Goodluck India Ltd. v. United States , 2019 CIT 110 ( 2019 )


Menu:
  •                                           Slip Op. 19-
    UNITED STATES COURT OF INTERNATIONAL TRADE
    GOODLUCK INDIA LIMITED,
    Plaintiff,
    v.
    UNITED STATES,
    Defendant,                Before: Gary S. Katzmann, Judge
    Court No. 18-00162
    and
    ARCELORMITTAL TUBULAR
    PRODUCTS, MICHIGAN SEAMLESS
    TUBE, LLC, PLYMOUTH TUBE CO. USA,
    PTC ALLIANCE CORP., WEBCO
    INDUSTRIES, INC. AND ZEKELMAN
    INDUSTRIES, INC,
    Defendant-Intervenors.
    OPINION
    [Plaintiff’s motion for judgment on the agency record is granted. The court remands to Commerce
    for further proceedings consistent with this opinion.]
    Dated:$XJXVW
    Ned H. Marshak and Michael S. Holton, Grunfeld, Desiderio, Lebowitz, Silverman & Klestadt,
    LLP, of New York, NY and Washington, DC, argued for plaintiff. With them on the brief were
    Andrew T. Schutz, Kavita Mohan, and Jordan C. Kahn.
    Ann C. Motto, Trial Attorney, Commercial Litigation Branch, Civil Division, U.S. Department of
    Justice, of Washington, DC, argued for defendant. With her on the brief were Joseph H. Hunt,
    Assistant Attorney General, Jeanne E. Davidson, Director, and Claudia Burke, Assistant Director.
    Of counsel was Caroline D. Bisk, Office of the Chief Counsel for Trade Enforcement &
    Compliance, Office of the General Counsel, U.S. Department of Commerce, of Washington, DC.
    Court No. 18-00162                                                                            Page 2
    Melissa M. Brewer and Kathleen M. Cusack, Kelley Drye & Warren, LLP, of Washington, DC,
    argued for defendant-intervenors. With them on the brief were R. Alan Luberda and David C.
    Smith.
    Katzmann, Judge: This case turns on distinguishing correctible importer mistakes from
    submissions of untimely new factual information. Before the court is whether the Department of
    Commerce (“Commerce”) abused its discretion by rejecting as untimely plaintiff Goodluck India
    Limited’s (“Goodluck”) corrections to information submitted as part of a less than fair value
    investigation on carbon and alloy steel from India and by subsequently relying on other sources of
    information to complete the factual record.
    As part of the investigation, Goodluck submitted sales data on cold-drawn mechanical
    tubing sold in both its home market of India and in the United States and applied particular product
    characteristic codes to the underlying data.      After Goodluck had begun preparing its data,
    Commerce revised its product characteristic coding guidance and extended Goodluck’s
    submission deadline. When Goodluck submitted its responses to Commerce, it had failed to revise
    the coding for 682 home market sales, resulting in cascading errors in Goodluck’s home sales and
    cost databases. Goodluck alerted Commerce to its errors -- which it characterized as correctible
    minor errors -- on the first day of verification, but Commerce rejected Goodluck’s updated
    submissions as untimely new factual information. Consequently, Commerce rejected all of
    Goodluck’s submitted data and issued Goodluck a final dumping margin of 33.80 percent based
    on total adverse facts available. Commerce also used another respondent’s export subsidy rate to
    calculate Goodluck’s export subsidy cash deposit offset, rather than the rate specifically calculated
    for Goodluck in the companion countervailing duty investigation. Goodluck appeals Commerce’s
    determination on each issue to this court.
    Court No. 18-00162                                                                           Page 3
    The court concludes that Commerce’s decision to reject Goodluck’s corrections was an
    abuse of discretion and remands to Commerce to consider Goodluck’s corrected submission as
    well as to explain why it has departed from its general practice for calculating cash deposit offset
    rates in this case.
    BACKGROUND
    I. Legal Background
    Dumping occurs when a foreign company sells a product in the United States for less than
    fair value. Sioux Honey Ass’n v. Hartford Fire Ins. Co., 
    672 F.3d 1041
    , 1046 (Fed. Cir. 2012).
    Similarly, a foreign country may artificially lower a product’s price through subsidies. U.S. Steel
    Grp. v. United States, 
    96 F.3d 1352
    , 1355 n.1 (Fed. Cir. 1996). To ameliorate distortions caused
    by these economic practices, Congress enacted the Tariff Act of 1930 (“Act”), which empowers
    Commerce to investigate potential dumping or subsidies, and if appropriate, issue orders imposing
    duties on the subject merchandise. Sioux Honey Ass’n, 
    672 F.3d at
    1046–47. These antidumping
    and countervailing duty actions are intended to be remedial, not punitive in nature, Chaparral Steel
    Co. v. United States, 
    901 F.2d 1097
    , 1103 (Fed. Cir. 1990), and it is Commerce’s duty to determine
    margins as accurately as possible, Rhone Poulenc, Inc. v. United States, 
    899 F.2d 1185
    , 1191 (Fed.
    Cir. 1990).
