to forego reliance on any Thai import statistics for cold-rolled steel inputs consumed by Dongyuan Kitchenware.” Dongyuan Br. 35. Raising various objections to the suitability and reliability of the Thai import data, Don-gyuan argues that Commerce should have valued the input using GTA import data for the Philippines and for Indonesia, which both investigated respondents placed on the record. Id. at 2-5, 14-34.

In response to Dongyuan’s challenge, defendant requests a voluntary remand so that Commerce may consider whether the Thai import data used in the Final Determination were aberrational. Def. Opp’n 7, 15-16. To address that question, defendant would compare the Thai GTA import data with GTA import data from other potential surrogate countries, namely Colombia, Indonesia, Peru, South Africa, Ukraine, and the Philippines. Id. at 15-16. Defendant explains that Commerce did not place on the record information pertaining to potential surrogate countries other than Indonesia,' the Philippines, and Thailand and that a remand would allow Commerce to “place the missing GTA data on the record.” Id. at 16. Defendant elaborates that “[gjranting a remand will enable the parties to make arguments concerning the omitted evidence[] and will enable Commerce to address those arguments in the first instance.” Id.

Dongyuan opposes defendant’s voluntary remand request, arguing that Commerce did not compile the list of surrogate countries based on average unit values (“AUVs”) of imports in the surrogate countries and that there is no record basis on which to conclude that the other potential surrogate countries have “usable data.” 8 Dongyuan’s Reply 6-9. Don-gyuan submits that the court should not permit Commerce to undergo the “burdensome and unreasonable task,” id. at 9, of recreating a record of world import prices of all surrogate countries listed on the surrogate country list or permit Commerce to “cherry-pick off-the-record data that it has concluded would bolster its Final Determination,” id. According to Dongyuan, allowing Commerce to reopen the record and examine import data from other surrogate countries would allow Commerce to “defend petitioners’ proposed high import AUVs after the close of the factual record and after that factual record is briefed by supplementing the record post-investigation.” Id. at 10. Dongyuan requests that the court instead rule that the existing “administrative record cannot support a finding that fairly traded Thai import statistics are non-aberrant,” id. at 22, and instruct Commerce “to disregard Thai import AUVs in selecting the ‘best available information’ for the surrogate value for Dongyuan Kitchenware’s cold-rolled steel,” id.

The court will grant the Department’s voluntary remand request so that Commerce may reconsider its use of the Thai import data to value the steel coil input and reopen the record to admit additional data. The court rejects Dongyuan’s position that the court, instead of granting the voluntary remand request, should issue a remand order that prohibits Commerce from using the Thai data to value the steel coil input. 9

*1375 Granting defendant’s request for a voluntary remand is appropriate because the Department’s concerns regarding the previous decision to use the Thai import data are substantial and legitimate. See SKF USA Inc. v. United States, 254 F.3d 1022, 1028-29 (Fed.Cir.2001). Because the decision to reopen the record in a remand proceeding ordinarily is a matter for an agency to decide, and. because Commerce has advanced a legitimate reason to reopen the record in this case, the court rejects Dongyuan’s argument that the court should not permit Commerce to admit additional import data to the record. The voluntary remand that defendant requests will permit Commerce to reconsider its previous valuation of the steel coil FOP on the basis of more complete information. In their comments on the remand redeter-mination, the parties to this litigation will have the opportunity to comment on any new evidence submitted to the record and on the decision Commerce reaches upon reconsideration.

B. The Court Orders a Remand on the Department’s Treatment of SG & A Labor Expenses

Section 773(e)(1) of the Tariff Act, 19 U.S.C. § 1677b(c)(l), provides generally that the normal value of subject merchandise produced in a non-market economy country shall be determined “on the basis of the value of the factors of production utilized in producing the merchandise and to which shall be added an amount for general expenses and profit plus the cost of containers, coverings, and other expenses.” Id. (emphasis added). The statute further provides that “the factors of production utilized in producing merchandise include, but are not limited to ... hours of labor required .... ” Id. § 1677b(c)(3)(A) (emphasis added). In this way, the statute provides for the separate valuation of the hours of labor required in producing the subject merchandise and of additional expenses (i.e., “general” and “other” expenses) in the NME normal value calculation. In identifying the “hours of labor required” as a factor of production, the statute does not distinguish between labor expended to produce subject merchandise (i.e., “production” labor) and labor expended in performing non-production activities (i.e., “non-production” labor), such as labor associated with the performance of selling, general, and administrative (“SG & A”) functions.

