A-552-817Administrative Review

POR: 09/01/2017 – 08/31/2018Public Document

E&C/OVI: FB

July 2, 2020

MEMORANDUM TO: Jeffrey I. KesslerAssistant Secretaryfor Enforcement and Compliance

FROM: James MaederDeputy Assistant Secretaryfor Antidumping and Countervailing Duty Operations

SUBJECT: Issues and Decision Memorandum for the Final Results of the2017-2018 Administrative Review of the Antidumping Duty Orderon Oil Country Tubular Goods from the Socialist Republic ofVietnam

______________________________________________________________________________

I. SUMMARY

We have analyzed the case and rebuttal briefs of interested parties in the 2017-2018administrative review of the antidumping (AD) duty order on certain oil country tubular goods(OCTG) from the Socialist Republic of Vietnam (Vietnam). As a result of our analysis, we findthat sales have not been made at less-than-normal value during the period of review (POR) byrespondent SeAH Steel VINA Corporation (SSV) and its U.S. affiliate Pusan Pipe America, Inc.We recommend that you approve the positions described in the “Discussion of the Issues”section of this memorandum. Below is the complete list of issues for which we receivedcomments and rebuttal comments from interested parties:

Comment 1: Brokerage and HandlingComment 2: Surrogate Value for WaterComment 3: Differential PricingComment 4: Financial StatementsComment 5: Particular Market SituationComment 6: Ministerial Errors

2

II. BACKGROUND

On November 15, 2019, the Department of Commerce (Commerce) published the PreliminaryResults of this administrative review in the Federal Register.1 On December 16, 2019,Commerce received a case brief from SSV,2 from Maverick Tube Corporation and Tenaris BayCity, Inc. (domestic interested parties),3 and from U.S. Steel Corporation (the petitioner).4However, on April 10, 2020, we rejected the case briefs from the petitioner and the domesticinterested parties because they contained new factual information filed after the due date for newfactual information.5 The petitioner and the domestic interested parties submitted redactedversions of their case briefs on April 14, 2020.6 On December 30, 2019, Commerce receivedrebuttal briefs from SSV7 and the domestic interested parties.8 However, on April 10, 2020,Commerce rejected SSV’s rebuttal brief because it contained new factual information filed afterthe due date for new factual information.9 SSV submitted a redacted version of its rebuttal briefon April 14, 2020.10

Commerce exercised its discretion to toll all deadlines affected by the partial federal governmentclosure from December 22, 2018, through the resumption of operations on January 29, 2019.11

On March 12, 2020, Commerce extended the final results of this administrative review until May

1 See Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Preliminary Results ofAntidumping Duty Administrative Review, 84 FR 62504 (November 15, 2019) (Preliminary Results), and accompanying Preliminary Decision Memorandum (PDM).2 See SSV Case Brief, “Administrative Review of the Antidumping Duty Order on Certain Oil Country TubularGoods from Vietnam – Case Brief of SeAH Steel VINA Corporation and Pusan Pipe America, Inc.” datedDecember 16, 2019 (SSV Case Brief).3 See Domestic Interested Parties’ Case Brief, Oil Country Tubular Goods from the Socialist Republic of Vietnam: Case Brief of Maverick Tube Corporation and Tenaris Bay City, Inc.,” dated December 16, 2019.4 See Petitioner’s Case Brief, “Oil Country Tubular Goods from the Socialist Republic of Vietnam: Case Brief ofUnited States Steel Corporation,” dated December 16, 2019.5 See Commerce Letter to the Petitioner, “Certain Oil Country Tubular Goods from the Socialist Republic ofVietnam: Rejection of Case Brief,” dated April 10, 2020, and Commerce Letter to Domestic Interested Parties,“Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Rejection of Case Brief,” dated April10, 2020.6 See Petitioner’s Case Brief, “Oil Country Tubular Goods from the Socialist Republic of Vietnam: Resubmissionof December 16th Case Brief of United States Steel Corporation,” dated April 14, 2020 (Petitioner’s Case Brief);Domestic Interested Party’s Case Brief, “Oil Country Tubular Goods from the Socialist Republic of Vietnam: Resubmission of Case Brief of Maverick Tube Corporation and Tenaris Bay City, Inc., dated April 14, 2020(Domestic Interested Parties’ Case Brief).7 See SSV Rebuttal Brief, “Administrative Review of the Antidumping Duty Order on Certain Oil Country TubularGoods from Vietnam – Rebuttal Brief of SeAH Steel VINA Corporation and Pusan Pipe America, Inc.,” datedDecember 30, 2019.8 See Domestic Interested Parties’ Rebuttal Brief, “Oil Country Tubular Goods from the Socialist Republic ofVietnam: Rebuttal Brief of Maverick Tube Corporation and Tenaris Bay City, Inc.,” dated December 30, 2019(Domestic Interested Parties’ Rebuttal Brief).9 See Commerce Letter, “Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Rejection ofRebuttal Brief,” dated April 10, 2020.10 See SSV Rebuttal Brief, “Administrative Review of the Antidumping Order on Certain Oil Country TubularGoods from Vietnam — Redacted Case {sic} Brief,” dated April 14, 2020 (SSV Rebuttal Brief).11 See Memorandum to the Record from Gary Taverman, Deputy Assistant Secretary for Antidumping andCountervailing Duty Operations, performing the non-exclusive functions and duties of the Assistant Secretary forEnforcement and Compliance, “Deadlines Affected by the Partial Shutdown of the Federal Government,” dated

3

13, 2020.12 On April 24, 2020, Commerce tolled all deadlines in administrative reviews by 50days, thereby extending the deadline for these results until July 2, 2020.13

III. SCOPE OF THE ORDER14

The merchandise covered by the order is certain oil country tubular goods, which are hollowsteel products of circular cross-section, including oil well casing and tubing, of iron (other thancast iron) or steel (both carbon and alloy), whether seamless or welded, regardless of end finish(e.g., whether or not plain end, threaded, or threaded and coupled) whether or not conforming toAmerican Petroleum Institute (API) or non-API specifications, whether finished (includinglimited service OCTG products) or unfinished (including green tubes and limited service OCTGproducts), whether or not thread protectors are attached. The scope of the order also coversOCTG coupling stock.

Excluded from the scope of the order are: casing or tubing containing 10.5 percent or more byweight of chromium; drill pipe; unattached couplings; and unattached thread protectors.The merchandise subject to the order is currently classified in the Harmonized Tariff Schedule ofthe United States (“HTSUS”) under item numbers: 7304.29.10.10, 7304.29.10.20,7304.29.10.30, 7304.29.10.40, 7304.29.10.50, 7304.29.10.60, 7304.29.10.80, 7304.29.20.10,7304.29.20.20, 7304.29.20.30, 7304.29.20.40, 7304.29.20.50, 7304.29.20.60, 7304.29.20.80,7304.29.31.10, 7304.29.31.20, 7304.29.31.30, 7304.29.31.40, 7304.29.31.50, 7304.29.31.60,7304.29.31.80, 7304.29.41.10, 7304.29.41.20, 7304.29.41.30, 7304.29.41.40, 7304.29.41.50,7304.29.41.60, 7304.29.41.80, 7304.29.50.15, 7304.29.50.30, 7304.29.50.45, 7304.29.50.60,7304.29.50.75, 7304.29.61.15, 7304.29.61.30, 7304.29.61.45, 7304.29.61.60, 7304.29.61.75,7305.20.20.00, 7305.20.40.00, 7305.20.60.00, 7305.20.80.00, 7306.29.10.30, 7306.29.10.90,7306.29.20.00, 7306.29.31.00, 7306.29.41.00, 7306.29.60.10, 7306.29.60.50, 7306.29.81.10, and7306.29.81.50.

The merchandise subject to the order may also enter under the following HTSUS item numbers:7304.39.00.24, 7304.39.00.28, 7304.39.00.32, 7304.39.00.36, 7304.39.00.40, 7304.39.00.44,7304.39.00.48, 7304.39.00.52, 7304.39.00.56, 7304.39.00.62, 7304.39.00.68, 7304.39.00.72,7304.39.00.76, 7304.39.00.80, 7304.59.60.00, 7304.59.80.15, 7304.59.80.20, 7304.59.80.25,7304.59.80.30, 7304.59.80.35, 7304.59.80.40, 7304.59.80.45, 7304.59.80.50, 7304.59.80.55,7304.59.80.60, 7304.59.80.65, 7304.59.80.70, 7304.59.80.80, 7305.31.40.00, 7305.31.60.90,7306.30.50.55, 7306.30.50.90, 7306.50.50.50, and 7306.50.50.70.

The HTSUS subheadings above are provided for convenience and customs purposes only. Thewritten description of the scope of the order is dispositive.

January 28, 2019. All deadlines in this segment of the proceeding have been extended by 40 days.12 See Memorandum, “Oil Country Tubular Goods from the Socialist Republic of Vietnam: Extension of Deadlinefor Final Results of Antidumping Duty Administrative Review,” dated March 12, 2020.13 See Memorandum, “Tolling of Deadlines for Antidumping and Countervailing Duty Administrative Reviews inResponse to Operational Adjustments Due to COVID-19,” dated April 24, 2020.14 See Certain Oil Country Tubular Goods from India, the Republic of Korea, Taiwan, the Republic of Turkey, andthe Socialist Republic of Vietnam: Antidumping Duty Orders; and Certain Oil Country Tubular Goods from theSocialist Republic of Vietnam: Amended Final Determination of Sales at Less Than Fair Value, 79 FR 53691(September 10, 2014).

4

IV. MARGIN CALCULATIONS

In these final results of review, based on our analysis of comments received, wemade the following changes to the Preliminary Results:

We revised our allocation methodology for calculating brokerage and handlingcosts. See Comment 1.

We removed the “document preparation” category of brokerage and handling costs.See Comment 1.

We did not use the financial statement of Spark Electrodes Private Limited in thecalculation of the surrogate financial ratios. See Comment 4.

We valued hot-rolled coil, varnish, and toluene using Global Trade Atlas importdata for India, rather than market-economy purchase prices. See Comment 5.

We corrected a ministerial error with respect to the cost of freight for twopacking inputs. See Comment 6.

V. DISCUSSION OF THE ISSUES

Comment 1: Brokerage and Handling

In the Preliminary Results, Commerce used Doing Business 2019: India (Doing Business) tocalculate the surrogate value for Indian brokerage and handling (B&H) costs.15 The total B&Hcosts consisted of the summation of costs for “documentary compliance,” “clearance andinspections required by agencies other than customs,” and “port or border handling.” Commercecalculated separate B&H costs for exports and imports.

SSV Case Brief:16

Commerce should exclude all costs for document preparation from the calculation of B&Hbecause none of the documents whose preparation is included in the Doing Business report isprepared by SSV’s export or import brokers. Some document preparation costs are incurredby SSV personnel, and these costs are included in the overhead “financial ratios” thatCommerce calculates from surrogate financial statements. Other document preparation costs(for exports) are incurred by the ocean shipping company, which is a market-economy (ME)supplier. These costs are captured in the fees that SSV pays to the ME supplier. Still otherdocument preparation costs (for imports) are incurred by the supplier of the raw material, orthe ocean shipping or marine insurance company. These costs are captured in the surrogatecountry import values, which are stated on a CIF basis.

15 See Preliminary Results PDM at 10.16 See SSV Case Brief at 7-18.

5

When Commerce values an input using an ME purchase price, and adds B&H costs to theME purchase price, Commerce is thereby incorrectly inflating the cost of B&H. Thisconclusion follows from how companies maintain their accounting records. When acompany’s accounting records do not allow materials handling costs to be identifiedspecifically with particular raw material items in inventory, the costs may be classified asmaterials overhead. For such companies, the handling costs would already be captured in therelevant financial ratios as part of overhead. This means that adding them separately to directmaterials costs would result in double-counting. This double-counting would occur whetherthe surrogate values for the material inputs are based on ME purchases or on surrogate-country import values. Consequently, the inclusion of import B&H costs when valuinginputs in an NME proceeding is never appropriate.If Commerce uses Doing Business to value B&H in the final results, it should allocate thecosts in a way that is consistent with judicial precedent.17 In CS Wind Vietnam, the Court ofInternational Trade (CIT) remanded to Commerce for recalculation or further explanation itscalculation of B&H in which it calculated document preparation costs based on the weight ofthe container.18 The CIT stated common sense indicates that a half-full twenty-foot containerwould incur the same document preparation expenses as a full twenty-foot container.19 Onremand, Commerce agreed to allocate document-preparation costs over the average quantityper export transaction.20 Furthermore, in DuPont Teijin, the CIT applied the same logic to“customs clearance and technical control.”21 Upon remand, Commerce agreed to allocate theB&H costs over the weight shipped per export transaction.22 Commerce should use the samemethodology here.In addition to the evidence from prior court rulings, price lists on the record of this reviewfrom Indian companies that provide export and import B&H services at Indian ports alsoshow that “document preparation” and “customs clearance and technical control” charges areassessed at a fixed amount per bill of lading.23 Thus, there is no basis for Commerce tocalculate either export or import B&H expenses (as it did in the Preliminary Results) on theassumption that “document processing” and “customs clearance and technical control” wouldvary with the weight of the shipment. However, unlike “document preparation” and“customs clearance and technical control,” Indian export and import B&H price lists on therecord indicate that “terminal handling charges” are set as a fixed amount per containerregardless of the weight of the container.24

Therefore, in light of this record evidence, Commerce should calculate per-unit export andimport “document preparation” and “customs clearance and technical control” using SSV’s

17 Id. at 9, 11, and 16 (citing CS Wind Vietnam v. United States, 971 F. Supp. 2d 1271 (CIT 2014) and Dupont TeijinFilms China v. United States, 7 F. Supp. 3d 1338, 1351-52 (CIT 2014)).18 Id. at 9-11 (citing CS Wind Vietnam v. United States, 971 F. Supp. 2d 1271 (CIT 2014) (CS Wind Vietnam)).19 Id.20 Id. at 11 (citing CS Wind Vietnam v. United States Results of Redetermination Pursuant to Court Order (July 28,2014) (CS Wind Vietnam Redetermination) at 21).21 Id. at 11 (citing Dupont Teijin Films China v. United States, 7 F. Supp. 3d 1338, 1351-52 (CIT Sept. 11, 2014)(DuPont).22 Id. (citing Redetermination Pursuant to Court Remand Order in DuPont Teijin Films China Limited et al. v.United States, Consol. Court No. 13-00229 (January 9, 2015) at 8).23 Id. at 11 and 16 (citing SSV March 26, 2019 Surrogate Value Submission (SSV SVS) at Exhibit 7-A).24 Id. at 15 (citing SSV SVS at Exhibit 7-A).