    A. Minor Corrections
    “Although Commerce has authority to place documents in the administrative record that it
    deems relevant, the burden of creating an adequate record lies with interested parties and not with
    Commerce.” Nan Ya Plastics Corp. v. United States, 
    810 F.3d 1333
    , 1337 (Fed. Cir. 2016)
    (quoting QVD Food Co. v. United States, 
    658 F.3d 1318
    , 1324 (Fed. Cir. 2011)). Commerce’s
    Court No. 18-00162                                                                                   Page 4
    regulations address submissions of new factual information1 by parties to an investigation, with
    the type of factual information determining the time limit for submission to Commerce under 
    19 C.F.R. § 351.301
    (c). Pertinent here, 
    19 C.F.R. § 351.301
    (c)(5) requires that miscellaneous new
    factual information must be submitted either 30 days before the scheduled date of the preliminary
    results, or 14 days before verification, whichever is earlier.2
    1
    
    19 C.F.R. § 351.102
    (b)(21) provides that:
    “Factual information” means:
    (i) Evidence, including statements of fact, documents, and data submitted
    either in response to initial and supplemental questionnaires, or, to rebut,
    clarify, or correct such evidence submitted by any other interested party;
    (ii) Evidence, including statements of fact, documents, and data submitted
    either in support of allegations, or, to rebut, clarify, or correct such evidence
    submitted by any other interested party;
    (iii) Publicly available information submitted to value factors under
    § 351.408(c) or to measure the adequacy of remuneration under
    § 351.511(a)(2), or, to rebut, clarify, or correct such publicly available
    information submitted by any other interested party;
    (iv) Evidence, including statements of fact, documents and data placed on
    the record by the Department, or, evidence submitted by any interested
    party to rebut, clarify or correct such evidence placed on the record by
    the Department; and
    (v) Evidence, including statements of fact, documents, and data, other
    than factual information described in paragraphs (b)(21)(i)-(iv) of this
    section, in addition to evidence submitted by any other interested party to
    rebut, clarify, or correct such evidence.
    2
    
    19 C.F.R. § 351.301
    (c)(5) provides that:
    Factual information not directly responsive to or relating to paragraphs (c)(1)-(4)
    of this section: Paragraph (c)(5) applies to factual information other than that
    described in § 351.102(b)(21)(i)-(iv). The Secretary will reject information filed
    under paragraph (c)(5) that satisfies the definition of information described in
    § 351.102(b)(21)(i)-(iv) and that was not filed within the deadlines specified above.
    All submissions of factual information under this subsection are required to clearly
    explain why the information contained therein does not meet the definition of
    factual information described in § 351.102(b)(21)(i)-(iv), and must provide a
    Court No. 18-00162                                                                             Page 5
    Apart from new factual information, “Commerce is free to correct any type of importer
    error—clerical, methodology, substantive, or one in judgment—in the context of making an
    antidumping duty determination, provided that the importer seeks correction before Commerce
    issues its final results and adequately proves the need for the requested corrections.” Timken U.S.
    Corp. v. United States, 
    434 F.3d 1345
    , 1353 (2006). 
    19 C.F.R. § 351.224
    (f) provides a definition
    of “ministerial error.”3 Additionally, this court has held that Commerce abuses its discretion by
    rejecting “corrective information,” which includes submissions “to correct information already
    provided [to Commerce],” Fischer S.A. Comercio v. United States, 
    34 CIT 334
    , 348, 
    700 F. Supp. 2d 1364
    , 1376 (2010), or to “clarif[y] information already in the record,” 
    id. at 1373
    , but not to
    “fill [ ] gap[s] caused by [a respondent’s] failure to provide a questionnaire response or evidence
    requested during verification,” 
    id. at 1377
    . Notably, no regulation addresses the circumstances
    under which corrections will be accepted or a time frame within which corrections should be
    submitted. Deacero S.A.P.I de C.V. v. United States, 42 CIT __, __, 
    353 F. Supp. 3d 1303
    , 1307
    (2018); see also Timken, 
    434 F.3d at 1353
     (noting that it “appears that Commerce has not issued
    any regulation addressing whether an importer can correct errors in the information it has
    submitted, [nor] restricted the types of importer errors that are eligible for such correction.”).
    detailed narrative of exactly what information is contained in the submission and
    why it should be considered. The deadline for filing such information will be 30
    days before the scheduled date of the preliminary determination in an investigation,
    or 14 days before verification, whichever is earlier, and 30 days before the
    scheduled date of the preliminary results in an administrative review, or 14 days
    before verification, whichever is earlier.
    3
    
    19 C.F.R. § 351.224
    (f) provides that “under this section, ministerial error means an error in
    addition, subtraction, or other arithmetic function, clerical error resulting from inaccurate copying,
    duplication, or the like, and any other similar type of unintentional error which
    the Secretary considers ministerial.”
    Court No. 18-00162                                                                            Page 6
    While Commerce is afforded discretion in deciding whether to accept a respondent’s
    corrective information, Deacero, 353 F. Supp. 3d at 1309, if “Commerce acted differently in this
    case than it has consistently acted in similar circumstances without reasonable explanation, then
    Commerce’s actions will have been arbitrary,” Consolidated Bearings Co. v. United States, 
    348 F.3d 997
    , 1007 (Fed. Cir. 2003) (citing RHP Bearings v. United States, 
    288 F.3d 1334
    , 1347 (Fed.
    Cir. 2002)). Moreover, “Commerce abuse[s] its discretion [when it] refus[es] to accept updated
    data when there [i]s plenty of time for Commerce to verify or consider it.” Papierfabrik August
    Koehler SE v. United States, 
    843 F.3d 1373
    , 1384 (Fed. Cir. 2016) (citing NTN Bearing Corp. v.
    United States, 
    74 F.3d 1204
    , 1208–09 (Fed. Cir. 1995) (requiring correction of typing errors) and
    Timken, 
    434 F.3d at 1353
     (expanding the holding in NTN to “any type of importer error—clerical,
    methodology, substantive, or one in judgment—... provided that the importer seeks correction
    before Commerce issues its final results and adequately proves the need for the requested
    corrections”)). While “a tension may arise between finality and correct result” at later stages of
    an investigation, “[p]reliminary determinations are ‘preliminary’ precisely because they are
    subject to change . . . the tension between finality and correctness [does] not exist at th[at] time.”