Although Commerce has discretion in determining the best method for constructing normal value (whether or not in the non-market economy context), it must exercise its discretion consistently with the purpose of the antidumping statute, that is, to “determine margins ‘as accurately as *1376 possible.’ ” Lasko Metal Products, Inc. v. United States, 43 F.3d 1442, 1446 (Fed.Cir.1994) (citing Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir.1990)); Shakeproof Assembly Components, Div. of III. Tool Works, Inc. v. United States, 268 F.3d 1376, 1382 (Fed.Cir.2001) (“Shakeproof”) (“In determining the valuation of the factors of production, the critical question is whether the methodology used by Commerce is based on the best available information and establishes anti-, dumping margins as accurately as possible.”). This principle applies even though the process of constructing normal value “is difficult and necessarily imprecise.” Shakeproof, 268 F.3d at 1381 (citing Nation Ford Chem. Co. v. United States, 166 F.3d 1373, 1377 (Fed.Cir.1999)).

Elkay’s challenge concerns the method Commerce used to account for certain non-production labor — specifically, the labor that the two investigated respondents expended in performing SG & A functions— in calculating the normal value of the subject merchandise of those respondents. Elkay claims that Commerce failed to include in the normal value calculation the total cost of the “hours of labor required,” as required by 19 U.S.C. § 1677b(c)(3)(A), or to capture satisfactorily “general expenses,” as required by 19 U.S.C. § 1677b(e)(l). Elkay Br. 10. Stated simply, Elkay’s claim is that Commerce im-permissibly understated normal value for the two investigated respondents by failing to capture adequately the labor component of the SG & A expenses of those respondents. Id. at 12. Elkay alleges, specifically, that the hours of labor reported by the two investigated respondents, a factor of production to which Commerce applied a surrogate value based on data pertaining to Thailand, included only hours of production labor and, therefore, did not include any hours of labor expended in performing SG & A functions. Id. at 8-9. According to Elkay’s argument, Commerce failed to capture the SG & A labor costs when valuing the hours of labor reported by the investigated respondents as a factor of production and also excluded the SG & A labor costs when it -performed its SG & A expense calculation. Id. at 9.

The Department’s regulations provide that “[f]or manufacturing overhead, general expenses, and profit, the Secretary normally will use non-proprietary information gathered from producers of identical or comparable merchandise in the surrogate country.” 19 C.F.R. § 351.408(c)(4). Commerce usually calculates separate values for factory overhead, SG & A expenses, and profit by calculating and applying financial ratios derived from financial statements of one or more producers of comparable merchandise in the primary surrogate country. See Dorbest Ltd. v. United States, 604 F.3d 1363, 1368 (Fed.Cir.2010) (“In addition to these methods of valuing the direct inputs to production, Commerce also values selling, general, and administrative expenses (“SG & A”), factory overhead, and profit .... using financial ratios derived from publicly-available financial statements of pro-. ducers of comparable merchandise in the surrogate country.”) (citing 19 C.F.R. § 351.408(c)(4)). According to the Department’s Antidumping Manual, the Department’s practice is to calculate a ratio for “general expenses” and to apply this ratio to the sum of the amount calculated for a respondent’s “total of materials, labor, energy, and factory overhead.” 10 *1377 Import Admin. Antidumping Manual, Chapter 10: Nan-Market Economies 18 (2009 ed.), available at http://enforeement. trade.gov/admanual/2009/Chapter% 2010% 20NME.pdf (last visited Dee. 16, 2014).