6

average per-invoice export quantity of OCTG and import quantity of hot-rolled coil (HRC)during the POR, respectively.With respect to “terminal handling charges” for exports, SSV did not export OCTG incontainers, and therefore did not incur any terminal handling charges for such shipments.However, it did export other circular welded products in containers, and SSV has reportedthe average shipment quantity per container for those shipments. If Commerce uses DoingBusiness to value B&H in the final results, it should calculate the per-ton cost for “terminalhandling and port charges” by dividing the amounts shown in Doing Business by SSV’sreported average export quantity.With respect to “terminal handling charges” for imports, SSV did not import HRC incontainers. Therefore, if Commerce decides to apply “terminal handling charges” to importsof steel coils, it should calculate per-ton costs using either the average weight per containerfor SSV’s exports of circular welded pipe products or the maximum weight per container setforth in published shipping documents.

Domestic Interested Parties’ Rebuttal Brief:25

Commerce should not remove the document preparation from the calculation of B&H.o Doing Business does not disaggregate costs for specific documents. Therefore,

Commerce should not attempt to adjust the Doing Business costs for any specificdocument because there is no way to know the costs associated with preparing specificdocuments. Moreover, Doing Business is based on a broad survey of many companiesand their different experiences preparing a variety of documents.

o Commerce already addressed this argument in the OCTG Vietnam AR1 Final, andconcluded it was not appropriate “to remove any of the cost categories from thesummation of B&H in Doing Business 2016,” and that Commerce was “unable to makean adjustment for individual documents because Doing Business 2016 does not providecosts for individual documents.”26

o With respect to SSV’s argument that document processing costs are captured in thefinancial ratios of the surrogate financial statements, this argument incorrectly assumesthat the surrogate financial ratios fully account for the document-processing functionsthat SSV asserts are part of the company’s own selling, general, and administrative(SG&A) expenses. Such a level of detail cannot be discerned from the financialstatements used in the Preliminary Results.

Commerce should continue to use a weight-based allocation in the final results.o Commerce has previously explained that it must allocate B&H expenses over the weight

of a standard container because this is the weight used in Doing Business, and thecalculation needs to be internally consistent with the original data’s reporting basis.27

25 See Domestic Interested Parties’ Rebuttal Brief at 4-15.26 Id. at 6 and 14 (citing Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Final Resultsof Antidumping Duty Administrative Review; 2014-2015, 82 FR 18611 (April 20, 2017 (OCTG Vietnam AR1 Final),and accompanying Issues and Decision Memorandum (IDM) at Comment 2).27 Id. at 9 and 15 (citing OCTG Vietnam AR1 Final IDM at Comment 2).

7

o In the litigation of the investigation stage of this proceeding, the CIT sustainedCommerce’s use of a weight-based calculation.28

o Using a weight-based calculation is consistent with Commerce’s practice in other recentcases.29

o SSV’s OCTG exports do not represent a “unique type of product, nor are its shippingconfigurations so unusual as to require Commerce to depart from the Doing Businessweight-based calculation in order to add more specificity and accuracy to the dumpingmargin calculation.30

With respect to SSV’s arguments that “terminal handing charges” are set as a fixed amountper container, and do not vary by weight, Commerce previously rejected this argument inOCTG Vietnam AR1 Final. Commerce has found that “customs clearance costs can changedepending on the size of the shipment,”31 and this approach has been upheld by the CIT.32

With respect to SSV’s argument that in an NME proceeding it is never appropriate to includeB&H costs when valuing inputs, SSV submitted this argument in OCTG Vietnam AR1 Final,and Commerce rejected it.33 Commerce stated that SSV had not shown that the B&H costsare included in any of the overhead reported on the surrogate financial statements on therecord.34 SSV has made no such showing here either, nor shown that B&H costs areincluded in overhead on its own financial statement.

Commerce Position:

In its Preliminary Results, Commerce applied the same methodology for calculating B&H costsas it did in the initial investigation of this proceeding, and also in the most recently completedadministrative review of the order.35 However, after parties filed case and rebuttal briefs in thisproceeding, the Court of Appeals for the Federal Circuit (CAFC) issued a ruling with respect tothe litigation on the investigation stage of this proceeding.36 In that ruling, the CAFC disagreedwith the methodology Commerce took in calculating B&H in the investigation. The CAFCagreed with an earlier CIT ruling that “Commerce’s {by-weight B&H allocation} methodologyincorrectly assumes that a shipment weighing less will incur lower document preparation andcustoms clearance costs, while a shipment weighing more will incur higher preparation costs.”37

28 Id. at 8 and 12 (citing SeAH VINA v. United States, 332 F. Supp. 3d 1314 (CIT 2018) (SeAH VINA) at 1330-31).29 Id. at 8 (citing Certain Polyethylene Terephthalate Resin from the People’s Republic of China: FinalDetermination of Sales at Less Than Fair Value, 81 FR 13331 (March 14, 2016), and accompanying IDM atComment 3; Citric Acid and Certain Citrate Salts from the People’s Republic of China: Final Results ofAntidumping Duty Administrative Review; 2013-2014, 80 FR 77323 (December 14, 2015), and accompanying IDMat Comment 6).30 Id. at 10 and 16 (citing CS Wind Redetermination at 16).31 Id at 10 (citing OCTG Vietnam AR1 Final IDM at Comment 2).32 Id. (citing SeAH VINA, 332 F. Supp. 3d at 1330).33 Id. (citing OCTG Vietnam AR1 Final IDM at Comment 2).34 Id.35 See Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Final Determination of Sales atLess Than Fair Value and Final Affirmative Determination of Critical Circumstances (OCTG Vietnam InvestigationFinal), 79 FR 41973 (July 18, 2014), and accompanying IDM at Comment 1; OCTG Vietnam AR1 Final IDM atComment 2.36 See SeAH Steel VINA Corporation v. United States, 950 F.3d 833, 847 (Fed. Cir. 2020) (SeAH VINA II).37 Id. at 847 (citing Dupont).

8

The CAFC concluded that “Commerce’s by-weight allocation methodology for B&H costs is, asapplied here, unsupported by substantial evidence.”38 In light of this CAFC ruling, we havereconsidered our per-unit B&H calculation methodology for these final results of review.

As an initial matter, as explained more fully in Comment 5 (below), Commerce has determinednot to value HRC or any other material input using SSV’s ME purchase prices from Korea inthese final results of review due to the existence of generally available export subsidies in Korea.Instead, we have valued inputs sourced from Korea using Indian import statistics obtained fromthe Global Trade Atlas (GTA). With respect to our treatment of B&H costs when using GTAdata, we note that in a previous antidumping investigation in which SSV was a respondent,Commerce stated, “We agree with {SSV} that it is not {Commerce’s} practice to add brokerageand handling costs to Indian {surrogate values}, as these costs are already captured in the{surrogate values}.”39 In accordance with this practice, we have not added B&H costs to theGTA surrogate values in these final results, and, SSV’s argument that “the inclusion of importB&H costs when valuing inputs in an NME proceeding is never appropriate” is moot.

Furthermore, we agree with SSV that we should not include the cost category “documentpreparation” in the calculation of total B&H costs, and have therefore removed this cost categoryin these final results of review. We made this change for these final results because SSV hasaccounted for the preparation of every document included in Doing Business B&H computation,and shown that they are prepared by either the ocean freight company or SSV personnel, or thatit is not a relevant document for its exports.40 Although the petitioner is correct that DoingBusiness does not disaggregate the costs of individual documents, it is not necessary that it do sobecause SSV has accounted for all of them. Commerce has stated, “{Commerce} willsometimes make an adjustment to surrogate value data to reflect an individual exporter’sexperience, including to B&H surrogate value data, but normally only when the item’s amount isclearly identified in the “Doing Business” report and the factors of production for self-preparation are accounted for.”41 Here, SSV has accounted for the entire expense category ofdocument preparation. Thus, disaggregating the costs of individual documents is unnecessary.

With respect to the remaining cost elements of export B&H, we have continued to use DoingBusiness data, and have adopted the calculation methodology that SSV proposes. Specifically:

o For “customs clearance and technical control,” we calculated a per-unit amount by dividingthe total costs by SSV’s reported average per-shipment export quantity.

o For “terminal handling charged,” we calculated a per-unit amount by dividing the total costsby the maximum container weight given in published shipping documents.

38 Id. at 846.39 See Circular Welded Carbon-Quality from the Socialist Republic of Vietnam: Final Determination of Sales atLess Than Fair Value, 81 FR 75042 (October 28, 2016), and accompanying IDM at Comment 6; see also OCTGVietnam AR1 Final IDM at Comment 2.40 See SSV Letter, “Administrative Review of the Antidumping Order Certain Oil Country Tubular Goods fromVietnam for the 2017-18 Review Period — Submission of Rebuttal Factor Value Information,” dated April 1, 2019,at Attachment 4.41 See OCTG Vietnam Investigation Final IDM at Comment 1 (citations omitted); see also OCTG Vietnam AR1Final IDM at Comment 2.

9

Using this methodology is consistent with record evidence of how B&H costs are incurred,42 anddoes not result in a computation in which B&H costs increase proportionately with the weight ofthe shipment. It is therefore consistent with the CAFC’s ruling in SeAH VINA II. For furtherdetails of the calculation, see SSV Final Analysis Memorandum.43

Comment 2: Surrogate Value for Water

In the Preliminary Results, Commerce treated water as a direct cost, and valued it usinginformation from the website of Maharashtra Industrial Development Corporation (MIDC). TheMIDC information included separate figures for “inside industrial area for industrial use” and for“outside industrial area for industrial use.”44 Commerce valued water using the average of thetwo figures.45

SSV Case Brief:46

As SSV has reported, SSV’s production facilities are located entirely within an industrialzone.47 The information obtained from MIDC confirms that the cost of water provided forindustrial use in industrial zones is lower than the cost in other areas, presumably becausesuch zones can have dedicated infrastructure and other economies of scale that reduce thesupplier’s costs. There is no reason to believe that industrial zones in Vietnam do not benefitfrom equivalent efficiencies. Consequently, if Commerce decides to assign a value to thewater that SSV uses, Commerce should use only the figures for usage “inside industrial areafor industrial use.”However, water is not incorporated in the subject merchandise during the production process,and the cost of water is typically considered a part of factory overhead. In fact, one of thefinancial statements Commerce used to calculate “financial ratios,” that of Swastik Pipe,explicitly listed “water charges” as a separate item of “manufacturing expenses.”48 Becausethe cost of water is included in the factory overhead rates calculated from the financialstatements of Indian producers of comparable merchandise, the inclusion of a separate valueto value water based on the reported usage and the Indian values downloaded from the MIDCwebsite would double-count the relevant cost. Therefore, Commerce should treat water asoverhead in its final results.

Domestic Interested Parties’ Rebuttal Brief:49

42 See SSV SVS at Exhibit 7-A.43 See Memorandum, “Analysis for the Final Results of the Fourth Antidumping Duty Administrative Review ofCertain Oil Country Tubular Goods from the Socialist Republic of Vietnam; SeAH Steel VINA Corporation,” datedJuly 2, 2020 (SSV Final Analysis Memorandum).44 Preliminary Determination Memorandum at 11.45 See Memorandum, “Surrogate Values for the Preliminary Results,” dated November 8, 2019 (SV Memorandum),at Exhibit 6.46 See SSV Case Brief at 18-19.47 Id. at 19 (citing SSV’s December 19, 2018, Section A Questionnaire Response (SSV December 19, 2018 AQR) at13).48 Id. (citing Swastik Pipe Financial Statements at 111, provided in SSV SVS at Attachment 11-A-10).49 See Domestic Interested Parties’ Rebuttal Brief at 16-19.

10

Although SSV operates within an industrial area in Vietnam, there is no record evidence toshow that companies located within industrial areas in Vietnam pay different rates for watercompared to companies outside industrial areas that also use water for industrial uses.Accordingly, there is no record evidence to demonstrate that the MIDC data is notrepresentative of SSV’s production experience or that it results in a less-accurate dumpingmargin calculation.Absent alternative surrogate value information for water, the average MIDC-based surrogatevalue for water used in the Preliminary Results represents the best available information onthe record for valuing SSV’s water consumption.Commerce has used this same surrogate value source and calculation methodology in twoprior segments of this proceeding,50 and there is no reason to depart from the MIDC-basedsurrogate value for water when calculating normal value for the final results.SSV’s questionnaire responses demonstrate unequivocally that it consumes water during theOCTG production process.51

Commerce’s practice is to value water as an energy input, not as factory overhead, evenwhen water is not incorporated directly into the finished product “when water is used formore than incidental purposes.”52 SSV’s questionnaire responses demonstrate that water is akey input for its production of OCTG.53 Commerce should follow its usual practice ofassigning a surrogate value to the water consumption that SSV reported on its factors ofproduction (FOP) database.There is no merit to SSV’s argument that Commerce should not value water consumptionbecause Swastik Pipe might have reported water as a manufacturing expense on its financialstatement. As described above, SSV consumes water as a direct input in the production ofOCTG, and Commerce’s established practice is to assign a surrogate value to water when itis consumed in the production process “for more than incidental purposes.”54 Moreover,Swastik Pipe’s financial statements do not appear to be publicly available as required byregulation, and should therefore be disregarded.55

Commerce Position:

We disagree with SSV that we should not assign a value to water in the calculation of NV.Although water may sometimes be classified as overhead, Commerce has stated, “Normally,{Commerce} values water directly and not in factory overhead when water is used for more thanincidental purposes, is required for a particular segment of the production process, or appears to

50 Id. at 18 (citing OCTG Vietnam Investigation Final IDM at Comment 4, and OCTG Vietnam AR1 Final IDM atComment 3).51 Id. (citing SSV’s February 19, 2019 Section D Questionnaire Response (SSV February 19, 2019 DQR) at 23).52 Id. (citing OCTG Vietnam AR1 Final IDM at Comment 3 (citing Automotive Replacement Glass Windshields fromthe People’s Republic of China: Final Results of Administrative Review, 69 FR 61790 (October 21, 2004) (GlassWindshields China Final).53 Id. at 19 (citing SSV February 19, 2019 DQR at 23 and Appendix D-5).54 Id. (citing OCTG Vietnam AR1 Final IDM at Comment 3).55 Id. (citing 19 CFR 351.408(c)(1), (4)).

11

be a significant input in the production process.”56 Although SSV states in its case brief that“water is not incorporated in the subject merchandise,”57 it previously submitted the informationthat water is used in the production process of OCTG. Specifically, SSV stated, “{SSV} useselectricity and water in its production of subject OCTG products.” 58 Therefore, because water isan integral part of the production process for OCTG, we have included water as a direct input inthese final results.

With respect to the calculation methodology used for water, we disagree with SSV that weshould use only the water rates for industrial zones. In OCTG Vietnam Investigation Final andOCTG Vietnam AR1 Final, we determined that although Vietnamese companies located withinindustrial areas may benefit from economies of scale that reduce the supplier’s costs, there wasno record evidence that Vietnamese producers located entirely within industrial areas are chargedlower water consumption rates.59 Furthermore, in the 2012 countervailing duty investigation ofcircular-welded pipe from Vietnam, Commerce found that there was no separate tariff water ratefor companies located within industrial zones.60 The record here is likewise devoid ofinformation establishing a separate tariff rate for companies located within industrial zones inVietnam. Therefore, in these final results of review we continue to find that the water SV used inthe Preliminary Results is the best information available on the record, and we have not revisedour calculation of the SV for water.