    NTN, 
    74 F.3d at 1208
    .
    B. Reliance on facts otherwise available and adverse facts available
    If a party fails to satisfactorily respond to Commerce’s requests for “necessary
    information” to calculate a dumping margin by withholding requested information, failing to
    provide information by the submission deadlines or in the form or manner requested, significantly
    impeding a proceeding, or providing information that cannot be verified, Commerce shall use facts
    otherwise available to calculate the margin. 19 U.S.C. § 1677e(a)(2). “The use of facts otherwise
    available . . . is only appropriate to fill gaps when Commerce must rely on other sources of
    information to complete the factual record.” Zhejiang Dunan Hetian Metal Co. v. United States,
    Court No. 18-00162                                                                           Page 7
    
    652 F.3d 1333
    , 1346 (Fed. Cir. 2011) (citing Nippon Steel Corp. v. United States, 
    337 F.3d 1373
    ,
    1381 (Fed. Cir. 2003)).
    Under 19 U.S.C. § 1677e(b)(1)(A), Commerce may apply adverse inferences as facts
    available (“AFA”) when Commerce “finds that an interested party has failed to cooperate by not
    acting to the best of its ability to comply with a request for information[.]” A respondent does not
    cooperate to the “best of its ability” when it fails to “put forth its maximum effort to provide
    Commerce with full and complete answers to all inquiries.” Dillinger France S.A. v. United States,
    42 CIT __, __, 
    350 F. Supp. 3d 1349
    , 1360 (2018) (quoting Nippon Steel, 
    337 F.3d at 1382
    ).
    “[W]here there is useable information of record but the record is incomplete,” Commerce applies
    partial AFA. Wash. Int’l Ins. Co. v. United States, 
    33 CIT 1023
    , 1035 n.18 (2009) (citing Yantai
    Timken Co., Ltd. v. United States, 
    31 CIT 1741
    , 1746–48, 
    521 F. Supp. 2d 1356
    , 1364–65 (2007),
    aff’d 
    300 Fed. Appx. 934
     (Fed. Cir. 2008)). In contrast, when “none of the reported data is reliable
    or usable,” Mukand, Ltd. v. United States, 
    767 F.3d 1300
    , 1305, 1307–08 (Fed. Cir. 2014), that is,
    when it “exhibit[s] pervasive and persistent deficiencies that cut across all aspects of the data,”
    Zhejiang, 
    652 F.3d at 1348
    , Commerce applies total AFA.
    C. Export subsidy cash deposit offset determinations
    “If Commerce issues a final determination that subject merchandise is being, or is likely to
    be sold in the United States at less than fair value, Commerce orders the posting of a cash deposit
    for each entry of the subject merchandise based on the estimated weighted average dumping
    margin.” Jinko Solar Co. v. United States, 41 CIT __, 
    229 F. Supp. 3d 1333
    , 1359 (2017) (citing
    19 U.S.C. §§ 1673d(a)(1), 1673d(c)(1)(B)(i)–(ii)). “The price used to establish export price and
    constructed export price shall be increased by the amount of any countervailing duty imposed on
    the subject merchandise . . . to offset an export subsidy.” 19 U.S.C. § 1677a(c)(1)(C). “Neither
    the statute nor Commerce’s regulations otherwise define how the cash deposit is to be calculated
    Court No. 18-00162                                                                                Page 8
    in an investigation,” and so “Commerce has discretion to establish a reasonable practice to
    calculate a cash deposit rate in investigations where there is no clear statutory directive.” Jinko,
    229 F. Supp. 3d at 1358 (citing United States v. Eurodif S.A., 
    555 U.S. 305
    , 316 (2009)).
    II. Factual Background and Procedural History
    On May 16, 2017, Commerce initiated less than fair value investigations on carbon and
    alloy steel from India, see Initiation of Less-Than-Fair-Value Investigations, 
    82 Fed. Reg. 22,491
    (May 16, 2017), Public Record (“P.R.”) 32, following a petition submitted on April 19, 2017 by
    ArcelorMittal Tubular Products, Michigan Seamless Tube, LLC, PTC Alliance Corp., Webco
    Industries, Inc., and Zekelman Industries, Inc. 
    Id.
    On June 19, 2017, Commerce designated Goodluck as a mandatory respondent.4 See U.S.
    Department of Commerce Questionnaire to Goodluck (June 19, 2017) (“ADD Questionnaire”),
    P.R. 81. Commerce solicited data from Goodluck via questionnaire on every sale of subject
    merchandise Goodluck made in its home market of India (“Section B”), to the United States
    (“Section C”), and product specific costs (“Section D”) during the period of investigation from
    4
    In antidumping duty investigations or administrative reviews, Commerce may select mandatory
    respondents pursuant to 19 U.S.C. § 1677f-1(c)(2), which provides:
    If it is not practicable to make individual weighted average dumping margin
    determinations [in investigations or administrative reviews] because of the large
    number of exporters or producers involved in the investigation or review, the
    administering authority may determine the weighted average dumping margins for
    a reasonable number of exporters or producers by limiting its examination to—
    (A) a sample of exporters, producers, or types of products that is statistically
    valid based on the information available to the administering authority at
    the time of selection, or
    (B) exporters and producers accounting for the largest volume of the subject
    merchandise from the exporting country that can be reasonably examined.
    Court No. 18-00162                                                                           Page 9
    April 1, 2016 to March 31, 2017. Id. Field 2.0 of the questionnaire instructed Goodluck to create
    and report a control number, or “CONNUM,” for “‘each unique product’ reported in the sales and
    cost data files,” using product characteristic-specific codes provided by Commerce.5 Id. at 109,
    136.