In the Preliminary Determination of this investigation, Commerce determined surrogate ratios for SG & A expenses, combined with interest expenses, for each of three companies .in the chosen surrogate country, Thailand, using data on those expenses reported in the financial statements of the three companies. Factor Valuation Mem. for Prelim. Determination 7-8, Attach. 5 (Sept. 27, 2012) (Admin.R.Doc. No. 342-348) (“Prelim. Factor Valuation Mem.”). The numerators of these SG & A/interest expense ratios calculated for the Preliminary Determination contained the SG & A and interest expenses listed in the surrogate financial statements. Id. The denominators of these ratios contained expenses listed in the surrogate financial statements for raw material, labor, energy, manufacturing overhead, and the cost of purchased goods and inventory. Id. From the three separate ratios, Commerce calculated an average ratio for use in determining surrogate SG & A/interest expenses for the two investigated respondents, Su-perte/Zhaoshun and Dongyuan. Id.

In the Final Determination, Commerce used the same three financial statements to calculate the three SG & A/interest expense ratios but applied a different method. This method resulted in an average SG & A/interest ratio that was substantially lower than the average ratio calculated in the Preliminary Determination. Decision Mem. 15. The financial statements of all three Thai companies reported production labor costs separately from certain other labor costs, which were itemized and categorized among the sales or administrative expenses in the surrogate financial statements. See Pet’r’s Submission of Surrogate Values Ex. 10 (Aug. 13, 2012) (Admin.R.Doc. No. 262) (“Elkay’s Surrogate Values Submission ”) (including financial statements from Stainless Steel Home Equipment Manufacturing Co., Ltd. and Diamond Brand Co., Ltd.); Don-gyuan’s Rebuttal Surrogate Values for the Prelim. Results Exs. 4-6 (Aug. 20, 2012) (Admin.R.Doc. No. 296) (including financial statements from Advance Stainless Steel Co., Ltd.). The record indicates that Commerce excluded a number of the labor costs identified in the financial statements as sales or administrative expenses (and described by Commerce as “SG & A labor costs”) from the numerators of the three SG & A/interest expense ratios and included these costs in the denominators of those ratios. 11 See Factor Valuations for the *1378 Final Determination, Attach. 1 (final surrogate value worksheets) (Feb. 19, 2013) (Admin.R.Doc. No. 422) (“Final Factor Valuations Mem.”). As defendant explains in its brief, reclassifying the line items in the surrogate financial statements that Commerce determined to correspond to SG & A labor “reduces the surrogate SG & A ratio by decreasing the total SG & A expenses in the numerator and increasing the labor expense in the denominator.” 12 Def. Opp’n 9. Reclassifying these SG & A labor costs reduced the 12.82% average SG & A/interest expense ratio of the Preliminary Determination to a 6.36% average ratio in the Final Determination, a reduction of more than half. Compare Prelim. Factor Valuation Mem. 8 with Final Factor Valuations Mem. 2.

The rationale Commerce provided in the Decision Memorandum for the decision to change the SG & A/interest expense ratio calculation relied in part on the surrogate labor rate with which Commerce valued the labor hours reported by the two investigated respondents. Decision Mem. 15. For the Preliminary Determination, Commerce valued the two respondents’ reported labor hours using a surrogate labor rate derived from 2005 data from Chapter 6A of the International Labor Organization (“ILO”) Yearbook of Labor Statistics. Prelim. Factor Valuation Mem. 5-6. The record indicates that the ILO labor rate Commerce applied in the Preliminary Determination was 141.2162 Thai Bhat (“THB”) per hour. Id. at Attach. 4. c

Following the Preliminary Determination, Dongyuan submitted for admission to the record certain data on the labor cost of “[m]anufacture of other fabricated metal products not enumerated elsewhere” contained in the “Industrial Census 2007” published by Thailand’s National Statistics Office (“NSO”). Dongyuan’s Final Surrogate Value Submission — Part I, Ex. SV-1 (Nov. 26, 2012) (AdmimR.Doc. No. 373) (“Dongyuan’s Final Surrogate Value Submission ”). For the Final Determination, Commerce applied the NSO labor rate as a surrogate labor rate to value the examined respondents’ reported labor hours. Decision Mem. 11. Commerce concluded that the NSO data were superior to the ILO data, finding “that the 2007 NSO data for labor cost of ‘manufacturing of other fabricated metal products’ (ISIC Rev.3 Code: 2899) is the best information available on the record to calculate the labor cost of the two respondents in the final determination.” Id. (citation omitted in original). Commerce explained that “[t]his is because the 2007 NSO data are the most product-specific and contemporaneous[ ] and provide a broader market average among all the data parties placed on the record.” Id. According to the Decision Memorandum, the hourly rate obtained from the NSO data was 54.61 THB per hour. Id. at 12. It appears from the record that the rate as adjusted for inflation and applied was 63.1158 THB per hour. See Final Factor Valuations Mem. 3 (“The SV [surrogate value] for labor is 63.12 THB/hr.”) & Attach. 1 (showing an applied rate of 63.1158); see also Dongyuan’s Final Surrogate Value Submission, Ex. SV-1. The surrogate labor rate applied in the Final *1379 Determination, therefore, was less than half of the 141.2162 THB/hour rate applied in the Preliminary Determination. 13