Comment 3: Differential Pricing

SSV Case Brief:61

Commerce is required to justify the numerical thresholds used in the differential pricing analysisbased on substantial evidence on the record.

In the Preliminary Results, Commerce applied its differential pricing analysis to determinewhether there was a pattern of prices for comparable merchandise that differed significantlyamong purchasers, regions, or time periods for SSV’s U.S. sales.62 Commerce’s use of thedifferential pricing analysis is both mathematically and legally improper.In order to adopt apparently arbitrary cut-offs (such as those used in the differential pricinganalysis) Commerce must either follow the requirements of the Administrative ProceduresAct (including its notice-and-comments requirements), or explain in each case why any of

56 See OCTG Vietnam Investigation Final IDM at Comment 4; see also Glass Windshields China Final IDM atComment 1.57 See SSV Case Brief at 19.58 See SSV February 19, 2019 DQR at 23.59 See OCTG Vietnam Investigation Final IDM at Comment 4; and OCTG Vietnam AR1 Final IDM at Comment 3.60 See Circular Welded Carbon-Quality Steel Pipe from the Socialist Republic of Vietnam: Final NegativeCountervailing Duty Determination, 77 FR 64471 (October 22, 2012), and accompanying IDM at 5 (“Hongyuan andSeAH VINA sourced their water from industrial development companies. We verified that both companies paid theapplicable tariff rates for their water and there was no separate tariff rate for companies located within the industrialzones”).61 See SSV Case Brief at 20-33.62 Id. at 20 (citing PreliminaryResults PDM at 14).

12

the numerical thresholds used in connection with the test are appropriate in the context ofeach specific case. Commerce has done neither of those things here.o This principle was recognized by the CIT and CAFC in cases addressing the de minimis

standard applied by Commerce in investigations.63 In the past, Commerce applied a 0.5percent de minimis standard as a matter of policy, but without a specific provision in theregulations for it at the time. In litigation, the CIT and CAFC held that because the deminimis standard had not been promulgated as a regulation in accordance with theprovisions of the APA, Commerce was not permitted to apply that rule automatically inevery case.64 Thus, under the principles recognized in Carlisle Tire and WashingtonRaspberries, Commerce’s use of the differential pricing analysis can be sustained only ifit provides both evidence and analysis showing why the cut-offs used the 0.8 cut-off usedfor the Cohen’s d test and the 33- and 66-percent cut-offs used for the “ratio test.”

The 0.8 cut-off used in the Cohen’s d test portion of the differential pricing analysis is notsupported by substantial evidence on the record.

When Commerce first applied its differential pricing analysis, it asserted that the reliance onthe Cohen’s d test (and, in particular, on the 0.8 cut-off for determining whether an effectsize is large) is appropriate because the cut-offs proposed by Professor Cohen “have beenwidely adopted.”65 However, Professor Cohen himself made clear that his proposed cut-offscould only appropriately be applied in specific circumstances —where “samples, each of ncases, have been randomly and independently drawn from normal populations,” and wherethe two samples do not have “substantially unequal variances” or “substantially unequalsample sizes (whether small or large).”66

In past cases, Commerce admitted that Professor Cohen placed such limitations on hisanalysis, but maintained that those limitations applied only to the section describing the“Introduction and Use” of his chapter on the “The T Test for Means.”67 According toCommerce, the “The T Test for Means” is irrelevant to the differential pricing analysis,because “this concerns the statistical significance of the difference in the means for twosampled sets of data, and is not relevant when considering whether this difference has apractical difference.”68

While Commerce may consider the “The T Test for Means” to be irrelevant to its goal ofidentifying “targeted dumping,” it was very much relevant to Professor Cohen’s developmentand presentation of his d statistic, and the various cut-offs he proposed for determiningwhether d is small, medium, or large. While the title of Chapter 2 of Statistical Power is

63 Id. at 21-22 (citing Carlisle Tire v. United States, 634 F. Supp. 419, 423 (CIT 1986) (Carlisle Tire); WashingtonRed Raspberry Commission. v. United States, 859 F. 2d 898, 903 (Fed. Cir. 1988) (Washington Raspberries); and IPSCO v. United States, 687 F. Supp. 614, 630-31 (CIT 1988) (IPSCO)).64 Id.65 Id. at 23 (citing Final Results of the Antidumping Duty Administrative Review: Welded ASTM A-312 StainlessSteel Pipe from the Republic of Korea; 2013-2014, 81 FR 46647 (July 18, 2016), and accompanying IDM at 15). 66 Id. at 24 (citing Certain Oil Country Tubular Goods from the Republic of Korea: Amended Final Results ofAntidumping Duty Administrative Review; 2014-2015, 82 FR 31750 (July 10, 2017) and OCTG Korea AR1 FinalIDM at 22 (citing Cohen, Statistical Power Analysis for the Behavioral Sciences (2nd ed. 1988) (Statistical Power)at 19-20).67 Id. (citing OCTG Korea AR1 Final IDM at 22).68 Id.

13

“The T Test for Means,” Professor Cohen’s entire explanation of the d statistic is found inChapter 2 and is the only subject of that chapter.In past cases Commerce argued that its differential pricing analysis is not meant to be a“power analysis,” and has claimed that the development and presentation of the d statisticwas meant to establish a framework for something very different from the differential pricinganalysis.69 In other words, Commerce has taken a statistical tool that it claims was intendedto provide a “T-Test for Means” in order to “guide researchers in their construction of aproject in order to obtain a prescribed ‘power,’” and used it for a purpose and in a situationthat Professor Cohen never intended. As a result, nothing in Statistical Power supportsCommerce’s claim that the cut-offs used are “fixed thresholds” or can provide justificationfor Commerce’s use of the d statistic in its differential pricing analysis in situations that arenot consistent with the limitations that Professor Cohen described.As noted above, the d statistic can only be used where “samples, each of n cases, have beenrandomly and independently drawn from normal populations,” and where the two samples donot have substantially unequal variances or sample sizes.70 Commerce made no suchdetermination with respect to SSV’s U.S. sales data meeting those requirements. In suchcircumstances, the d statistic simply does not provide meaningful results.As a separate matter, Commerce has asserted in the past that the conditions laid out byProfessor Cohen are irrelevant because it is analyzing the complete population of therespondent’s U.S. sales, and not just a sample. Mathematically, that assertion is untenable.Normal distributions — whether consisting of a sample or the entire population — havemathematical characteristics that non-normal distributions do not have. Professor Cohenexplicitly relied on the mathematical characteristics of normal distributions in thedevelopment and justification of his proposed cut-offs.71

Commerce’s assertion that Professor Cohen’s proposed cutoffs can be used whenever acomplete population is being analyzed is completely unsupported by any evidence on therecord. A number of academic analyses demonstrate that, when the conditions set out byProfessor Cohen are not met, the d statistic ceases to be a useful measure of effect sizes.Commerce has not cited to any evidence to support its novel assertion that a parametric testdesigned for the analysis of two normally-distributed data sets with roughly equal number ofdata points and roughly equal variance can be used when none of those conditions exist, aslong as the entire population is being considered.

The 33- and 66-percent cut-offs used in the “ratio test” portion of the differential pricinganalysis are not supported by substantial evidence on the record.

Commerce has never explained why the thresholds used in the “ratio test” should be 33 and66 percent, and not some other numbers (such as 40 and 80 percent, 50 and 90 percent, orany two other numbers between 0 and 100), nor has it explained why a ratio between 33 and66 percent calls for consideration of the transaction-to-average methodology only for the

69 Id. at 25 -26 (citing OCTG Korea AR1 Final IDM at 22, note 61; Welded Line Pipe from Korea: Final NegativeCountervailing Duty Determination, 80 FR 61365 (October 13, 2015), and accompanying IDM at VI.B.1 andComment 1).70 Id. at 27 (citing OCTG Korea AR1 Final IDM at 22, note 61).71 Id. at 28 (citing OCTG Korea AR1 Final IDM at 22-23, note 62).

14

sales that “pass” the Cohen’s d test, while a ratio of 66 percent or more calls for theapplication of the transaction-to-average methodology for all sales.72 Commerce’s numericalthresholds have not been established through notice-and-comment rulemaking, then they canonly be upheld if supported by substantial evidence on the record in each case in which theyare applied.73

The “differential pricing analysis” fails to explain why any patterns of price differences werenot, or could not be, taken into account using an average-to-average comparison.

The statute permits Commerce to depart from the normal A-to-A comparison to account fortargeted dumping (in investigations) only if it “explains why such differences cannot betaken into account using” an A-to-A or transaction-to-transaction (T-to-T) calculationmethodology.74

The price differences that give rise to a finding of targeted dumping are primarily a functionof the different treatment of negative dumping margins under Commerce’s “standard”methodology where “zeroing” is not used, and its “alternate” methodologies where negativemargins are “zeroed.” Differences in dumping margins generated by the application of“zeroing” are not the same as differences in dumping margins caused by patterns of pricedifferences by customer, region, or time period, and Commerce has failed to explain whythose patterns cannot be addressed using the normal comparison methodologies.In addition, Commerce has not provided any support for its assertion that the difference inweighted-average dumping margins is “meaningful” when the weighted-average dumpingmargin crosses the de minimis threshold when using the alternative calculation instead of theaverage-to-average and alternative calculation method. Therefore, without a reasonable basisfor that numerical threshold, Commerce’s use of the de minimis measure to decide whichmargin calculation to apply is inherently arbitrary and improper.

Under the relevant provisions of the Act, Commerce is not permitted to utilize an average-to-transaction comparison methodology for any of SSV’s U.S. sales.

The Act normally requires Commerce to calculate dumping margins in investigations usingone of two methodologies: by comparing an average NV to an average U.S. price, or bycomparing the NVs for individual transactions to the U.S. prices for individual transactions.75

As a general rule, the Act does not permit Commerce to compare an average NV to U.S.prices for individual transactions in an investigation.The statute provides an exception to this general rule when targeted dumping is found toexist, but that exception, which might permit Commerce to calculate dumping margins bycomparing an average NV to U.S. prices for individual transactions, applies only when thereis a pattern of export prices (or constructed export prices) for comparable merchandise thatdiffer significantly among purchasers, regions, or periods of time, and if the administering

72 Id. at 29 (citing e.g., OCTG Korea AR1 Final IDM at 25). 73 Id. at 30 (citing Carlisle Tire, 634 F. Supp. at 423; Washington Raspberries, 859 F. 2d at 903; and IPSCO, 687 F.Supp. at 630-1).74 Id. at 31 (citing section 777A(d)(1)(B) of the Act). 75 Id. at 32 (citing section 777A(d)(1)(A) of the Act).

15

authority explains why such differences cannot be taken into account using a methoddescribed in section 777A (d)(1)(A)(i) or (ii) of the Act.76 If those conditions are not met,Commerce is not permitted to depart from the A-to-A (or T-to-T) methodology that isnormally required in investigations.The conditions permitting the use of an average-to-transaction (A-to-T) comparisonmethodology are not satisfied in this case and the exception set forth in section777A(d)(1)(B) of the Act does not apply, and Commerce is required by statute, to continue tocalculate dumping margins using the A-to-A methodology for all of SSV’s U.S. sales in itsfinal results.

Petitioner’s Rebuttal Brief:77

As in other proceedings, SSV challenges Commerce’s use of its differential pricing analysisto determine whether there was a pattern of prices for comparable merchandise that differssignificantly among purchasers, regions, or time period as mathematically and legallyimproper. More specifically, in this review, SSV repeats the arguments it submitted mostrecently in OCTG Korea AR3 Final: (a) Commerce has not sufficiently explained or justifiedits analytical bases for the use of the certain cutoffs in the Cohen’s d test (0.8 percent) andthe ratio test (33 and 66 percent, respectively); (b) Commerce must explain its use of themethodology in every case; (c) that the differential pricing analysis must explain why an A-to-A or T-to-T methodology is insufficient to calculate a dumping margin; and (d) thatCommerce’s use of an A-to-T comparison methodology is contrary to the statute.78 In thefirst three administrative reviews of the antidumping duty order on OCTG from Korea,Commerce specifically addressed and rejected these challenges.79

The CIT and CAFC have also upheld Commerce’s application of the differential pricingmethodology.80 In fact, the CIT recently upheld Commerce’s differential pricingmethodology in an appeal of OCTG Korea AR2 Final, in which SSV likewise challenged theexact same aspects of the differential pricing methodology that it now raises beforeCommerce in its case brief.81

Commerce Position:

76 Id. at 33 (citing section 777A(d)(1)(B) of the Act). 77 See Domestic Interested Parties’ Rebuttal Brief at 19-24.78 Id. at 23 citing Certain Oil Country Tubular Goods from the Republic of Korea: Final Results of AntidumpingDuty Administrative Review; 2016-2017, 84 FR 24085 (May 24, 2019), (OCTG Korea AR3 Final), and accompanying IDM at Comment 3).79 Id. at 21 (citing Certain Oil Country Tubular Goods from the Republic of Korea: Final Results of AntidumpingDuty Administrative Review: 2014-2015, 82 FR 18105 (April 17, 2017) (OCTG Korea AR1 Final), and accompanying IDM at Comment 2; Certain Oil Country Tubular Goods From the Republic of Korea: Final Resultsof Antidumping Duty Administrative Review and Final Determination of No Shipments; 2015-2016, 83 FR 17146(April 18, 2018) (OCTG Korea AR2 Final), and accompanying IDM at Comment 8; and OCTG Korea AR3 FinalIDM at Comment 3).80 Id. at 20 (citing Apex Frozen Foods Private Lt. v. United States, 862 F. 3d 1337, 1344-51 (Fed. Cir. 2017) (ApexIV); NEXTEEL Co. v. United States, 355 F. Supp. 3d 1336, 1354-57 (CIT) (NEXTEEL I); NEXTEEL Co. v. UnitedStates, 392 F. Supp. 3d 1276, 1294-97 (CIT) (NEXTEEL II); and APEX Frozen Foods Private Ltd. v. United States,144 F. Supp. 3d 1308, 1314-37 (CIT 2016) (Apex II).81 Id. at 24 (citing NEXTEEL II, 392 F. Supp. 3d at 1294-1297).

16

As an initial matter, we note that there is nothing in section 777A(d) of the Act that mandateshow Commerce measures whether there is a pattern of prices that differs significantly or explainswhy the average-to-average (A-to-A) method or the transaction-to-transaction (T-to-T) methodcannot account for such differences. On the contrary, carrying out the purpose of the Act82 hereis a gap filling exercise properly conducted by Commerce.83 As explained in the PreliminaryResults, as well as in various other proceedings,84 Commerce’s differential pricing analysis isreasonable, including the use of the Cohen’s d test as a component in this analysis, and it is in noway contrary to the law.