    When Commerce first provided the questionnaire to Goodluck on June 19, 2017,
    Commerce had not yet determined how certain physical characteristics should be coded for the
    purposes of constructing the 24-digit CONNUMs. See U.S. Department of Commerce CONNUM
    Letter (July 6, 2017), P.R. 95 (“First CONNUM Letter”). On July 6, 2017, Commerce issued to
    Goodluck the product characteristic codes via letter. Id. At that time, Goodluck’s Section B, C,
    and D responses were due to Commerce on August 7, 2017. Id.
    Field 2.5 required Goodluck to provide nominal wall thickness in millimeters for each tube
    Goodluck sold in both its home market of India in its Section B questionnaire response and in the
    United States in Section C. Id. at Attach. 1. Field 3.5 required Goodluck to apply a two-digit code
    to the numeric wall thickness reported in Field 2.5 according to thickness ranges provided by
    Commerce in its July 6, 2017 letter. Id. The two-digit codes reported in Field 3.5 were then
    combined with codes for additional physical characteristics contained in other fields to create the
    24-digit CONNUMs reported in Field 2.0. See ADD Questionnaire at 109, 136. Goodluck
    reported weighted average CONNUM-specific costs and used CONNUM-specific expenses to
    create the Section D database. Id. at 162.
    5
    Commerce instructed that “identical products should be assigned the same control number in
    each record in every file in which the product is referenced (e.g., products with identical physical
    characteristics reported in the foreign market sales file and the U.S. market sales file should have
    the same control number.” ADD Questionnaire at 109, 136.
    Court No. 18-00162                                                                        Page 10
    On July 12, 2017, Petitioners submitted comments to Commerce contesting the
    correspondence of certain product characteristics to the Commerce-devised codes.               See
    Petitioners’ CONNUM Comments (July 12, 2017), P.R. 97. Specifically, Petitioners asserted that
    Commerce’s wall thickness ranges “were too broad to accurately capture cost and price differences
    between products” and would undermine cost of production and product matching. Id. at 4–5.
    Following Petitioners’ comments, Commerce sent Goodluck a Revised Product Characteristics
    Letter on August 7, 2017. See U.S. Department of Commerce Revised Product Characteristics
    Letter (Aug. 7, 2017), P.R. 153 (“Revised CONNUM Letter”). This letter modified and increased
    the number of coding ranges for Goodluck to apply to the nominal wall thickness in Field 2.5 to
    create the two-digit code reported in Field 3.5. See id. at Attach. 1; see also First CONNUM Letter
    at Attach. 1. On August 25, 2017, Goodluck timely submitted its questionnaire responses for
    Sections B, C, and D to Commerce, see Goodluck Sections BC&D Questionnaire Response (Aug.
    25, 2017), P.R. 165–68, C.R. 62–131, reporting costs for 385 CONNUMs of subject merchandise,
    see, e.g., id. at 788–95.
    On September 20, 2017 and October 2, 2017, Commerce asked Goodluck for additional
    supplemental information on Section D and Sections B and C, respectively. See U.S. Department
    of Commerce Section D Supplemental Questionnaire to Goodluck (Sept. 20, 2017), P.R. 190, C.R.
    168; see also U.S. Department of Commerce Section B-C Supplemental Questionnaire to
    Goodluck (Oct. 2, 2017), P.R. 196, C.R. 170. Goodluck submitted its responses to the Section B
    and C supplemental questionnaires on October 20 and 23, 2017. See Goodluck Supplemental
    Section B-C Questionnaire Responses (Oct. 20, 2017), P.R. 220, C.R. 207–11; see also Goodluck
    Supplemental Section B-C Questionnaire Responses (Additional Question) (Oct. 23, 2017), P.R.
    222, C.R. 220–21. In these exchanges, neither Commerce nor Goodluck raised the manner in
    Court No. 18-00162                                                                       Page 11
    which Goodluck had reported nominal wall thickness in Field 2.5 or coded the wall thickness in
    Field 3.5.
    On November 22, 2017, Commerce issued its preliminary determination finding that
    Goodluck’s dumping margin was de minimis. See Certain Cold-Drawn Mechanical Tubing of
    Carbon and Alloy Steel From India: Preliminary Affirmative Determination of Sales at Less Than
    Fair Value, in Part, Postponement of Final Determination, and Extension of Provisional Measures,
    
    82 Fed. Reg. 55,567
     (Nov. 22, 2017), P.R. 240. On November 20, 2017, Petitioners filed a
    ministerial error allegation requesting Commerce to correct a clerical error in the coding of two
    steel products sold in the home market. See Petitioners Ministerial Error Allegation (Nov. 20,
    2017), P.R. 238, C.R. 238. On November 27, 2017, Goodluck filed a ministerial error letter asking
    Commerce to correct this mistake. See Goodluck Ministerial Error Letter (Nov. 27, 2017), P.R.
    249, C.R. 242. Commerce agreed to do so on January 3, 2018, and revised Goodluck’s preliminary
    dumping margin to 4.2 percent. See U.S. Department of Commerce Amended Preliminary Margin
    Calculation Memo (Jan. 3, 2018), P.R. 276, C.R. 380–84.
    On November 22 and 27, 2017, Commerce sent its Sales Verification Agenda-Outline and
    Cost Verification Agenda-Outline, respectively, to Goodluck, instructing:
    [V]erification is not intended to be an opportunity for submission of new factual
    information. New information will be accepted at verification only when: (1) the
    need for that information was not evident previously; (2) the information makes
    minor corrections to information already on the record; or (3) the information
    corroborates, supports, or clarifies information on the record.
    See U.S. Department of Commerce Sales Verification Agenda-Outline for Goodluck (Nov. 22,
    2017), P.R. 244, C.R. 240; see also U.S. Department of Commerce Cost Verification Agenda-
    Outline for Goodluck (Nov. 27, 2017), P.R. 246, C.R. 241.