According to the Decision Memorandum, Dongyuan argued during the investigation that “[t]he Department should treat selling, general and administrative (‘SG & A’) labor line items (sales, administrative and managerial staff and directors’ salaries) in the surrogate financial statements as manufacturing labor, not SG & A labor, in its financial ratio calculations because the underlying ILO Chapter 6A data and the NSO Industrial Census data include these expenses.” Decision Mem. 14. In the Final Determination, Commerce agreed with Dongyuan’s proposed change, stating that “[a]fter examining the record, we agree with Dongyuan that the NSO data include total labor costs (i.e., manufacturing and SG & A)-” Id. at 15.

Commerce explained that it intended to account for SG & A labor by applying to the hours of labor reported by each examined respondent a surrogate labor rate that included both manufacturing and SG & A labor costs. Id. Commerce also stated that “[u]sing a surrogate financial ratio that includes SG & A labor costs in addition to the NSO-based surrogate labor rate would double-count those costs in normal value because both include an amount for SG & A labor.” Id. Commerce concluded its discussion by stating that “to avoid double-counting SG & A labor, we have determined to exclude labor costs from the SG & A surrogate financial ratios for the final determination.” Id.

Elkay argues that Commerce should be required to calculate the SG & A/interest expense ratios as Commerce did in the Preliminary Determination — by placing in the numerator of the SG & A/interest expense ratios all SG & A expenses reported in the surrogate companies’ financial statements, including SG & A labor costs. El-kay Br. 12. Elkay submits that although “the NSO labor rate was an average based on compensation for all types of labor, including both production and non-production labor,” id. at 8-9, the “labor rate was only applied to production labor hours,” id. at 9, and not to the non-production labor expended in performing SG & A functions. Elkay asserts that because Commerce “applied the revised surrogate labor rate to a labor FOP that includes only production hours, Commerce failed to capture any SG & A labor costs in the calculation of normal value.” Id. at 6.

There can be no dispute that Commerce intentionally used an average ratio for SG & A and interest expenses that captured less than the full amount of the SG & A expenses as reported in the surrogate financial statements. It also appears from the record that the labor hours reported as a factor of production by both investigated respondents included only hours of production labor. 14 But it does not necessarily follow, as Elkay argues, *1380 that “Commerce failed to capture any SG & A labor costs in the calculation of normal value.” Elkay Br. 6. Elkay’s characterization assumes that Commerce omitted SG & A labor costs' entirely from the normal value determination considered as a whole. Because the surrogate labor rate Commerce applied to the quantities of labor hours reported by the two investigated respondents bears some relationship to non-production labor costs, Elkay has not demonstrated that the calculation Commerce used to value the examined respondent’s labor hours failed to capture any SG & A labor costs. Nevertheless, as discussed below, the court cannot sustain the principal conclusion underlying the Department’s decision.

1. The Principal Conclusion Underlying the Department’s Decision to Adjust the SG & A/Interest Expense Ratios is Not Supported by Substantial Record Evidence

In recounting the Department’s reasons for making the adjustments, the Decision Memorandum states a finding that the NSO data include “total labor costs (ie., manufacturing and SG & A) ....” 15 Decision Mem. 15. From this finding, Commerce reached the conclusion that “[u]sing a surrogate financial ratio that includes SG & A labor costs in addition to the NSO-based surrogate labor rate would double-count those costs- in normal value because both include an amount for SG & A labor.” Id. Referring implicitly to the specific adjustment that it made to the surrogate financial ratios, Commerce further concluded that this adjustment was necessary in order “to avoid double-counting SG & A labor.” Id. As the court discusses below, the record lacks substantial evidence to support the Department’s conclusion that “double-counting” of SG & A labor expenses required the specific downward adjustments to the SG & A/interest expense ratios that Commerce found necessary when calculating the normal value for the merchandise of Dongyuan and Su-perte/Zhaoshun. 16 The court, therefore, *1381 must hold invalid the principal conclusion underlying the decision to make the adjustments to the SG & A/interest expense ratios.