We disagree with SSV that the differential pricing analysis, including the Cohen’s d test, isunreasonable, unlawful, or arbitrary. As an initial matter and to the contrary, we note that theCAFC has upheld key aspects of Commerce’s differential pricing analysis, including theapplication of the “meaningful difference” standard, which compares the rate calculated usingthe A-to-A method not using zeroing and the rate calculated using an alternative comparisonmethod based on the A-to-T method using zeroing; the reasonableness of Commerce’scomparison method in fulfilling the relevant statute’s aim; Commerce’s use of a “benchmark” toillustrate a meaningful difference between the A-A and A-T rates; Commerce’s justification forapplying the A-to-T methodology to all sales instead of just those that pass the Cohen’s d test;Commerce’s use of zeroing in applying the A-to-T methodology to all transactions; that thestatute does not directly apply to reviews; Congress did not dictate how Commerce shoulddetermine if the A-to-A method accounts for targeted or masked dumping; Commerce mayconsider all sales in its “meaningful difference” analysis and consider all sales when calculatinga final rate using the A-to-T method; and it is acceptable to apply zeroing when using the A-to-Tmethod.85 In NEXTEEL, the CIT rejected SSV’s challenge to our differential pricing analysisand held that “the steps underlying the differential pricing analysis as applied by Commerce

82 See Koyo Seiko Co., Ltd. v. United States, 20 F. 3d 1156, 1159 (Fed. Cir. 1994) (“The purpose of the antidumpingstatute is to protect domestic manufacturing against foreign manufacturers who sell at less than fair market value.Averaging U.S. prices defeats this purpose by allowing foreign manufacturers to offset sales made at less-than-fairvalue with higher priced sales. Commerce refers to this practice as ‘masked dumping.’ By using individual U.S.prices in calculating dumping margins, Commerce is able to identify a merchant who dumps the productintermittently—sometimes selling below the foreign market value and sometimes selling above it. We cannot saythat this is an unfair or unreasonable result.” (internal citations omitted)).83 See Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842-43 (1984) (Chevron) (recognizingdeference where a statute is ambiguous and an agency’s interpretation is reasonable); see also Apex Frozen FoodsPrivate Limited v. United States, 37 F. Supp. 3d 1286, 1302 (Apex I) (applying Chevron deference in the context of Commerce’s interpretation of section 777A(d)(1) of the Act).84 See, e.g., Welded Line Pipe from the Republic of Korea: Final Determination of Sales at Less Than Fair Value,80 FR 61366 (October 13, 2015), and accompanying IDM at Comment 1; Circular Welded Non-Alloy Steel Pipefrom the Republic of Korea: Final Results of Antidumping Duty Administrative Review; 2012-2013, 80 FR 32937(June 10, 2015), and accompanying IDM at Comments 1 and 2; Welded ASTM A–312 Stainless Steel Pipe from theRepublic of Korea: Final Results of Antidumping Duty Administrative Review; 2013–2014, 81 FR 46647 (July 18,2016), and accompanying IDM at Comment 4; Certain Hot-Rolled Steel Flat Products from Japan: FinalDetermination of Sales at Less Than Fair Value and Final Affirmative Determination of Critical Circumstances, 81FR 53409 (August 12, 2016), and accompanying IDM at Comment 8. 85 See Apex II; Apex Frozen Foods Private Ltd. v. United States, 862 F. 3d 1322 (Fed. Cir. 2017) (Apex III); ApexIV; Mid Continent Steel & Wire, Inc. v. United States, 940 F. 3d 662 (Fed. Cir. 2017) (Mid Continent).

17

{are} reasonable.”86 As explained in the Preliminary Results, Commerce continues to developits approach pursuant to its authority to address masked dumping.87 In carrying out this statutoryobjective, Commerce determines whether “there is a pattern of export prices (or constructedexport prices) for comparable merchandise that differs significantly among purchasers, regions,or periods of time, and…. why such differences cannot be taken into account using {the A-to-Aor T-to-T comparison method}.”88 With the statutory language in mind, Commerce relied on thedifferential pricing analysis to determine whether these criteria are satisfied such that applicationof an alternative methodology may be appropriate.89

Because the statute does not explicitly discuss how Commerce should conduct its determinationof less than normal value in reviews,90 carrying out the purpose of the statute, here, is a gapfilling exercise properly conducted by Commerce.91 Commerce finds that the purpose of section777A(d)(1)(B) is to evaluate whether the A-to-A method is the appropriate measure to determinewhether, and if so to what extent, a given respondent is dumping the merchandise at issue in theU.S. market.92 While “targeting” and “targeted dumping” may be used as a general expression todenote this provision of the statute, these terms impose no additional requirements beyond thosespecified in the statute for Commerce to otherwise determine that the A-to-A method is notappropriate based upon a finding that the two statutory requirements have been satisfied.93 TheCIT and the CAFC have upheld Commerce’s application of its differential pricing analysis toevaluate the statutory requirements.94

SSV presents several arguments regarding Commerce’s differential pricing analysis in thePreliminary Results, the first of which is that Commerce should follow the APA to justify thenumerical thresholds used in the differential pricing analysis, i.e., the 0.8 cut-off used for theCohen’s d test and the 33- and 66-percent cut-offs used for the ratio test. As explained in pastdeterminations, the notice and comment requirements of the APA do not apply “to interpretativerules, general statements of policy, or rules of agency organization, procedure, or practice.”95

Further, Commerce normally makes these types of changes in practice (e.g., the change from thetargeted dumping analysis to the current differential pricing analysis) in the context of our

86 See NEXTEEL II.87 See Preliminary Results PDM at 14. 88 See section 777A(d)(1)(B) of the Act (emphasis added); see also Tri Union, 163 F. Supp. 3d 1255, 1302 (“{h}adCongress intended to impose upon Commerce a requirement to ensure statistical significance, Congress presumablywould have used language more precise than ‘differ significantly.’”). 89 See 19 CFR 351.414(c)(1).90 See Timken Co. v. United States, 968 F. Supp. 2d 1279, 1286 and n. 7 (CIT 2014).91 See Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44 (1984).92 See 19 CFR 351.414(c)(1).93 See, e.g., Samsung v. United States, 72 F. Supp. 3d 1359, 1364 (CIT 2015) (“Commerce may apply the A-to-Tmethodology ‘if (i) there is a pattern of export prices (or constructed export prices) for comparable merchandise thatdiffer significantly among purchasers, regions, or periods of time, and (ii) the administering authority explains whysuch differences cannot be taken into account using the A-to-A or T-to-T methodologies. Id. 1677f-1(d)(1)(B). Pricing that meets both conditions is known as ‘targeted dumping.’”).94 See, e.g., JBF RAK LLC v. United States, 991 F. Supp. 2d 1343 (CIT 2014); and JBF RAK LLC v. United States,790 F. 3d 1358 (Fed. Cir. 2015).95 See, e.g., OCTG Korea AR2 Final IDM at Comment 8 (citing 5 U.S.C. 553(b)(3)(A)).

18

proceedings, on a case-by-case basis.96 As the CAFC has recognized, Commerce is entitled tomake changes and adopt a new approach in the context of its proceedings, provided it explainsthe basis for the change, and the change is a reasonable interpretation of the statute.97 The CAFChas also held that Commerce’s meaningful difference analysis was reasonable.98 Moreover, theCIT in Apex II held that Commerce’s change in practice (from targeted dumping to itsdifferential pricing analysis) was exempt from the APA’s rule making requirements, stating:

Commerce explained that it continues to develop its approach with respect to theuse of {A-to-T} “as it gains greater experience with addressing potentially hiddenor masked dumping that can occur when {Commerce} determines weighted-average dumping margins using the {A-to-A} comparison method. Commerceadditionally explained that the new approach is “a more precise characterizationof the purpose and application of {19 U.S.C. 1677f-1(d)(1)(B)}” and is theproduct of Commerce’s “experience over the last several years… further research,analysis and consideration of the numerous comments and suggestions on whatguidelines, thresholds, and tests should be used in determining whether to applyan alternative comparison method based on the {A-to-T} method.” Commercedeveloped its approach over time, while gaining experience and obtaining input.Under the standard described above, Commerce’s explanation is sufficient.Therefore, Commerce’s adoption of the differential pricing analysis was notarbitrary.99

Moreover, the CIT acknowledged in Apex III that as Commerce “gains greater experience withaddressing potentially hidden or masked dumping that can occur when {Commerce} determinesweighted-average dumping margins using the {A-to-A} comparison method, {Commerce}expects to continue to develop its approach with respect to the use of an alternative comparisonmethod.”100 Further developments and changes, along with further refinements, are expected inthe context of our proceedings based upon an examination of the facts and the parties’ commentsin each case.

Regarding SSV’s arguments concerning our reliance on the Cohen’s d test and on the 0.8 cut-offfor determining whether an effect size is large, we disagree. As stated in the PreliminaryResults, the purpose of the Cohen’s d test is to evaluate “the extent to which the prices to aparticular purchaser, region, or time period differ significantly from the prices of all other salesof comparable merchandise.”101 The Cohen’s d coefficient is a recognized measure whichgauges the extent (or “effect size”) of the difference between the means of two groups.Commerce has previously noted that the Cohen’s d coefficient provides “a simple way of

96 See Differential Pricing Analysis; Request for Comments, 79 FR 26720, 26722 (May 9, 2014) (DifferentialPricing Comment Request).97 See Saha Thai Steel Pipe Company v. United States, 635 F. 3d 1335, 1341 (CAFC 2011); WashingtonRaspberries, 859 F. 2d at 902-03; and Carlisle Tire, 634 F. Supp. at 423 (discussing exceptions to the notice andcomment requirements of the APA).98 See Apex IV, 862 F. 3d at 1347-1351. 99 See Apex II. 100 See Differential Pricing Comment Request, 79 FR at 26722.101 See Preliminary Results PDM at 14.

19

quantifying the difference between two groups and has many advantages over the use of tests ofstatistical significance alone.”102 Commerce has also previously noted that Robert Coe, in EffectSize, has stated that “{e}ffect size quantifies the size of the difference between two groups, andmay therefore be said to be a true measure of the significance of the difference.”103 Commercehas noted that Effect Size points out that the precise purpose for which Commerce relies on theCohen’s d test to satisfy the statutory language, to measure whether a difference is significant.104

Moreover, the CAFC has affirmed this 0.8 threshold as reasonable.105

Further, in describing “effect size” and the distinction between effect size and statisticalsignificance, Commerce stated in Shrimp from Vietnam:

Dr. Paul Ellis, in Guide to Effect Sizes, introduces effect size by asking a question:“So what? Why do this study? What does it mean for the man on the street?” Dr.Ellis continues:

A statistically significant result is one that is unlikely to be the resultof chance. But a practically significant result is meaningful in the realworld. It is quite possible, and unfortunately quite common, for aresult to be statistically significant and trivial. It is also possible for aresult to be statistically nonsignificant and important. Yet scholars,from PhD candidates to old professors, rarely distinguish between thestatistical and the practical significance of their results.

In order to evaluate whether such a practically significant result is meaningful, Dr. Ellisstates that this “implies an estimation of one or more effect sizes.”

An effect size refers to the magnitude of the result as it occurs, orwould be found, in the population. Although effects can be observedin the artificial setting of a laboratory or sample, effect sizes exist inthe real world.106

Commerce further stated in Shrimp from Vietnam:

As recognized by Dr. Ellis in the quotation above, the results of an analysis may havestatistical and/or practical significance, and that these two distinct measures ofsignificance are independent of one another. In its case brief, VASEP {VietnamAssociation of Seafood Exporters} accedes to the distinction and meaning of “effect size”

102 See OCTG Korea AR3 Final IDM at Comment 3 (citing Robert Coe, It’s the Effect Size, Stupid: What “EffectSize” Is and Why It Is Important, 2002 Annual Conference of the British Educational Research Association,University of Exeter, Exeter, Devon, England, September 12–14, 2002 (Effect Size).103 Id.104 Id.105 See Mid Continent.106 See Certain Frozen Warmwater Shrimp from the Socialist Republic of Vietnam: Final Results of AntidumpingDuty Administrative Review, 2014-2015, 81 FR 62717 (September 12, 2016) (Shrimp from Vietnam), and accompanying IDM at 16-17 (citations omitted) (citing P. Ellis, The Essential Guide to Effect Sizes (Guide to EffectSizes) at 3-5.

20

when it states “{w}hile application of the t-test {a measure of statistical significance} inaddition to Cohen’s d might at least provide the cover of statistical significance, it stillwould not ensure practical significance.” {Commerce} agrees with this statement --statistical significance is not relevant to {Commerce’s} examination of an exporter’s U.S.prices when examining whether such prices differ significantly. {Commerce’s}differential pricing analysis, including the Cohen’s d test, includes all U.S. sales whichare used to calculate a respondent’s weighted-average dumping margin; therefore,statistical significance, as discussed above, is inapposite. The question is whether there isa practical significance in the differences found to exist in the exporter’s U.S. pricesamong purchasers, regions or time periods. Such practical significance is quantified bythe measure of “effect size.”107

Lastly, in Shrimp from Vietnam, Commerce again pointed to where Dr. Ellis addressedpopulations of data, stating that, “Dr. Ellis also states in his publication that the ‘best way tomeasure an effect is to conduct a census of an entire population but this is seldom feasible inpractice.’”108

There are two separate concepts and measurements when analyzing whether the means of twosets of data are different. The first measurement, when these two sets of data are samples of alarger population, is whether this difference is statistically significant, as measured by a t-test.This will determine whether this difference rises above the sampling error (or in other words,noise or randomness) in selecting the sample. This will answer the question of whether picking asecond (or third or fourth) set of samples will result in a different outcome than the first set ofsamples. When the t-test results in determining that the difference is statistically significant (i.e.,the null hypothesis is false), then these results rise above the sampling error and are statisticallysignificant.

The second measurement is whether there is a practical significance of the difference betweenthe means of the two sets of data, as measured by an “effect size” such as the Cohen’s dcoefficient. As noted above, this measures the real-world relevance of this difference “and maytherefore be said to be a true measure of the significance of the difference.”109 This is the basisfor Commerce’s determination of whether prices in a test group differ significantly from pricesin a comparison group.

SSV claims that Commerce’s use of Cohen’s stated thresholds to determine whether Cohen’smeasurement of effect size is significant is not appropriate. SSV states that:

Professor Cohen himself made clear that his proposed cut-offs could onlyappropriately be applied in specific circumstances —where ‘samples, each of ncases, have been randomly and independently drawn from normal populations,’and where the two samples do not have “substantially unequal variances” or‘substantially unequal sample sizes (whether small or large).’110

107 Id.108 See Shrimp from Vietnam at 17 (citing Guide to Effect Sizes).109 See OCTG Korea Final IDM at Comment 3 (citing Effect Size).110 See SSV Case Brief at 24 (citing OCTG Korea AR1 Final IDM at 22 (citing Statistical Power at 19-20)).