    Court No. 18-00162                                                                        Page 12
    While preparing for verification, Goodluck discovered that it had reported incorrect codes
    in Field 3.5 for 682 observations in its home market sales database provided in its Section B
    questionnaire. See Goodluck Sales Verification Exhibits (Dec. 12, 2017), P.R. 269, C.R. 290–93.
    The incorrect codes in Field 3.5 reflected Commerce’s coding instructions provided in the original
    July 6, 2017 letter, rather than the updated coding guidance contained in Commerce’s August 7,
    2017 letter. 
    Id.
     at Exhibit 16.
    On the first morning of Sales Verification on December 4, 2017, Goodluck informed
    Commerce of its error and provided Commerce with a worksheet identifying the erroneous data
    entries for Field 3.5, which presented the incorrectly reported two-digit product characteristic
    codes side-by-side with the corrected codes. See Pl.’s Br. at Attach. One; see also Cold-Drawn
    Mechanical Tubing of Carbon and Alloy Steel From India: Final Affirmative Determination of
    Sales at Less than Fair Value (“Final Determination”), 
    83 Fed. Reg. 16,296
     (Apr. 16, 2018), P.R.
    317, accompanying Issues and Decision Memorandum at 6–8 (Apr. 9, 2018) (“IDM”), P.R. 311.
    On the first morning of Cost Verification on December 11, 2017, Goodluck again directed
    Commerce to the errors. IDM at 6–8. Commerce’s Cost Verification Report noted “[c]orrection
    of these errors would result in changes to the physical characteristics of 24 CONNUMs and the
    addition of 13 new CONNUMs” that were not included in Goodluck’s original Section D cost
    database submission. See U.S. Department of Commerce Cost Verification Report (Jan. 18, 2018),
    P.R. 278, C.R. 385.
    On February 15, 2018, both Goodluck and Petitioners filed Case Briefs with Commerce.
    See Goodluck Case Brief (Feb. 15, 2018), P.R. 288, C.R. 389; see also Petitioners’ Case Brief
    (Feb. 15, 2018), P.R. 291, C.R. 392. Commerce rejected Goodluck’s case brief on February 20,
    2018 for including untimely new factual information on Goodluck’s identified questionnaire
    Court No. 18-00162                                                                      Page 13
    errors. See U.S. Department of Commerce Letter re: Rejection of New Factual Information (Feb.
    20, 2018), P.R. 295. Goodluck refiled a redacted case brief on February 21, 2018. See Goodluck
    Redacted Case Brief (Feb. 21, 2018), P.R. 296, C.R. 406. Commerce rejected Goodluck’s redacted
    case brief on March 7, 2018, see U.S. Department of Commerce Letter re: Rejection of New
    Factual Information (Mar. 7, 2018), P.R. 302, and Goodluck’s corrected database on March 19,
    2018 for containing untimely new factual information, see Letter from U.S. Department of
    Commerce re: Rejection of New Factual Information (Mar. 19, 2018), P.R. 306.
    On April 16, 2018, Commerce issued its final determination announcing a dumping margin
    for Goodluck of 33.80 percent and a cash deposit rate of 33.70 percent. See Final Determination,
    
    83 Fed. Reg. 16,296
    . In reaching the Final Determination, Commerce rejected all of the data
    submitted by Goodluck in its questionnaires and instead relied on total AFA to calculate
    Goodluck’s dumping margin. IDM at 10–14.
    In determining Goodluck’s cash deposit rate of 33.70 percent, Commerce employed the
    .10 percent export subsidy rate calculated for Tube Products of India, Ltd. (“TPI”), a separate
    respondent in the investigation. See Goodluck Final Determination Ministerial Error Comments
    (Apr. 17, 2018), P.R. 316. On April 17, 2018, Goodluck timely filed ministerial error comments
    to challenge Commerce’s use of TPI’s export subsidy rate instead of 4.85 percent, the rate
    calculated specifically for Goodluck in the accompanying countervailing duty investigation. 
    Id.
    Commerce rejected Goodluck’s claim via memorandum on May 16, 2018. See U.S. Department
    of Commerce Ministerial Error Allegation Memorandum (May 16, 2018), P.R. 318.
    On July 10, 2018, Goodluck filed its complaint with this court appealing from Commerce’s
    Final Determination. Compl., ECF No. 4. Goodluck filed its brief on December 14, 2018. Mot.
    for J. on the Agency R. (“Pl.’s Br.”), ECF Nos. 21–22. Defendant the United States (“the
    Court No. 18-00162                                                                         Page 14
    Government) and Defendant-Intervenors filed their responses on April 24, 2019. Resp. to Mot.
    for J. on the Agency R. (“Def.’s Br.”), ECF No. 27; Resp. in Opp’n to Mot. for J. on the Agency
    R. (“Def.-Inters.’ Br.”), ECF No. 28. Goodluck filed its reply on May 28, 2019. Reply Br. (“Pl.’s
    Reply”), ECF Nos. 30–31. The court heard oral argument on July 11, 2019. ECF No. 43.
    JURISDICTION AND STANDARD OF REVIEW
    The court has jurisdiction over this dispute pursuant to 
    28 U.S.C. § 1581
    (c) and 19 U.S.C.
    § 1516a. The court may review final affirmative determinations in countervailing duty or
    antidumping duty proceedings under 19 U.S.C. § 1561a(a)(2)(B)(i) and will hold unlawful those
    agency determinations which are unsupported by substantial evidence on the record or otherwise
    not in accordance with law under 19 U.S.C. § 1515a(b)(1)(B)(i).