Record evidence indicates that the NSO labor rate is derived from data on “remuneration” paid to “[pjersons engaged” in the “manufacture of other fabricated metal products.” Dongyuan’s Final Surrogate Value Submission, Ex. SV-2 (containing the NSO’s description of its methodology and definitions). The NSO defined “[pjer-sons engaged” to include: (1) “[ujnpaid workers” who are “owners or business partners who managed or participated in the management of the establishment;” (2) “[ojperatives” that “were directly engaged in the production or other related activities of the establishment and received pay;” and (3) “other employees” referring to “all employees other than operatives,” including “administrative, technical and clerical personnel such as salaried managers and directors, laboratory and research workers, clerks, typists, book-keepers, administrative supervisors, salesman and the like.” Id. The NSO’s definitions specify that excluded from the.“persons engaged” are: (1) “managers or directors paid solely for their attendance at meeting[s] of the board of director[s];” (2) [pjersons from other establishments] working at this establishment; (3) “[hjome workers;” (4) “[pjersons who were on leave for military services or one who had obtained long leave or were on strike”; and (5) “[pjersons who were employed to work occasionally such as laborer[s] and sale agents who do not receive regular pay.” Id.

The NSO information supports a finding that the NSO rate was derived from an average remuneration paid for “persons engaged” in various production-related and non-production-related activities. It also supports a finding that the NSO rate is a much broader average than one representing only wages and salaries. The NSO definition of “[rjemuneration” includes: (1) “[wjages/salaries;” (2) “[ojvertime, bonus, special payment, cost of living allowance and commission;” (3) “fringe benefits;” and (4) “[ejmployer’s contribution to social security” (including as examples, payments to a “social security fund, workmen’s compensation fund[,j and health insurance, *1382 etc.”)- Id. Based on the record evidence, Commerce could conclude that the NSO labor rate is higher than it would have been had it been derived solely from data on wages and salaries.

Less clear is that the NSO labor rate is higher than it would have been had it been derived solely from data on production workers. It may be reasonable to infer that some non-production employees, e.g., high-level salaried managerial employees, receive higher remuneration than persons engaged in production. The record data, however, do not support an actual finding that the NSO labor rate was higher — or by what percentage it was higher — than it would have been had it been derived solely from Thai data on production labor rather than from a combination of Thai data on production labor and various types of non-production labor. Apparently, missing from the record are the raw data from which the NSO rate was derived, which possibly could support such conclusions.

The court has considered the record evidence, summarized above, on the derivation of the NSO labor rate and on the derivation, and the magnitude, of the downward adjustments Commerce made to the three SG & A/interest expense ratios (and, accordingly, to the average SG & A/interest expense ratio applied in determining the normal value of the two investigated respondents). On this administrative record, the Department’s reliance on the extent of any “double-counting,” Decision Mem. 15, was too much a matter of speculation. The record lacks substantial evidence to support the Department’s conclusion that the rate Commerce applied to the hours of production labor reported by the investigated respondents overstated the value of those labor hours to such an extent as to justify the specific, compensatory adjustments that Commerce made to the SG & A/interest expense ratios.

2. The Department’s Reliance on the Labor Methodologies Notice Does Not Justify the Downward Adjustments to the SG & A/Interest Expense Ratios

In both the Preliminary Determination and in the Decision Memorandum, Commerce based its treatment of SG & A labor expenses in part on a notice (“Labor Methodologies ”) Commerce published in 2011 announcing a new methodology for valuing labor in non-market economy (NME) proceedings. 17 See Antidumping Methodologies in Proceedings Involving Non-Market Economies: Valuing the Factor of Production: Labor, 76 Fed.Reg. 36,092 (Int’l Trade Admin. June 21, 2011) (“Labor Methodologies ”). The Department’s reasoning, and its conclusion, changed considerably from the Preliminary to the Final Determination. 18