21

We find SSV’s claim to be misplaced. As an initial matter, discussions of sampling, sample size,and statistical inferences are irrelevant to the Cohen’s d test as there is no sampling involved init. There are no estimates of the means and variances of the test and comparison groups.Furthermore, as Commerce noted in OCTG Korea Final, SSV’s quotation is from section 2.1 ofDr. Cohen’s text, “Introduction and Use” of “The T Test for Means.”111 As described above, thisconcerns the statistical significance of the difference in the means for two sampled sets of dataand is not relevant when considering whether this difference has a practical difference. This isnot to say that sample size and sample distribution have no impact on the description of “effectsize” for sampled data,112 but that is not the basis for Commerce’s analysis of SSV’s U.S. salesprice data.

Further, Commerce has noted that the subject of Statistical Power and the discussion therein is“statistical power analysis.”113 Power analysis involves the interrelationship between statisticaland practical significance to attain a specified confidence or “power” in the results of one’sanalysis. Indeed, the beginning of the “Introduction and Use” of “The T Test for Means,”including SSV’s first quotation, is:

The arithmetic mean is by far the most frequently used measure of location bybehavioral scientists, and hypotheses about means the most frequently tested. Thetables have been designed to render very simple the procedure for power analysisin the case where two samples, each of n cases, have been randomly andindependently drawn from normal populations, and the investigator wishes to testthe null hypothesis that their respective population means are equal…114

Again, Commerce is not conducting a “power analysis” which guides researchers in theirconstruction of a project in order to obtain a prescribed “power” (i.e., confidence level, certainty)in the researchers’ results and conclusions. This incorporates a balance between the samplingtechnique, including sample size and potential sampling error, with the stipulated effect size.The Cohen’s d test and Dr. Cohen’s thresholds in these final results only measure thesignificance of the observed differences in the mean prices for the test and comparison groupswith no need to draw statistical inferences regarding sampled price date or the “power” ofCommerce’s results and conclusions.

The 0.8 threshold for the Cohen’s d coefficient, which establishes whether the price differencebetween the test and comparison groups is significant (i.e., the “large” effect size), is arbitrary.Commerce addressed the same argument by the respondent Deosen in Xanthan Gum, stating:

Deosen’s claim that the Cohen’s d test’s thresholds of “small,” “medium,” and“large” are arbitrary is misplaced. In “Difference Between Two Means,” the

111 See OCTG Korea AR3 Final IDM at Comment 3.112 Id. (citing Statistical Power at 21-23, section 2.2.1, where Dr. Cohen quantifies the “nonoverlap” of sampled setsof data. The calculation of the overlap must rely on certain assumptions, such as normal distributions and equalvariances in order to determine the common or non-common overlap of the two datasets).113 Id.114 Id. (citing Statistical Power at 19 (emphasis in italics, SSV’s quotation underlined)).

22

author states that “there is no objective answer” to the question of whatconstitutes a large effect. Although Deosen focuses on this excerpt for theproposition that the “guidelines are somewhat arbitrary,” the author also notes thatthe guidelines suggested by Cohen as to what constitutes a small effect size,medium effect size, and large effect size “have been widely adopted.” The authorfurther explains that Cohen’s d is a “commonly used measure” to “consider thedifference between means in standardized units.” At best, the article may indicatethat although the Cohen’s d test is not perfect, it has been widely adopted. Andcertainly, the article does not support a finding, as Deosen contends, that theCohen’s d test is not a reasonable tool for use as part of an analysis to determinewhether a pattern of prices differ significantly.115

As Commerce explained in the Preliminary Results, the magnitude of the price differences asmeasured with the Cohen’s d coefficient:

… can be quantified by one of three fixed thresholds defined by the Cohen’s dtest: small, medium or large (0.2, 0.5 and 0.8, respectively). Of these thresholds,the large threshold provides the strongest indication that there is a significantdifference between the mean of the test and comparison groups, while the smallthreshold provides the weakest indication that such a difference exists. For thisanalysis, the difference is considered significant, and the sales in the test groupare found to pass the Cohen’s d test, if the calculated Cohen’s d coefficient isequal to or exceeds the large (i.e., 0.8) threshold.116

Commerce has relied on the most conservative of these three thresholds to determine whether thedifference in prices is significant. Dr. Cohen further provided examples which demonstrate the“real world” understanding of the small, medium and large thresholds, where a “large”difference “is represented by the mean IQ difference estimated between holders of the Ph.D.degree and typical college freshmen, or between college graduates and persons with only a 50-50chance of passing an academic high school curriculum. These seem like grossly perceptible andtherefore large differences, as does the mean difference in height between 13- and 18-year-oldgirls…”117 In other words, Dr. Cohen was stating that it is obvious on its face that there aredifferences in intelligence between highly educated individuals and struggling high schoolstudents, and between the height of younger and older teenage girls. Likewise, the “large”threshold is a reasonable yardstick to determine whether prices differ significantly.

115 See Xanthan Gum from the People’s Republic of China: Final Determination of Sales at Less Than Fair Value,78 FR 33351 (June 4, 2013), and accompanying IDM at Comment 3 (quoting Dave Lane, et al., “Effect Size,”Section 2 “Difference Between Two Means”); Certain Activated Carbon from the People’s Republic of China: Final Results of Antidumping Duty Administrative Review; 2011-2012, 78 FR 70533 (November 26, 2013), and accompanying IDM at Comment 4 (quoting same); and Certain Steel Nails from the People’s Republic of China: Final Results of the Fourth Antidumping Duty Administrative Review, 79 FR 19316 (April 8, 2014), and accompanying IDM at Comment 7 (quoting same)).116 Nonetheless, these thresholds, as with the approach incorporated in the differential pricing analysis itself, may bemodified given factual information and argument on the record of a proceeding. See Preliminary Results PDM at14.117 See OCTG Korea AR3 Final IDM at Comment 3 (citing Statistical Power at 27).

23

Therefore, Commerce disagrees with SSV’s arguments that its application of the Cohen’s d testin this review is improper. As a general matter, Commerce finds that the U.S. sales data whichSSV has reported to Commerce constitutes a population. As such, sample size, sampledistribution, and the statistical significance of the sample are not relevant to Commerce’sanalysis. Furthermore, Commerce finds that Dr. Cohen’s thresholds are reasonable, and the useof the “large” threshold is reasonable and consistent with the requirements of section777A(d)(1)(B) of the Act.

Finally, we note that, in the Preliminary Results, we requested that interested parties “presentarguments and justifications in relation to the above-described differential pricing approach usedin these preliminary results, including arguments for modifying the group definitions used in thisproceeding.”118 SSV submitted no factual evidence or argument that these thresholds should bemodified or that any other aspects of the differential pricing analysis should be changed for SSVin this review. Accordingly, SSV’s arguments at this late stage of the review are unsupported bythe record and appear to only convey SSV’s disagreement with the results of Commerce’sapplication of a differential pricing analysis in this review, rather than to truly identify someaspect of this approach which is unreasonable or inconsistent with the statute.

We disagree with SSV’s contention that Commerce has never explained the 33- and 66-percentthresholds used in the ratio test. Specifically, in OCTG from India, we addressed theestablishment of the 33- and 66-percent thresholds as follows:

In the differential pricing analysis, {Commerce} reasonably established a 33percent threshold to establish whether there exists a pattern of prices that differsignificantly. {Commerce} finds that when a third or less of a respondent’s U.S.sales are not at prices that differ significantly, then these significantly differentprices are not extensive enough to satisfy the first requirement of the statute…

Likewise, {Commerce} finds reasonable, given its growing experience ofapplying section 777A(d)(1)(B) of the Act and the application of the A-to-Tmethod as an alternative to the A-to-A method, that when two thirds or more of arespondent’s sales are at prices that differ significantly, then the extent of thesesales is so pervasive that it would not permit {Commerce} to separate the effectof the sales where prices differ significantly from those where prices do not differsignificantly. Accordingly, {Commerce} considered whether, as an appropriatealternative comparison method, the A-to-T method should be applied to all U.S.sales. Finally, when {Commerce} finds that between one third and two thirds ofU.S. sales are at prices that differ significantly, then there exists a pattern of pricesthat differ significantly, and that the effect of this pattern can reasonably beseparated from the sales whose prices do not differ significantly. Accordingly, inthis situation, {Commerce} finds that it is appropriate to address the concern ofmasked dumping by considering the application of the A-to-T method as an

118 See Preliminary Results PDM at 15.

24

alternative to the A-to-A method for only those sales which constitute the patternof prices that differ significantly.119

Although the selection of these thresholds is subjective, Commerce’s stated reasons behind the33- and 66-percent thresholds does not render them arbitrary. In its case brief, SSV suggestsseveral pairs of other possible thresholds but without reasoning or support to argue that thesevalues are more appropriate than those used by Commerce in this review. Likewise, during thecourse of this review, SSV has submitted no factual evidence or argument that these thresholdsshould be modified. Accordingly, SSV’s arguments at this late stage of the review areunsupported by the record and appear only to convey SSV’s disagreement with the results ofCommerce’s application of a differential pricing analysis in this review, rather than to trulyidentify some aspect of this approach which is unreasonable or inconsistent with the statute.

Commerce disagrees, in part, with SSV that “the mere existence of different results is plainlyinsufficient, by itself, to satisfy the statutory requirements”120 whether the A-to-A method canaccount for significant price differences which are imbedded in SSV’s pricing behavior in theU.S. market. Commerce does agree with SSV that this difference is due to the offsetting oflower U.S. prices with higher U.S. prices. When using the A-to-A method, this offsetting occursimplicitly within the average U.S. prices which is compared with normal value, and thisoffsetting occurs explicitly when offset are granted for non-dumped, negative comparisonresults. The A-to-T method, with zeroing, eliminates the masking of dumping by each type ofoffsetting. When the A-to-T method in not used in conjunction with zeroing, it will always yieldidentical results to the results under the A-to-A method.

The difference in the calculated results specifically reveals the extent of the masked, or“targeted,” dumping which is being concealed when applying the A-to-A method.121 Thedifference in these two results is caused by higher U.S. prices offsetting lower U.S. prices wherethe dumping, which may be found on lower priced U.S. sales, is hidden or masked by higherU.S. prices,122 such that the A-to-A method would be unable to account for such differences.123

Such masking or offsetting of lower prices with higher prices may occur implicitly within theaveraging groups or explicitly when aggregating the A-to-A comparison results. Therefore, inorder to understand the impact of the unmasked “targeted dumping,” Commerce finds that the

119 See Final Determination of Sales at Less Than Fair Value and Final Negative Determination of CriticalCircumstances: Certain Oil Country Tubular Goods from India, 79 FR 41981 (July 18, 2014) (OCTG India Final), and accompanying IDM at Comment 1.120 See SSV Case Brief at 31.121 See Koyo Seiko Co., Ltd. v. United States, 20 F. 3d 1156, 1159 (Fed. Cir. 1994) (“The purpose of theantidumping statute is to protect domestic manufacturing against foreign manufacturers who sell at less than fairmarket value. Averaging U.S. prices defeats this purpose by allowing foreign manufacturers to offset sales made atless-than-fair value with higher priced sales. Commerce refers to this practice as ‘masked dumping.’ By usingindividual U.S. prices in calculating dumping margins, Commerce is able to identify a merchant who dumps theproduct intermittently—sometimes selling below the foreign market value and sometimes selling above it. Wecannot say that this is an unfair or unreasonable result.” (internal citations omitted)).122 See SAA at 842.123 See Union Steel v. United States, 713 F. 3d 1101, 1108 (Fed. Cir. 2013) (“{the A-to-A} comparison methodologymasks individual transaction prices below normal value with other above normal value prices within the sameaveraging group.”).

25

comparison of each of the calculated weighted-average dumping margins using the standard andalternative comparison methodologies exactly quantifies the extent of the unmasked “targeteddumping.”

The simple comparison of the two calculated results belies the complexities in calculating andaggregating individual dumping margins (i.e., individual results from comparing EP, or CEP,with NV). It is the interaction of these many comparisons of EP or CEP with NV, and theaggregation of these comparison results, which determine whether there is a meaningfuldifference in these two calculated weighted-average dumping margins. When using the A-to-Amethod, lower-priced U.S. sales (i.e., sales which may be dumped) are offset by higher-pricedU.S. sales. Congress was concerned about offsetting and that concern is reflected in the SAA,which states that “targeted dumping” is a situation where “an exporter may sell at a dumpedprice to particular customers or regions, while selling at higher prices to other customers orregions.”124 The comparison of a weighted-average dumping margin based on comparisons ofweighted-average U.S. prices that also reflects offsets for non-dumped sales, with a weighted-average dumping margin based on comparisons of individual U.S. prices without such offsets(i.e., with zeroing) precisely examines the impact on the amount of dumping which is hidden ormasked by the A-to-A method. Both the weighted-average U.S. price and the individual U.S.prices are compared to an NV that is independent from the type of U.S. price used forcomparison, and the basis for NV will be constant because the characteristics of the individualU.S. sales125 remain constant, whether weighted-average U.S. prices or individual U.S. prices areused in the analysis.

Consider the simple situation where there is a single, weighted-average U.S. price, and thisaverage is made up of a number of individual U.S. sales which exhibit different prices, and thetwo comparison methods under consideration are the A-to-A method with offsets (i.e., withoutzeroing) and the A-to-T method with zeroing.126 The NV used to calculate a weighted-averagedumping margin for these sales will fall into one of five scenarios with respect to the range ofthese different, individual U.S. sale prices:

1) the NV is less than all U.S. prices and there is no dumping;

2) the NV is greater than all U.S. prices and all sales are dumped;

3) the NV is nominally greater than the lowest U.S. prices such that there is a minimalamount of dumping and a significant amount of offsets from non-dumped sales;127

124 See SAA at 842.125 These characteristics include may include such items as product, level-of-trade, time period, and whether theproduct is considered as prime- or second-quality merchandise.126 The calculated results using the A-to-A method with offsets (i.e., no zeroing) and the calculated results using theA-to-T method with offsets (i.e., no zeroing) will be identical. Accordingly, this discussion is effectively betweenthe A-to-T method with offsets and the A-to-T method with zeroing. See SSV Final Analysis Memorandum atAttachment 2, pages 61-63, which identifies the specific calculation results for SSV in these final results. 127 As discussed further below, please note that scenarios 3, 4 and 5 imply that there is a wide enough spreadbetween the lowest and highest U.S. prices so that the differences between the U.S. prices and normal value canresult in a significant amount of dumping and/or offsets, both of which are measured relative to the U.S. prices.

26

4) the NV is nominally less than the highest U.S. prices, such that there is a significantamount of dumping and a minimal amount of offsets generated from non-dumped sales;

5) the NV is in the middle of the range of individual U.S. prices, such that there is both asignificant amount of dumping and a significant amount of offsets generated from non-dumped sales.