    DISCUSSION
    Goodluck contends that Commerce’s determination was unsupported by substantial
    evidence and contrary to law because (1) Commerce abused its discretion by rejecting Goodluck’s
    correction submission at verification; (2) Commerce’s reliance on adverse facts available was not
    warranted where there were no gaps in the record and where Goodluck did not significantly impede
    the investigation; and (3) Commerce deviated from its typical practice when calculating
    Goodluck’s export subsidy cash deposit without adequate explanation. The court concludes that
    Commerce’s decision to reject Goodluck’s corrections was an abuse of discretion and remands to
    Commerce to consider Goodluck’s corrected submission as well as to explain why it has departed
    from its general practice for calculating export subsidy cash deposit offset rates in this case. The
    court declines to reach Commerce’s reliance on facts otherwise available and adverse inferences
    at this time.
    Court No. 18-00162                                                                        Page 15
    I.       Commerce’s rejection of Goodluck’s updated data was not supported by substantial
    evidence or in accordance with law.
    The Government contends that Commerce properly rejected Goodluck’s corrections
    submitted at verification because the errors Goodluck identified amounted to untimely new factual
    information under 
    19 C.F.R. § 351.301
    (c)(5) and the verification instructions.6 Specifically,
    according to the Government, the errors identified were so fundamental and systematic that they
    compromised the integrity of the dataset as a whole.
    The court finds this argument unpersuasive in light of existing case law and Commerce’s
    own past practice. For example, in NTN Bearing Corp. v. United States, the Federal Circuit held
    that Commerce abused its discretion when it refused to consider NTN’s corrections of errors in its
    submissions because of the “untimely” submission of the corrective information. NTN, 
    74 F.3d at 1208
    . When converting its sales data to the manner requested by Commerce, NTN entered some
    improper code numbers. 
    Id.
     “The effect of these mistakes was compounded because [Commerce]
    used a sampling method to determine the extent of dumping, and the sample data contained the
    clerical errors.” 
    Id. at 1205
    . In concluding that Commerce abused its discretion by rejecting
    NTN’s corrections, the Federal Circuit noted that “NTN responded in a timely manner to the
    preliminary determination” when alerting Commerce to its errors, and that “[a] straightforward
    mathematical adjustment was all that was required. Failure to make it resulted in the imposition
    6
    Commerce’s November 22, 2017 Sales and Costs Verification Agenda-Outline instructions:
    [V]erification is not intended to be an opportunity for submission of new factual
    information. New information will be accepted at verification only when: (1) the
    need for that information was not evident previously; (2) the information makes
    minor corrections to information already on the record; or (3) the information
    corroborates, supports or clarifies information on the record.
    See U.S. Department of Commerce Sales Verification Agenda-Outline for Goodluck (Nov. 22,
    2017), P.R. 244, C.R. 240; see also U.S. Department of Commerce Cost Verification Agenda-
    Outline for Goodluck (Nov. 27, 2017), P.R. 246, C.R. 241.
    Court No. 18-00162                                                                         Page 16
    of many millions of dollars in duties not justified under the statute” in light of the antidumping
    laws’ remedial, and not punitive, purposes. 
    Id. at 1208
    .
    Applying NTN, this court in Fischer S.A. v. United States distinguished correctible
    importer error entailing “a mistaken previous submission” to the record from new factual
    information filling “[a] gap caused by failure to provide a questionnaire response or evidence
    during verification.” 
    700 F. Supp. 2d at 1377
    . In that case, Fischer reported the wrong factor --
    called a “Brix level” -- for Commerce to use in its formula for converting its United States and
    home market sales to the same measurement system so Commerce could make an accurate
    comparison. 
    Id.
     at 1370–72. The Fischer court held that
    On the authority of Timken and NTN Bearing, the Court holds that Commerce
    abused its discretion in rejecting Fischer’s additional agreement pages as untimely.
    Doing so was an abuse of discretion because (1) no finality concerns demanded
    exclusion of the additional data at the preliminary results stage; (2) failure to
    consider the additional pages to correct information already provided was a
    violation of Commerce’s duty to determine Fischer’s dumping margin as accurately
    as possible; (3) consideration of the additional data is necessary to ensure that the
    remedial, non-punitive nature of the antidumping laws is not violated by imposition
    of inaccurately high antidumping duties on Fischer despite the evidence that was
    rejected; and (4) the recalculation of Fischer’s dumping margin could be
    accomplished by simply replacing the actual Brix levels reported by Fischer in its
    database with the standard Brix level of 11.8 degrees, should Commerce determine
    upon remand that the sales agreement pages in fact substantiate that Brix levels
    above 11.8 degrees did not increase the United States unit price of Fischer’s NFC.
    
    Id.
     at 1376–77.
    Here, like the respondents in NTN and Fischer, Goodluck sought to rectify reporting
    mistakes contained in its previous submission to the record and not to fill gaps caused by an
    omission or withholding of requested information from Commerce. Goodluck did not seek to
    provide additional underlying production data -- which was correctly reported in its Section B and
    C questionnaire -- but merely to remedy Goodluck’s misapplication of Commerce-devised codes
    to the underlying data. As in NTN, Goodluck’s miscoding of nominal wall thickness in Field 3.5,
    Court No. 18-00162                                                                        Page 17
    which impacted two digits of a 24-digit CONNUM, could have been addressed through a
    “straightforward mathematical adjustment” even though the “effect of these mistakes was
    compounded” by how Commerce used the incorrect CONNUMs. See NTN, 
    74 F.3d at 1205
    ;
    Certain Hot-Rolled Steel Flat Products from the United Kingdom, 
    81 Fed. Reg. 53,436
     (Aug. 12,
    2016), accompanying Issues and Decision Memorandum at Comment 1 (noting that whether a
    reporting error “affected many transactions and cascaded into other fields in [a] database” is not
    dispositive on the error’s categorization).