*1383 The Labor Methodologies notice announces that “the Department will value the NME respondent’s labor input using industry-specific labor costs prevailing in the primary surrogate country, as reported in Chapter 6A of the ILO Yearbook of Labor Statistics.” 19 Id. at 36,094. The notice creates “the rebuttable presumption that Chapter 6A data better accounts for all direct and indirect labor costs.” Id. at 36,093. The notice further explains that Commerce would discontinue its current use of the data in ILO Chapter 5B: “Unlike Chapter 6A data that reflects all costs related to labor including wages, benefits, housing, training, etc., Chapter 5B data reflects only direct compensation and bonuses.” Id.

The Decision Memorandum cited the Labor Methodologies notice in support of the position that “[i]n deriving surrogate financial ratios, ‘it is the Department’s longstanding practice to avoid double-counting costs where the requisite data are available to do so.’” Decision Mem. 15 (citation omitted) (emphasis added to original in quoted source). Commerce noted that “in Labor Methodologies, we said that ‘the Department will adjust the surrogate financial ratios when the available record information — in the form of itemized indirect labor costs — demonstrates that labor costs are overstated.’ ” 20 Id. (citing Labor Methodologies, 76 Fed.Reg. at 36,093-94).

For two reasons, the Labor Methodologies notice is not an adequate justification for the decision challenged in this case. First, Commerce departed from the methodology announced in the notice by rejecting the ILO Chapter 6A data in the Final Determination in favor of the NSO data, *1384 Decision Mem. 13-14, resulting in a rate that was lower by more than half than one derived from ILO Chapter 6A. 21 Compare Prelim. Factor Valuation Mem., Attach. 4 (applying a surrogate rate of 141.2162 THB/hr), with Final Factor Valuations Mem. 3 (applying a surrogate rate of 63.12 THB/hr). Second, the Labor Methodologies notice explains that “[i]f there is evidence submitted on the record by interested parties demonstrating that the NME respondent’s cost of labor is overstated, the Department will make the appropriate adjustments to the surrogate financial statements subject to the available information on the record.” Labor Methodologies, 76 Fed.Reg. at 36,094. As the court discussed previously, the record in this case does not contain substantial evidence to support the Department’s conclusion that when calculating the SG & A/interest expense ratios Commerce made appropriate adjustments to the surrogate financial statements to compensate for an overstated labor cost.

S. The Court Rejects Defendant’s and Dongyuan’s Arguments in Support of the Department’s Decision on SG & A Labor Expenses

Before the court, defendant and Don-gyuan make several arguments in favor of the Department’s decision to adjust the SG & A/interest expense ratios. The court rejects these arguments for the reasons discussed below.

Defendant states that “[ujnfortunately, the respondents did not report SG & A labor hours” and that “[bjased on the limitations of the data on the record, Commerce was faced with two imperfect methodologies for calculating labor.” Def. Opp’n 6. Defendant argues that “[i]n such circumstances, Commerce has broad discretion to choose between the competing methodologies,” id. at 7, and that the court must defer to the Department’s reasonable choice for labor and SG & A expense calculations, so long as that choice is accompanied by a reasoned explanation, id. at 15. See also Dongyuan Opp’n 9 (“Even if there is another way to capture this cost, the Court must uphold this methodology if the Court finds it to be reasonable-”). Regarding the choice Commerce made, defendant explains that “[ajlthough double counting was not a certainty under the record in this case, Commerce’s actions recognized the possibility of double-counting,” Def. Opp’n 14, and that “[a]s a result, Commerce reasonably excluded ‘labor costs from the SG & A surrogate financial ratios for the final determination,’ ” id. (quoting Decision Mem. 15).

Defendant’s argument is unconvincing because it is based on a mischaracterization of the principal conclusion as stated in the Decision Memorandum: Commerce concluded that double-counting would occur absent the adjustments it made to the ratios, and this was the basis upon which Commerce determined that those adjustments were appropriate to avoid the double-counting it found to exist. See Decision Mem. 15 (“Using a surrogate financial ratio that includes SG & A labor costs in addition to the NSO-based surrogate labor rate would double-count those costs in normal value because both include an amount *1385 for SG & A labor.” (emphasis added)). And even were the Department’s decision presumed to rest upon a “reasoned explanation,” Def. Opp’n 15, that explanation cannot overcome the Department’s basing its decision on a conclusion that, for the reasons discussed, is not supported by substantial record evidence.