Under scenarios (1) and (2), either there is no dumping, or all U.S. sales are dumped such thatthere is no difference between the weighted-average dumping margins calculated using offsets orzeroing and there is no meaningful difference in the calculated results and the A-to-A methodwill be used. Under scenario (3), there is a minimal (i.e., de minimis) amount of dumping, suchthat the application of offsets will result in a zero or de minimis amount of dumping (i.e., the A-to-A method with offsets and the A-to-T method with zeroing both result in a weighted-averagedumping margin which is either zero or de minimis) and which also does not constitute ameaningful difference and the A-to-A method will be used. Under scenario (4), there is asignificant (i.e., non-de minimis) amount of dumping with only a minimal amount of non-dumped sales, such that the application of the offsets for non-dumped sales does not change thecalculated results by more than 25 percent or cause the weighted-average dumping margin to bede minimis, and again there is not a meaningful difference in the weighted-average dumpingmargins calculated using offsets or zeroing and the A-to-A method will be used. Lastly, underscenario (5), there is a significant, non-de minimis amount of dumping and a significant amountof offsets generated from non-dumped sales, such that there is a meaningful difference in theweighted-average dumping margins calculated using offsets or zeroing. Only under the fifthscenario can Commerce consider the use of an alternative comparison method.

Only under scenarios (3), (4) and (5) are the granting or denial of offsets relevant to whetherdumping is being masked, as there are both dumped and non-dumped sales. Under scenario (3),there is only a de minimis amount of dumping, such that the extent of available offsets will onlymake this de minimis amount of dumping even smaller and have no impact on the outcome.Under scenario (4), there exists an above-de minimis amount of dumping, and the offsets are notsufficient to meaningfully change the results. Only with scenario (5) is there an above-deminimis amount of dumping with a sufficient amount of offsets such that the weighted-averagedumping margin will be meaningfully different under the A-to-T method with zeroing ascompared to the A-to-A / A-to-T method with offsets. This difference in the calculated results ismeaningful in that a non-de minimis amount of dumping is now masked or hidden to the extentwhere the dumping is found to be zero or de minimis or to have decreased by 25 percent of theamount of the dumping with the applied offsets.

This example demonstrates that there must be a significant and meaningful difference in U.S.prices in order to resort to an alternative comparison method. These differences in U.S. pricesmust be large enough, relative to the absolute price level in the U.S. market, where not only isthere a non-de minimis amount of dumping, but there also is a meaningful amount of offsets toimpact the identified amount of dumping under the A-to-A method with offsets. Furthermore,the NV must fall within an even narrower range of values (i.e., narrower than the pricedifferences exhibited in the U.S. market) such that these limited circumstances are present (i.e.,scenario (5) above). This required fact pattern, as represented in this simple situation, must then

27

be repeated across multiple averaging groups in the calculation of a weighted-average dumpingmargin in order to result in an overall weighted-average dumping margin which changes to ameaningful extent.

Further, for each A-to-A comparison result which does not result in the set of circumstances inscenario (5), the “meaningfulness” of the difference in the weighted-average dumping marginsbetween the two comparison methods will be diminished. This is because for these A-to-Acomparisons which do not exhibit a meaningful difference with the A-to-T comparisons, therewill be little or no change in the amount of dumping (i.e., the numerator of the weighted-averagedumping margin) but the U.S. sales value of these transactions will nonetheless be included inthe total U.S. sales value (i.e., the denominator of the weighted-average dumping margin). Theaggregation of these intermediate A-to-A comparison results where there is no “meaningful”difference will, thus, dilute the significance of other A-to-A comparison results where there is a“meaningful” difference, which the A-to-T method avoids.

Therefore, we find that the meaningful difference test reasonably fills the gap in the statute toconsider why, or why not, the A-to-A method (or T-to-T method) cannot account for thesignificant price differences in SSV’s pricing behavior in the U.S. market. Congress’s intent ofaddressing “targeted dumping,” when the requirements of section 777A(d)(1)(B) of the Act aresatisfied,128 would be thwarted if the A-to-T method without zeroing were applied since this willalways produce the identical results when the standard A-to-A method without zeroing isapplied. Under that scenario, both methods would inherently mask dumping. It is for this reasonthat we find that the A-to-A method cannot take into account the pattern of prices that differsignificantly for SSV, i.e., Commerce identified conditions where “targeted” or masked dumping“may be occurring” in satisfying the pattern requirement, and Commerce demonstrated that theA-to-A method could not account for the significant price differences, as exemplified by thepattern of prices that differ significantly. Thus, we continue to find that application of the A-to-Tmethod, with zeroing, is an appropriate tool to address masked “targeted dumping.”129

We disagree with SSV that it has failed to satisfy the statutory requirements of section777A(d)(1)(B) of the Act and consider the application of an alternative comparison methodbased on the A-to-T method. As set forth in the Preliminary Results130 and as further discussedin these final results, Commerce’s differential pricing analysis for SSV in this administrativereview is both lawful, reasonable, and completely within Commerce’s discretion in executing thetrade statute.

Comment 4: Financial Statements

In the Preliminary Results, Commerce used the financial statements of six Indian producers ofcomparable merchandise to calculate surrogate financial ratios for purposes of calculating SSV’snormal value. Those financial statements were those of Hi-Tech Pipes Limited (Hi-Tech), JTLInfra Limited (JTL), Rama Steel Tubes Limited (Rama), Spark Electrodes Private Limited

128 See SAA at 842-843.129 See Apex I, 37 F. Supp. 3d at 1296.130 See Preliminary Results PDM at 13-15.

28

(Spark Electrodes), Swastik Pipes Limited (Swastik) and Tamilnadu Steel Tubes Limited(Tamilnadu).131

Domestic Interested Parties’ Case Brief:132

Commerce is required by statute to use the best available information from an ME countrywhen calculating surrogate financial ratios.133 In selecting the best available information,Commerce’s practice is to use publicly available information pertaining to manufacturers ofidentical or comparable merchandise, and to select financial statements based on thespecificity, contemporaneity, and quality of the data.134 In addition, Commerce typicallyexcludes the financial statements of potential surrogate companies that receivedcountervailable subsidies.135 The six financial statements Commerce used in the PreliminaryResults do not meet Commerce’s criteria.o None of the six companies whose financial statements Commerce used in its Preliminary

Results have API-5CT certification, which means they do not produce OCTG.136 Thisdistinction between API-certified OCTG and other tubular products is not insignificant.OCTG generally commands a higher price in the market than other pipe (with theexception of specialized products). Furthermore, Commerce has stated its preference “touse the financial statements of producers of identical merchandise.”137

o The financial statements of Spark Electrodes and Swastik do not appear to be publiclyavailable or otherwise do not constitute the best information available.

In Steel Pipe Vietnam 2016 Final, Commerce rejected Swastik’s financial statementas not publicly available after finding that “there is no realistic way for {Commerce}or the petitioners to replicate the process or otherwise obtain the financial statementsof{Swastik} independently.”138 Furthermore, Swastik’s financial statement ispresented in a template format, similar to those Commerce rejected in Steel PipeVietnam 2016 Final. Notably, it is also similar in format to the financial statementsof Nezone Tubes Limited and Ravindra Tubes Private Limited, both of whichCommerce eliminated from consideration in Steel Pipe Vietnam 2016 Final for notbeing publicly available.139 These financial statements suffer from irregularities that

131 See SV Memorandum at 8 and Exhibit 8.132 See Domestic Interested Parties’ Case Brief at 1-18.133 Id. at 5 (citing section 773(c)(1) of the Act).134 Id. at 7 (citing Preliminary Results PDM at 12 (citing Silicon Metal from the People’s Republic of China: FinalResults and Partial Rescission of Antidumping Duty Administrative Review, 75 FR 1592 (January 12, 2010) (SiliconMetal China Final), and accompanying IDM at Comment 4)).135 Id. at 7-8 (citing Preliminary Results PDM at 12-13 (citing Crystalline Silicon Photovoltaic Cells, Whether orNot Assembled into Modules, from the People’s Republic of China: Final Results of Antidumping DutyAdministrative Review and Final Determination of No Shipments; 2016-2017, 84 FR 36886 (July 30, 2019), andaccompanying IDM at Comment 14).136 Id. at 8 (citing Domestic Interested Parties’ Letter, “Certain Oil Country Tubular Goods from the SocialistRepublic of Vietnam: Rebuttal Comments on Surrogate value Information,” dated April 1, 2019, at Exhibit 1.A).137 Id. at 10 (citing OCTG Vietnam Investigation Final IDM at Comment 2).138 Id. at 11 (citing Circular Welded Carbon-Quality Steel Pipe from the Socialist Republic of Vietnam: FinalDetermination of Sales at Less Than Fair Value, 81 FR 75042 (October 28, 2016) (Steel Pipe Vietnam 2016 Final), and accompanying IDM at Comment 1).139 Id. at 12 (citing Steel Pipe Vietnam 2016 Final IDM at Comment 1).

Preliminary Results

o

o

o

IdId.Id OCTG Vietnam AR1 Final Welded Stainless Pressure Pipe from the Socialist Republic

of Vietnam: Final Determination of Sales at Less Than Fair Value Pressure Pipe Vietnam Final Utility Scale Wind Towers from the Socialist Republic Vietnam: Final Results of Antidumping Duty Administrative Review; 2013–2014

Certain Frozen Warmwater Shrimp from the Socialist Republic of Vietnam: Preliminary Results of Antidumping Duty Administrative Review; 2017–2018

Certain Activated Carbon from the People’s Republic of China: Final Results of Antidumping Duty Administrative Review; 2016–2017 Certain Corrosion-Resistant Steel Products from the People’s Republic of China: Final Determination of Sales at Less Than Fair Value, and Final Affirmative Critical Circumstances Determination, in Part

Id Certain Frozen Fish Fillets from the Socialist Republic of Vietnam: Final Results, and Final Results of No Shipments of the Antidumping Duty Administrative Review 2016–2017

Laminated Woven Sacks from the Socialist Republic of Vietnam: Final Determination of Sales at Less Than Fair Value

Id

30

o Commerce used the financial statement of Surya Roshni in OCTG Vietnam AR1 Final.145

SSV Rebuttal Brief:146

Commerce properly disqualified Ratnamani’s financial statement.o Ratnamani received subsidies under India’s Duty Entitlement Passbook (DEPB) and

Export Packing Credit schemes during the POR.147 Commerce has found both of theseprograms to be countervailable.148

o Commerce has previously held that the mere fact that a company is licensed by the APIto produce OCTG does not prove that it actually produced OCTG.149

o Even if Ratnamani did produce OCTG, its financial statement indicates that that OCTGproduction represented an insignificant portion of its overall operations.150 Furthermore,Ratnamani derives significant revenues from its sale of power generated from wind millsand sale of services.151 Moreover, the carbon steel pipes that Ratnamani producesinclude products that are quite different from the OCTG that SSV produces.152

Commerce properly disqualified Surya Roshni’s financial statement.o Surya Roshni received subsidies under the Export Promotion Capital Goods (EPCG)153

and Investment Promotion Assistance schemes.154 Commerce has found both of theseprograms to be countervailable,155 and because of Surya Roshni’s receipt of EPCG,rejected Surya Roshni’s financial statement in OCTG Vietnam Investigation Final.156

o The Domestic Interested Parties are wrong in suggesting that SSV agreed to the inclusionof Surya Roshni’s financial statement. Although SSV submitted Surya Roshni’s financialstatement, it never argued that Surya Roshni’s financial statement could properly be used.

145 Id. at 16 (citing OCTG Vietnam AR1 Final IDM at Comment 1).146 See SSV Rebuttal Brief at 15-27.147 Id. at 17 (citing Ratnamani’s financial statement at 109 and 151, respectively, found in the Domestic InterestedParties’ SVS at Exhibit 1.E (Ratnamani’s financial statement)).148 Id. (citing Polyethylene Terephthalate Film, Sheet, and Strip from India: Final Results of Countervailing DutyAdministrative Review, 73 FR 75672 (December 12, 2008) (PET Film India Final), and accompanying IDM; Noticeof Final Determination of Sales at Less Than Fair Value: Carbazole Violet Pigment 23 from India, 69 FR 67306(November 14, 2004) (Pigment 23 India Final), and accompanying IDM; Pressure Pipe Vietnam Final IDM atComment 1).149 Id. at 18 (citing OCTG Vietnam Investigation Final IDM at Comment 2; Certain Oil Country Tubular Goodsfrom the Socialist Republic of Vietnam: Final Results of Redetermination Pursuant to Court Remand (May 2,2017)). 150 Id. (citing Ratnamani’s financial statement at 12).151 Id. (citing Ratnamani’s financial statement at 114).152 Id. (citing Ratnamani’s financial statement at 16.)153 Id. at 19 (citing Surya Roshni’s financial statement at 121, found in the Domestic Interested Parties’ SVS atExhibit 1.F (Surya Roshni’s financial statement).154 Id. (citing Surya Roshni’s financial statement at 118 and 121).155 Id. (citing PET Film India Final IDM; Large Diameter Welded Pipe from India: Final AffirmativeCountervailing Duty Determination, 83 FR 56819 (November 14, 2018) (Welded Pipe India Final), andaccompanying IDM).156 Id. at 19-20 (citing OCTG Vietnam Investigation Final IDM at Comment 2).

31

o Commerce did not use Surya Roshni’s financial statement in OCTG Vietnam AR1 Final.It used the financial statement of Surya Global Steel Tubes Limited, which is not thesame company as Surya Roshni.157

The six surrogate financial statements Commerce used in the Preliminary Results wereappropriate. Commerce’s past decisions, as well as those by reviewing courts, make clearthat the surrogate company need not itself produce the specific subject merchandise.158

API-5CT certification has proven to be an unreliable indicator of whether or not a companyactually produces OCTG. Furthermore, the scope language in this case is clear that the scopeis not limited to API-certified OCTG. The scope language defines the scope of the order toinclude an expansive range of products including pipe that does not conform to the APIspecifications.In its investigation, the International Trade Commission found that both line pipe andstructural pipe (which do not carry the API certifications for OCTG) can be substituted forlow-grade OCTG in certain applications.159 In this review, SSV’s exports consisted of onlyone of the lowest grades of OCTG.160 The six surrogate companies that Commerce used inthe Preliminary Results produced line pipe and structural pipe. Some of them produced pipeused in water wells. The Domestic Interested Parties have not explained how such productsdiffer from the low-grade OCTG that SSV exported, or why the line pipe, structural pipe, andwater well pipe that the selected surrogate companies produce are not acceptable substitutesfor OCTG in the sort of undemanding oil wells in which SSV’s OCTG could be used.The Domestic Interested Parties’ contention that the financial statements of Spark Electrodesand Swastik are not publicly available is premised on the assumption that those financialstatements could not be obtained by the public. Commerce rejected a virtually identicalclaim in OCTG Vietnam AR1 Final.161 The Swastik statements to which the DomesticInterested Parties object are in the exact same format as the other statements obtained fromthe Indian Ministry of Industry website to which the Domestic Interested Parties have notobjected. In light of the Domestic Interested Parties’ failure to object to the public nature ofthose other financial statements, there is no basis for its selective complaints.Commerce properly relied on multiple financial statements in order to reflect the broaderexperience of the surrogate industry. Although Commerce has sometimes relied on only oneor two financial statements to calculate surrogate financial ratios when there were only alimited number of usable statements, Commerce has also used financial data from more thanone surrogate company on multiple occasions. Furthermore, Commerce has also stated thatit is its preference to do so in order to “reflect the broader experience of the surrogateindustry.”162 This rationale is in keeping with Commerce’s regulatory mandate of using

157 Id. at 20 (citing OCTG Vietnam AR1 Final IDM at 3).158 Id. at 21 (citing Yantai Xinke Steel Structure v. United States, 36 ITRD 129 (2014); Viraj Forgings, Ltd. v. UnitedStates, 283 F. Supp. 2d 1335, 1347 (CIT 2003); and NTN Bearing Corp. of Am. v. United States, 127 F. 3d 1061,1063 (Fed. Cir. 1997)).159 Id. at 23 (citing Certain Oil Country Tubular Goods from India, Korea, the Philippines, Taiwan, Thailand,Turkey, Ukraine, and Vietnam, Inv. Nos. 701-TA-499-500 and 731-TA-1215-1217 and 1219-1223 (Final), USITCPub. 4489 (Sept. 2014) at II-17).160 Id. (citing SSV December 19, 2018, AQR at 32).161 Id. (citing OCTG Vietnam AR1 Final IDM at Comment 1).162 Id. at 26-27 (citing Fresh Garlic from the People’s Republic of China: Final Results of New Shipper Review, 75FR 61130 (October 4, 2010), and accompanying IDM at Comment 4; Brake Rotors from the People’s Republic of

32

“best available information” by selecting financial statements that are representative of abroad market average.163

Commerce Position:

We agree with the domestic interested parties that it is appropriate to revise our selection of thesurrogate financial statements used to value financial ratios in the Preliminary Results. We donot, however, agree with them that the financial statements of Surya Roshni and Ratnamanirepresent the best information on the record for purposes of calculating surrogate financial ratios.