    The Government argues that accepting Goodluck’s submission would have “involve[d]
    numerous corrections,” including “revising the weighted average CONNUM costs,” see IDM at 8,
    which rendered it new factual information and a systematic error. However, whether the “effect
    of these mistakes was compounded” by how Commerce used the incorrect CONNUMs does not
    transform a minor correction into substantial new factual information. See NTN, 
    74 F.3d at 1205
    ;
    Certain Hot-Rolled Steel Flat Products from the United Kingdom, 
    81 Fed. Reg. 53,436
     (Aug. 12,
    2016), accompanying Issues and Decision Memorandum at Comment 1 (noting that whether a
    reporting error “affected many transactions and cascaded into other fields in [a] database” is not
    dispositive on the error’s categorization). As Commerce has previously stated -- and this court has
    affirmed -- “the value of the errors as a percentage of total sales, or the number of instances of
    errors” is not decisive. See The Coalition for the Preservation of American Brake Drum and Rotor
    Aftermarket Manufacturers v. United States, 
    23 CIT 88
    , 93, 
    44 F. Supp. 2d 229
    , 236 (1999).
    Commerce articulated a standard for delineating what constitutes a “substantial revision of [a]
    response” in Certain Coated Paper Suitable for High Quality Print Graphics Using Sheet-Fed
    Presses from the People’s Republic of China, 
    75 Fed. Reg. 59,217
     (Sept. 27, 2010), accompanying
    Issues and Decision Memorandum at Comment 10. In Certain Coated Paper, despite declaring
    Court No. 18-00162                                                                         Page 18
    that the impact of the respondent APP-China’s miscoding of a particular product characteristic7
    “may not be minor,” Commerce nevertheless accepted the correction at verification as a minor
    correction because it “d[id] not involve production data, or call into question [Factors of
    Production8] and cost reconciliations.” 
    Id.
     Like with APP-China, parties here are not contesting
    the accuracy of Goodluck’s underlying production, input,9 or cost data. On costs specifically, in
    the section entitled “Cost Reconciliations” of Commerce’s Cost Verification Report, Commerce
    stated: “We reconciled the total [period of investigation (“POI”)] costs from the financial
    statements to the total POI cost of units that produced [merchandise under consideration].” See
    U.S. Department of Commerce Cost Verification Report at 8 (Jan. 18, 2018), P.R. 278, C.R. 385.
    While it is true that some of the “weighted average CONNUM-specific costs” provided by
    Goodluck were inaccurate, IDM at 12, this does not implicate the accuracy of Goodluck’s
    underlying data, but rather was a cascading result of Goodluck’s coding error in field 3.5 of its
    home market sales database. As with APP-China, the impact of Goodluck’s coding error “may
    7
    The correction at issue in Certain Coated Paper involved revising the “reporting of a particular
    ‘finish’ characteristic for several of APP-China’s products from characteristic ‘2’ – indicating a
    finish between 65 and 74.99, to characteristic ‘1’ – indicating a finish of 75 or greater.” 
    75 Fed. Reg. 59,217
     (September 27, 2010), accompanying Issues and Decision Memorandum at Comment
    10.
    8
    “Factors of production are the inputs consumed to produce the subject merchandise.” Nat’l Nail
    Corp. v. United States, Slip Op. 19-71, 
    2019 WL 2537931
    , at *3 n.11 (CIT June 12, 2019).
    9
    The court notes that Commerce’s IDM stated: “[P]arties commented on Goodluck’s reporting of
    nominal field lengths, steel grade, credit, insurance, indirect selling expenses, and inventory
    carrying costs, U.S. destinations, date of sale, quality rejections, cost specificity, intra-company
    transfers, and rebates.” However, Commerce did not reach these points in light of its decision to
    rely on total AFA. IDM at 13. It is possible that these additional contentions could undermine the
    accuracy of Goodluck’s reported production data, factors of production, and/or cost
    reconciliations, and thereby render Goodluck’s submission a “substantial revision” under the
    standard articulated in Certain Coated Paper. However, because Commerce did not fully explain
    the implications of this statement, the court will not speculate.
    Court No. 18-00162                                                                         Page 19
    not [have] be[en] minor;” nevertheless, under the standard articulated in Certain Coated Paper,
    correction of Goodluck’s errors should not require a “substantial revision of [its] response” given
    the accuracy of Goodluck’s underlying data.
    The Government further contended that accurately coded product characteristics are
    indispensable to less than fair value investigations, and within product characteristics, accurately
    coded wall thickness are paramount to cold-drawn mechanical tubing because misreporting wall
    thickness dramatically skews costs. See Oral Argument, July 11, 2019, ECF No. 43. In the
    Government’s estimation, errors in such crucial elements of the investigation cannot be considered
    minor. As an initial matter, the court notes that Commerce did not specifically emphasize the
    importance of wall thickness in its IDM. Rather, Commerce more generally asserted that
    “[w]ithout accurate reporting of physical characteristics [generally] and matching CONNUMs in
    Goodluck’s databases, Commerce does not have the primary components to perform an accurate,
    reliable margin calculation for Goodluck” and stated that such “errors in a factor as fundamental
    as the control number invalidate[] the allocations, matches, and calculations that follow.” IDM at
    8–9.