Further to its argument that Commerce made a reasonable choice given two imperfect alternatives, defendant argues that “Elkay has not established that its preferred methodology — retaining the SG & A labor costs in the SG & A financial ratio calculation ... while using NSO data that indisputably includes SG & A labor— would have provided a more accurate surrogate value for labor.” Def. Opp’n 14 (citation omitted). Defendant argues further that Elkay “wholly ignores the facial risk of double-counting that would be created by its proposed approach.” Id.

Defendant’s argument fails because the court must evaluate the Department’s decision according to the applicable standard of review and cannot sustained a decision not grounded in substantial record evidence. The argument is also misguided in positing that Commerce had only the two choices defendant describes. Commerce had other choices, including choices that did not involve use of the NSO data or choices not involving the particular adjustments it made to the SG & A/interest expense ratios.

In support of the Final Determination, Dongyuan argues that the Department’s methodology captured SG & A labor expenses because the NSO rate “applied to the production labor, is artificially higher than the labor rate would be because it includes more labor expenses than production labor would normally cover.” Don-gyuan Opp’n 8-9. As the court discussed previously, the record shows that the NSO rate was derived from an average remuneration paid for “persons engaged” in various production-related and non-production-related activities and that it is a broader representation of labor cost than one limited to wages and salaries. But even if the NSO rate were presumed to be higher than it would be if it had not included non-production labor (a presumption for which the record lacks substantial evidence), the record would not support the Department’s conclusion that the adjustments the Department made to the SG & A/interest expense ratios were appropriate adjustments for the double-counting of SG & A labor that the Department found would occur absent those adjustments.

k. On Remand, Commerce Must Reconsider its Decision to Adjust the SG & A/Interest Expense Ratios

On remand, Commerce must reconsider its decision to adjust the SG & A/interest expense ratios and the particular way in which it accomplished those adjustments. Additionally, because Commerce grounded its decision in part on its choice of a surrogate labor cost, it may consider on remand alternative data sources with which to value the labor hours reported by the two investigated respondents.

III. Conclusion and Order

For the reasons discussed in the foregoing, the court remands the final determination (“Final Determination”) of the International Trade Administration, U.S. Department of Commerce (“Commerce” or the “Department”) in Drawn Stainless Steel Sinks From the People’s Republic of China: Investigation, Final Determination, 78 Fed.Reg. 13,019 (Int’l Trade Admin. Feb. 26, 2013) (“Final Determination”), as amended, 78 Fed.Reg. 21,592 (Int’l Trade Admin. Apr. 11, 2013). Therefore, upon consideration of all papers and proceedings in this case, and upon due deliberation, it is hereby

*1386 ORDERED that the Final Determination be, and hereby is, set aside as unlawful and remanded for reconsideration and redetermination in accordance with this Opinion and Order; it is further

ORDERED that Commerce shall issue, within ninety (90) days of the date of this Opinion and Order, a new determination upon remand (“Remand Redetermination”) that conforms to this Opinion and Order and reconsiders the use of surrogate information from Thailand to value cold-rolled stainless steel coil when determining the normal value of Dongyuan’s subject merchandise; it is further

ORDERED that if Commerce places additional surrogate value data on the record on remand, Commerce must provide parties to this litigation the opportunity to submit comments concerning those data and the Department’s decision addressing the valuation of cold-rolled stainless steel coil; it is further

ORDERED that in the Remand Rede-termination, Commerce shall reconsider its method of accounting for selling, general, and administrative (“SG & A”) labor costs in its calculation of normal value and, as necessary, revise the antidumping duty margins for both the investigated and separate rate respondents; it is further

ORDERED that Elkay Manufacturing Company and Guangdong Dongyuan Kitchenware Industrial Co., Ltd. each may file comments on the Remand Redetermi-nation within thirty (30) days from the date on which the Remand Redetermination is filed with the court; and it is further

ORDERED that defendant may file a response within fifteen (15) days from the date on which the last of any such comments is filed with the court.