Commerce’s regulations state that for calculating manufacturing overhead, general expenses, andprofit, Commerce will normally use non-proprietary information gathered from producers ofidentical or comparable merchandise in the surrogate country.164 Furthermore, in choosingsurrogate financial ratios, it is Commerce’s practice to use data from ME surrogate companiesbased on the “specificity, contemporaneity, and quality of the data.”165 Commerce prefers to usethe financial statements of companies who produce identical merchandise.166 Furthermore,Commerce has stated, “It is {Commerce’s} practice to disregard financial statements ofcompanies that evidence receipt of countervailable subsidies and where there are other usabledata on the record.”167 The domestic interested parties argue that the financial statements ofSurya Roshni and Ratnamani meet all of these criteria.

We disagree with the domestic interested parties. Contrary to the domestic interested parties’assertion, the financial statements of Surya Roshni and Ratnamani evidence receipt of subsidiespreviously found to be countervailable. Specifically, Surya Roshni’s financial statement shows itreceived benefits under the Export Promotion Capital Goods scheme168 and the InvestmentPromotion Assistance scheme.169 Commerce has found both of these programs

China: Final Results and Partial Rescission of the Sixth Antidumping Duty Administrative Review and FinalResults of the Ninth New Shipper Review, 69 FR 42039 (July 13, 2004), and accompanying IDM at Comment 2;Final Results of First New Shipper Review and First Antidumping Duty Administrative Review: Certain PreservedMushrooms from the People’s Republic of China, 66 FR 31204 (June 11, 2001), and accompanying IDM atComment 3; and Certain Oil Country Tubular Goods from the People’s Republic of China: Final Determination ofSales at Less Than Fair Value, Affirmative Final Determination of Critical Circumstances and Final Determinationof Targeted Dumping, 75 FR 20335 (April 19, 2010) (OCTG China Final), and accompanying IDM at Comment13).163 Id. at 27 (citing section 773(c) of the Act).164 See 19 CFR 351.408(c)(4).165 See OCTG Vietnam AR1 Final IDM at Comment 1 (citing Silicon Metal China Final IDM at Comment 4).166 See OCTG Vietnam Investigation Final IDM at Comment 2.167 Id. (citing Certain Steel Threaded Rod from the People’s Republic of China: Final Determination of Sales atLess than Fair Value, 74 FR 8907, (February 27, 2009), and accompanying IDM at Comment 1; and OCTG ChinaFinal IDM at Comment 13; Certain Steel Racks and Parts Thereof from the People’s Republic of China: FinalAffirmative Determination of Sales at Less Than Fair Value, 84 FR 35995 (July 24, 2019) (“It is Commerce’spractice to reject the financial statements of a company that we have reason to believe or suspect may have receivedcountervailable subsidies from a program previously investigated by Commerce, particularly when other sufficient,reliable, and representative data are available for calculating surrogate financial ratios”)).168 See Surya Roshni’s financial statement at 121.169 Id. at 118 and 122.

33

countervailable.170 Furthermore, Ratnamani has received benefits under the export packingcredits program,171 which Commerce has also found countervailable.172 Therefore, if otherusable information exists on the record, the financial statements of Surya Roshni and Ratnamaniwould not be the best information available on the record with which to calculate financial ratios.

Upon review of the record, we determine that other usable information does exist on the record.Specifically, most of the companies whose financial statements we used in the PreliminaryResults meet Commerce’s criteria. They are all producers of pipe and tube products, whichconstitute merchandise comparable to OCTG. Their financial statements are alsocontemporaneous to the POR and (unlike the financial statements of Surya Roshni andRatnamani) do not evidence receipt of any subsidies Commerce has found to be countervailable.These factors are all among the criteria Commerce considers when evaluating financialstatements, and the domestic interested parties have not disputed them.

The domestic interested parties have disputed only the criterion of public availability, and onlywith respect to the financial statements of Spark Electrodes and Swastik. However, theinformation on the record does not necessarily demonstrate that these financial statements are notpublicly available. The factors to which the domestic interested parties cite (i.e., the irregularformat of the financial statements, their similarity in format to other financial statementsCommerce has previously found to be not publicly available, the scan notation at the bottom ofthe Spark Electrodes statement) do not compel the conclusion that these financial statementswere not intended for public review or that they are not the best available information on therecord.

Furthermore, that Commerce found the Swastik financial statement to be not publicly availablein Steel Pipe Vietnam 2016 Final is not determinative in this proceeding. The Swastik financialstatement at issue in Steel Pipe Vietnam 2016 Final was not for the same year as the Swastikfinancial statement on the record in this review. Additionally, in Steel Pipe Vietnam 2016 FinalCommerce was dissatisfied with the submitter’s explanation of how it obtained the financialstatement. In this review, unlike Steel Pipe Vietnam 2016 Final, the record does not containinformation as to how the submitter obtained Swastik’s financial statement. Moreover, eventhough the domestic interested parties have cited to a page containing a table that extends beyondthe edges of the page, thus partially cutting off some data, the page at issue does not contain datawhich we would use in our calculation of the financial ratios. Commerce has stated that itspractice is “not to rely on financial statements that are missing significant elements, or which areotherwise deficient, when there are other, more complete financial statements on the record.”173

170 See PET Film India Final IDM at Comments 2-6; Welded Pipe India Final IDM at 10-11 and Comment 2.171 See Ratnamani’s financial statement at 151.172 See PET Film India Final IDM at 4-5; Pigment 23 India Final IDM at Comment 2; Pressure Pipe VietnamFinal IDM at Comment 1; Steel Pipe Vietnam 2016 Final IDM at Comment 1.173 See Tapered Roller Bearings and Parts Thereof, Finished and Unfinished, from the People’s Republic of China: Final Results of Antidumping Duty Administrative Review; 2016-2017, 84 FR 6132 (February 26, 2019), and accompanying IDM at Comment 10 (citing Notice of Final Determination of Sales at Less Than Fair Value: Silicomanganese from Kazakhstan, 67 FR 15535 (April 2, 2002), and accompanying IDM at Comment 3; see alsoWooden Bedroom Furniture from the People’s Republic of China: Final Results of the 2004-2005 Semi-AnnualNew Shipper Reviews, 71 FR 70739 (December 6, 2006), and accompanying IDM at Comment 2, where Commerce

34

The data in the table at issue here concerns Swastik’s corporate social responsibility, an elementthat does not figure into the financial ratio calculations or affect Commerce’s evaluation of thesuitability of the financial statement for calculating financial ratios. Therefore, we do not find itto be a significant element.

Nevertheless, upon further review of Spark Electrodes’ financial statement, we agree with thedomestic interested parties that portions of it are indeed illegible.174 Therefore, because it isCommerce’s practice not to use illegible financial statements,175 and because we have otherusable financial statements on the record, we have not used Spark Electrodes’ financial statementin these final results of review.

However, we do not agree with the domestic interested parties that our use of six financialstatements in the Preliminary Results, rather than only one or two, should raise any concernswith respect to Commerce’s practice. Commerce has stated that it “prefers relying on multiplefinancial statements to derive surrogate ratios in order to obtain broader industryrepresentation.”176 Our use of six financial statements in the Preliminary Results, and of fivefinancial statements in these final results of review, is consistent with this preference.Furthermore, Commerce does not have a maximum number of financial statements that it willuse for calculating surrogate financial ratios, and has used multiple surrogate financial statementsin previous cases.177

The domestic interested parties have also argued that the surrogate financial ratios calculatedfrom the financial statements of Surya Roshni and Ratnamani are similar to the surrogatefinancial ratios used in prior segments of this order. However, we disagree that this is a relevantconsideration in selecting the best information on the record of this review. We enumerated(above) our criteria for selecting the best information available on the record for surrogatefinancial ratios, and numerical consistency with prior reviews of the order is not among thoseconsiderations.

Moreover, we also do not agree with the domestic interested parties that SSV’s having placed thefinancial statement of Surya Roshni on the record of this review necessarily implies that SSVagrees that it is a suitable financial statement for calculating surrogate financial ratios. In itsrebuttal brief, SSV cited a different reason as its motivation for putting Surya Roshni’s financial

stated, “It is {Commerce}’s practice to disregard incomplete financial statements as a basis for calculating surrogatefinancial ratios where ‘...the statement is missing key sections, such as sections of the auditor’s report, that are vitalto our analysis and calculations,’” citing Final Determination of Sales at Less Than Fair Value: Steel ConcreteReinforcing Bars from Belarus, 66 FR 33528 (June 22, 2001), and accompanying IDM at Comment 2).174 See Spark Electrode’s financial statement at 13, 14, 16, and 17.175 See Certain New Pneumatic Off-The-Road Tires from the People’s Republic of China: Final AffirmativeDetermination of Sales at Less Than Fair Value and Partial Affirmative Determination of Critical Circumstances,73 FR 40485 (July 15, 2008), and accompanying IDM at Comment 17A.176 See Certain Helical Spring Lock Washers from the People’s Republic of China: Final Results of AntidumpingDuty Administrative Review, 75 FR 29720 (May 27, 2010), and accompanying IDM at Comment 1.177 See, e.g., Circular Welded Carbon-Quality Steel Pipe from the Socialist Republic of Vietnam: Notice of FinalDetermination of Sales at Less Than Fair Value, 77 FR 64483 (October 22, 2012) (Steel Pipe Vietnam 2012 Final), and accompanying IDM at Comment 2 (in which Commerce used five financial statements to calculate surrogatefinancial ratios).

35

statement on the record. Even if all parties did agree that Surya Roshni’s financial statementwere a suitable financial statement from which to calculate surrogate financial ratios, it is forCommerce to determine the best available information on the record and, for the reasons notedabove, we have not found Surya Roshni’s financial statement to qualify based on our selectioncriteria. Furthermore, we agree with SSV that in the final results of OCTG Vietnam AR1 Final,we used the financial statement of Surya Global Steel Tubes Limited, and not Surya Roshni.178

Accordingly, we determine that the financial statements of Hi-Tech, JTL, Rama, Swastik, andTamilnadu represent the best information available on the record for calculating surrogatefinancial ratios because they meet all of Commerce’s criteria for surrogate values and, unlike thefinancial statements of Surya Roshni and Ratnamani, evince no receipt of subsidies we havepreviously found to be countervailable. Therefore, in these final results of review, we have usedthe financial statements of only these five companies in calculating the surrogate financial ratios.

Comment 5: Particular Market Situation

In the Preliminary Results, Commerce valued HRC using SSV’s purchase prices from MEsuppliers 179 located in Korea.180

Petitioner’s Case Brief181

Commerce has made recent decisions that a particular market situation (PMS) distorts theprice of HRC in the Korean market.182 Commerce should account for this distortion byadjusting the ME purchase prices upward.o Commerce is required under section 773(c)(1) of the Act to use the best available

information to value FOPs. Furthermore, Commerce has broad discretion to determinewhat constitutes the best information available.183 Commerce has determined that theKorean HRC market is distorted by, inter alia, low-priced Chinese imports, domestic andglobal overcapacity, government subsidization, strategic alliances among steelmakers,industrial restructuring, and government control over the Korean electricity market.184

Thus, if Commerce does not adjust the ME purchase prices of HRC to account for thePMS distortion in Korea, it will have failed to comply with the statutory requirement of

178 See OCTG Vietnam AR1 Final IDM at Comment 1.179 See SV Memorandum at 3.180 The petitioner kept the location of the ME supplier proprietary in its case brief, but SSV made it public in itsrebuttal brief. Therefore, we are treating it as public information here.181 See Petitioner’s Case Brief at 1-12.182 Id. at 3 (citing OCTG Korea AR3 Final IDM at Comments 1-A, 1-B, 1-C; Circular Welded Non-Alloy Steel Pipefrom the Republic of Korea: Final Results of Antidumping Duty Administrative Review; 2016-2017, 84 FR 26401(June 6, 2019), and accompanying IDM at Comment 1; Certain Corrosion-Resistant Steel Products from theRepublic of Korea: Final Results of Antidumping Duty Administrative Review; 2016-2017, 84 FR 10784 (March 22,2019), and accompanying IDM at Comment 1; Certain Oil Country Tubular Goods from the Republic of Korea: Preliminary Results of Antidumping Duty Administrative Review; 2017-2018, 84 FR 63615 (November 18, 2019)(OCTG Korea AR4 Prelim) and accompanying PDM at 14-15.183 Id. at 5; and QVD Food Co., Ltd., v. United States, 658 F. 3d 1318, 1323 (Fed. Cir. 2011).184 Id. at 6 (citing OCTG Korea AR4 Prelim PDM at 14-15).

36

using “the best available information” to value factors.185 To make this adjustment,Commerce can adjust SSV’s reported ME purchase prices using the final PMSadjustment rate from OCTG Korea AR3 Final or OCTG Korea AR4 Prelim.

In addition to Korean HRS values being distorted, record evidence shows that Indian HRCdata are also distorted. Specifically, Indian GTA data shows that Korean imports constitutedeighty percent or more of total imports of HRC over the period covered for each of threeHTS numbers under which HRC is shipped.186 Therefore, if Commerce decides to valueHRC using an Indian surrogate value, rather than SSV’s ME purchase prices, it should adjustthe Indian surrogate value for PMS distortion that exists in the Korean market. This is amethodology Commerce recently used in Pipes and Tubes India Prelim.187

SSV Rebuttal Brief188

Commerce has had a longstanding practice of not using purchases from Korea to determinesurrogate values due to Commerce’s findings of general export subsidies in Korea.189 IfCommerce were to follow that practice here, the petitioner’s complaints about a PMS anddistortions would be irrelevant.In accordance with its regulations, Commerce established specific deadlines for parties tosubmit comments on the appropriate surrogate country to value FOP and on proposedsurrogate values in this review.190 The petitioner’s comments and supporting factualinformation in its case brief regarding a PMS in India and Korea (in which it raised the issuefor the first time), are well past the established deadlines. Moreover, although section351.301(c)(2)(v) and (c)(5) provides parties an opportunity to submit information that is notcovered by specific deadlines in Commerce’s questionnaires or other provisions of theregulations, the deadline for such submissions is 30 days before the scheduled date of thepreliminary determination. Therefore, Commerce should reject the petitioner’s commentsand supporting information on PMS as untimely.Commerce may make an adjustment for a PMS only when it uses its constructed valuemethodology for calculating normal value.191 Here, however, Commerce used its FOPmethodology to calculate normal value. Therefore, it must calculate normal value usingeither actual purchase prices from ME suppliers or surrogate value information fromcountries at a comparable level of economic development,192 and may not adjust SSV’s coilprices due to a finding of a PMS in Korea.

185 Id. at 6 (citing section 773(c)(1) of the Act).186 Id. at 9 (citing SSV SVS at Attachment 2-B).187 Id. at 10 (citing Welded Carbon Steel Standard Pipes and Tubes from India: Preliminary Results of AntidumpingDuty Administrative Review; 2017-2018, 84 FR 33916 (July 16, 2019) (Pipes and Tubes India Prelim), andaccompanying IDM at 19-21).188 See SSV Rebuttal Brief at 1-15.189 Id. at 3 (citing Glass Windshield China Final IDM at Comment 5).190 Id. at 4-5 (citing Memorandum, “Certain Oil Country Tubular Goods from the Socialist Republic of Vietnam: Request for Comments re: (1) Economic Development; (2) Surrogate Country, and (3) Surrogate ValueInformation,” dated February 21, 2019, and Memorandum, “Extension of Deadlines for Surrogate Country andSurrogate Value Submissions,” dated February 28, 2019).191 Id. at 7-8 (citing section 773(e) of the Act; and Saha Thai Steel Pipe Public Co. Ltd. v. United States, Slip Op.19-165 at 11 (December 18, 2019)).192 Id. (citing section 773(c) of the Act).

37

Even if Commerce were to consider the petitioner’s PMS allegation, and even if Commercewere to find that a PMS adjustment is allowed under a FOP analysis, the evidence in this caseis not sufficient to justify an adjustment to the prices that SSV paid for purchases of HRCfrom Korean suppliers.o Commerce’s PMS findings in OCTG Korea AR3 Final (to which the petitioner cites) are

currently on appeal before the CIT. The CIT has not yet issued its ruling. However, ithas ruled on Commerce’s findings of a PMS in the first and second administrativereviews of OCTG from Korea, and found them unsupported by sufficient evidence.193

o Commerce’s finding of a PMS in Korea in OCTG Korea AR3 Final was based on anadverse facts available (AFA) finding of subsidies to POSCO, one of the Koreanproducers of HRCs.194 The adjustment Commerce applied in that review was calculatedfrom the POSCO subsidy rate found in the countervailing duty investigation of hot-rolledsteel flat products from Korea for the 2014 calendar year. In the subsequent review forthe 2016 calendar year, Commerce found only a 0.54 percent subsidy rate for POSCOand a 0.58 percent subsidy rate for the other selected respondent for the 2016 calendaryear.195 In these circumstances, Commerce cannot reasonably rely on the findings in pastOCTG cases that relied on the now-outdated AFA subsidy rate for POSCO to determinethat a PMS existed during the POR. Furthermore, relying on an AFA determination dueto the unresponsiveness of an unaffiliated producer that SSV cannot control would beinconsistent with basic principles of fair procedures, and would therefore be unlawful.196

o The past findings of a PMS in Korea were largely based on factual informationconcerning prices for imports of Chinese steel coils into Korea and the consequentresponse by Korean steel producers and the Korean government to meet that competition.In making its PMS determination in OCTG Korea AR3 Final, Commerce explicitly reliedon these interactions.197 However, nothing in Commerce’s determination indicated thatthe alleged distortion would occur in the absence of the specific conditions that werefound to exist in the Korean market. There was no finding that Korean exports of HRCto OCTG producers outside of Korea were distorted.

o The petitioner’s suggestion that Commerce adjust Korean export prices using the PMSadjustment applied in the preliminary results of OCTG Korea AR4 Prelim is contrary toCommerce’s practice. Specifically, Commerce has stated that it would not use rates frompreliminary determinations in other proceedings.198

The arguments the petitioner placed on the record are the same arguments it placed on therecord in the fourth administrative review of the antidumping order of OCTG from Korea.However, in that proceeding the petitioners placed supporting information on the record that

193 Id. (citing NEXTEEL I, 355 F. Supp. 3d at 1364 (CIT), reconsideration denied, 389 F. Supp. 3d 1343 (CIT 2019);NEXTEEL II, 392 F. Supp. 3d at 1297; Hyundai Steel v. United States, Slip Op. 19-148 (CIT November 25, 2019)).194 Id. (citing OCTG Korea AR3 Final IDM at Comment 1-C).195 Id. (citing Certain Hot-Rolled Steel Flat Products from Korea: Final Results of Countervailing DutyAdministrative Review, 2016, 84 FR 28461 (June 19, 2019), as amended by Countervailing Duty Order on CertainHot-Rolled Steel Flat Products from the Republic of Korea: Amended Final Results of the First AdministrativeReview, 84 FR 35604 (July 24, 2019)).196 Id. at 11 (citing SKF USA v. United States, 675 F. Supp. 2d 1264, 1276 (CIT 2009); and Carpenter Tech. v.United States 26 CIT 830, 843 (CIT 2002).197 Id. (citing OCTG Korea AR3 Final IDM at Comment 1-B).198 Id. at 13 (citing Welded Line Pipe from the Republic of Korea: Final Results of Antidumping DutyAdministrative Review and Final Determination of No Shipments; 2016-2017, 84 FR 27762 (June 14, 2019), and accompanying IDM at Comment 3).

38

the respondents rebutted at length. Here, because the petitioner’s arguments for a PMSadjustment to Korean coil prices have never been placed on the record, SSV has not had anyopportunity to respond to the arguments or place information rebutting the petitioner’sanalysis on the record. It would be a fundamental violation of SSV’s rights for Commerce tomake an adjustment based on arguments and information submitted in another case that SVhas not been given an opportunity to rebut.Numerous flaws exist in the petitioner’s argument (based on Pipes and Tubes India Prelim)that Commerce should adjust Indian import statistics if it determines to use such statistics inthe final results:o As noted earlier, the statute does not permit a PMS adjustment when normal value is

based on FOP.o Commerce has expressly stated that it will not rely on preliminary determinations made

in one case as the basis for its final determinations in another.o None of the argument or information relied upon in Pipes and Tubes India Prelim is on

the record of this proceeding. Thus, SSV has not had opportunity to rebut the petitioner’sfactual allegations, and Commerce has no way to assess whether the specific factualcircumstances that were found to exist in Pipes and Tubes India Prelim also apply to thepresent review.

Commerce Position:

In these final results of review, we have valued HRC differently from how we valued it in thePreliminary Results, but we disagree with the petitioner that we should make a PMS adjustmentto the revised calculation of HRC prices.

As a preliminary matter, we agree with SSV that our practice is to not value inputs using MEpurchase prices from Korea. In Structural Steel China Final, Commerce stated:

As we stated in the Preliminary Determination, in accordance with the OmnibusTrade & Competitiveness Act, Commerce continues to apply its long-standingpractice of disregarding surrogate values if it has a reason to believe or suspect thesource data may be subsidized. In this regard, Commerce has previously found thatit is appropriate to disregard such prices from South Korea because we determinedthat South Korea maintains broadly available, non-industry specific export subsidies.Based on the existence of these subsidy programs that were generally available to allexporters and producers in these countries at the time of the POI, Commerce findsthat it is reasonable to infer that all exporters from South Korea may have benefitedfrom these subsidies. This is consistent with past practice, where Commerce hasrejected market economy purchase prices from South Korea.199

199 See Certain Fabricated Structural Steel from the People’s Republic of China: Final Affirmative Determinationof Sales at Less Than Fair Value, 85 FR 5376 (January 30, 2020) (Structural Steel China Final), and accompanyingIDM at Comment 6 (citations omitted.) See also Certain Coated Paper Suitable for High-Quality Print GraphicsUsing Sheet-Fed Presses from the People’s Republic of China: Final Determination of Sales at Less Than FairValue, 75 FR 59217 (September 27, 2010), and accompanying IDM at Comment 17; Final Determination of Salesat Less Than Fair Value: Certain Automotive Replacement Glass Windshields from the People’s Republic of China,67 FR 6482 (February 12, 2002), and accompanying IDM at Comment 1; Notice of Final Determination of Sales at

39

Therefore, consistent with our decision in Structural Steel China Final, in these final results ofreview we have determined not to value HRC using SSV’s ME purchase prices from Korea aswe did in the Preliminary Results. Thus, the petitioner’s arguments for making a PMSadjustment because of Commerce’s previous finding of the existence of a PMS in Korea forHRC, as well as SSV’s arguments against making a PMS adjustment, are moot.

Instead, consistent with our practice of using the best information available on the record tovalue factors of production in NME proceedings, we have looked for alternative surrogate valueinformation on the record. In evaluating such data, Commerce’s practice, in accordance withsection 773(c)(1) of the Act, is to select, to the extent practicable, SVs which are product-specific, representative of a broad-market average, publicly available, contemporaneous with thePOR, and tax and duty exclusive.200

On March 26, 2019, SSV placed GTA Indian import data on the record for all inputs.201 As thepetitioner placed no alternative data on the record, the Indian GTA data is the only data on therecord with which to value HRC. The petitioner has argued that if Commerce values HRC usingan Indian surrogate value, it should make a PMS adjustment to the surrogate value consistentwith Pipes and Tubes India Prelim.202 However, the petitioner has not placed on the record anydata with which Commerce can perform such an adjustment. The petitioner instead cites to adocument not on the record of this proceeding that they state contains an adjustment factor thatCommerce could use to adjust the Indian data for a PMS in India for HRC.203 However, as thedocument to which the petitioner cites is not on the record of this proceeding, we are unable touse any data allegedly found in it to make an adjustment to the HRC surrogate value in thisreview.

If the petitioner believes that Korea HRC prices are not suitable for valuing SSV’s purchases ofHRC because of the existence of a PMS in Korea, it could, as an alternative, have placed ontothe record of this proceeding surrogate value data for HRC that is not affected by a PMS (e.g.,data concerning an Indian product similar to HRC for which no PMS exists, or HRC data from asecondary surrogate country). The petitioner did not do so, and it is now well past the deadlinefor submission of data with which to value factors of production.204

Section 773(c)(1) of the Act states, “the valuation of the factors of production shall be based onthe best available information regarding the values of such factors in a market economy country

Less Than Fair Value: Lawn and Garden Steel Fence Posts from the People’s Republic of China, 68 FR 20372(April 25, 2003), and accompanying IDM at Comment 2.200 See Structural Steel China Final IDM at Comment 6.201 See SSV SVS at Attachment 2.202 SSV’s argument that we should not make a PMS adjustment to Indian surrogate values because the Indian PMSdetermination was only preliminary is now moot because in January 2020, after parties in this proceeding submittedtheir case and rebuttal briefs, Commerce made its final determination that a PMS for HRC exists in India. SeeWelded Carbon Steel Standard Pipes and Tubes from India: Final Results of Antidumping Duty AdministrativeReview; 2017-2018, 85 FR 2715 (January 16, 2020), and accompanying IDM at Comment 1.203 See Petitioner’s Case Brief at 11, n. 27.204 See Memorandum, “Extension of Deadlines for Surrogate Country and Surrogate Value Submissions,” datedFebruary 28, 2019.

40

or countries considered to be appropriate by the administering authority.” As noted above, SSVplaced GTA data on the record for all factors of production, including HRC. Commerce hasstated that GTA data “is a source that is regularly used by Commerce because the data thereinmeet Commerce’s {surrogate value} criteria.”205 Therefore, because the GTA data meet oursurrogate value criteria, and there is no alternate information on the record with which to valueHRC, and also no adjustment factor on the record with which to adjust the GTA data for a PMS,we find the unadjusted Indian GTA data to constitute the best available information with whichto value HRC, and have done so in these final results of review. Additionally, consistent withour valuation of HRC, in these final results of review we have also valued varnish and diluent(which we valued in the Preliminary Results using SSV’s Korean-market purchase prices) usingGTA import data for India.206

Comment 6: Ministerial ErrorsSSV Case Brief:207

Commerce erred in the Preliminary Results by adding a surrogate value for B&H, which wasdenominated in dollars per metric ton, to the inputs diluent and varnish, which SSV reportedon a per-kilogram basis. Thus, Commerce overstated the B&H costs for these inputs.208

Commerce erred in the Preliminary Results by adding a surrogate truck freight value, whichwas denominated in Indian rupees per metric ton per kilometer, to the packing inputs steelband and steel clip, for which the surrogate values (which were based on GTA import data)were denominated in Indian rupees per kilogram.209 Therefore, Commerce overstated thetrucking costs for these inputs.

No other party commented on these issues.

Commerce Position:

The first ministerial error that SSV has noted is moot because, as explained in response toComments 1 and 5 (above), in these final results of review we did not value inputs purchasedfrom Korea using ME purchase prices, but instead valued them using surrogate values. See SSVFinal Analysis Memorandum for further information.

With respect to the second alleged ministerial error, we agree with SSV, and have corrected thiserror in these final results of review.

205 See Chlorinated Isocyanurates from the People’s Republic of China: Final Results of Antidumping DutyAdministrative Review; 2017-2018, 85 FR 10411 (February 24, 2020), and accompanying IDM at Comment 3.206 See SSV Final Analysis Memorandum for details.207 See SSV Case Brief at 5-7.208 Id. at 6 (citing SV Memorandum at 7).209 Id. (citing Preliminary Determination PDM at 6; and Memorandum, “Analysis for the Preliminary Results of theFourth Antidumping Duty Administrative Review of Certain Oil Country Tubular Goods from the SocialistRepublic of Vietnam; SeAH Steel VINA Corporation,” dated November 8, 2019, at 3).

41

VI. RECOMMENDATION

Based on our analysis of the comments received, we recommend adopting all of the abovepositions and adjusting the margin calculation program for SSV accordingly. If accepted, wewill publish the final results of review and the final dumping margins in the Federal Register.

____________ _____________Agree Disagree

X

Signed by: JEFFREY KESSLERJeffrey I. KesslerAssistant Secretaryfor Enforcement and Compliance