    Notwithstanding these contentions, Commerce has previously deemed errors in product
    characteristic coding, wall thickness coding, and CONNUM creation to be minor errors at
    verification, despite the impact of the errors on “allocations, matches, and calculations that
    follow.” In a case similar to the one at bar, National Steel Corp. v. United States, Commerce found
    errors committed by the respondent Hoogovens to be neither “systematic in nature” nor
    “amount[ing] to a failure to provide information” -- despite the fact that, like Goodluck,
    Hoogovens informed Commerce on the first day of verification that it had discovered inaccuracies
    affecting “product characteristics submitted to Commerce,” the majority of which “involved the
    Court No. 18-00162                                                                          Page 20
    miscalculating of thickness [which] placed a number of sales in the wrong thickness group.” 
    18 CIT 1126
    , 1127, 
    870 F. Supp. 1130
    , 1132 (1994). As with Goodluck, Commerce found that
    Hoogovens’ error affected home market sales CONNUMs, and correspondingly, “a significant
    percentage of margin calculations” and “the model matching hierarchy.” Id. at 1133. Yet, unlike
    with Goodluck, Commerce concluded that Hoogovens’ “errors themselves were minor” and
    Hoogoven “did not omit data, but only provided inaccurate information.”10 Id. at 1134. While
    each case turns on its own facts and circumstances, treating Goodluck’s strikingly similar error
    differently is arbitrary.11
    10
    Given the case law that governed in 1994, Commerce decided not to accept Hoogovens’
    corrections due to time constraints, despite finding that Hoogovens’ errors were minor. National
    Steel, 870 F. Supp. at 1133. Such a determination by Commerce -- finding a respondent’s
    computer conversion errors to be minor, and yet declining to correct them -- would soon be
    precluded by the Federal Circuit case NTN Bearing Corp. v. United States in 1995, which held
    that Commerce had abused its discretion in refusing to correct a respondent’s typographical errors
    identified prior to the issuance of final results. 
    74 F.3d at
    1207–08. All of that aside, the salient
    point in this case comparison is that Commerce’s reasoning is inexplicably inconsistent:
    Commerce declared Hoogovens’ CONNUM reporting errors to be neither “systematic” nor
    “amount[ing] to a failure to provide information,” National Steel, 870 F. Supp. at 1134, and yet
    found Goodluck’s strikingly similar error to be “systematic,” amounting to a provision of new
    factual information.
    11
    When questioned about this difference, the Government responded in Oral Argument that
    because Hoogovens’ mistakes resulted from a computer conversion error, it therefore qualified as
    a clerical error (stemming from “inaccurate copying, duplication, or the like” under 
    19 C.F.R. § 351.224
    (f)); while Goodluck’s error “entailed more than copying or duplicating,” but properly
    analyzing and following directions. Def.’s Br. at 21; Oral Argument, July 11, 2019, ECF No. 43;
    IDM at 8. The court is unpersuaded. To produce the erroneously coded 682 home market sales
    observations, Goodluck still copied and/or duplicated codes provided by Commerce, but simply
    the wrong ones -- as evidenced by the fact that Goodluck’s errors reflected Commerce’s original
    July 6, 2017 coding guidance instead of the revised guidance. See First CONNUM Letter at
    Attach. 1; see also Revised CONNUM Letter.
    Moreover, the Government acknowledged that had Goodluck simply mistyped
    Commerce’s updated coding guidance in the process of revising the 682 home market sales
    observations -- with the same cascading effects on CONNUMs and weighted average CONNUM-
    specific costs -- such mistyping would be a correctible clerical error at the preliminary
    investigation phase, but maintained that verification was too late to make such corrections.
    However, this delineation in timing is inconsistent with NTN, which held that Commerce abused
    Court No. 18-00162                                                                            Page 21
    Finally, the Government contests Commerce’s assertion that “the rejected information was
    available in the record” because “the code was ‘derived’ from the data reported in a separate field,
    field 2.5,” pointing out that “nowhere in its case brief does Goodluck -- or can it -- point to a place
    on the record in which this new information [the corrected CONNUM codes] can be found.” Def.’s
    Br. at 25. The Government argues that the 13 unreported CONNUMs stemming from Goodluck’s
    coding error necessarily dictate a finding of new factual information. The court is not convinced
    by this argument. As discussed above, in NTN, some code numbers were incorrect,12 and yet the
    its discretion in declining to correct typographical coding errors identified before Commerce had
    issued its final results. 
    74 F.3d at 1209
    .
    Additionally, the court is unconvinced by Commerce’s explanation for why this error
    constitutes a failure to adhere to instructions. In its IDM at 8, Commerce states that “Goodluck
    did not adhere to Commerce’s coding instructions, despite having the correct information on
    hand.” The court notes that for any correctible error, the respondent necessarily has the correct
    information on hand, but inadvertently reports the wrong information instead and thus seeks to
    correct that mistake. In addition, as discussed, this is not a situation where a company was
    unresponsive, provided fraudulent information, or clearly ignored Commerce’s instructions;
    rather, Goodluck believed it had reported the correct information in accordance with Commerce’s
    instructions -- and largely did so -- but made a mistake. It is thus unclear from Commerce’s
    explanation what renders Goodluck’s error here a failure to follow instructions rather than a
    correctible error.
    12
    The Government emphasizes the Federal Circuit’s use of the word ‘transposed’ to suggest that
    Goodluck’s error was distinguishable from those at issue in NTN. However, as discussed in
    footnote 11, the Government’s attempt to draw a divide between pure typographical errors and
    Goodluck’s error is a distinction without a difference. Moreover, Federal Circuit progeny cases
    establish that the holding of NTNis not limited to pure typographical, or even clerical, errors.
    Papierfabrik August Koehler SE v. United States -- the most recent Federal Circuit treatment of
    NTN Bearing -- stated:
    We have held that Commerce abused its discretion in refusing to accept updated
    data when there was plenty of time for Commerce to verify or consider it. NTN
    Bearing Corp. v. United States, 
    74 F.3d 1204
    , 1207–8 (Fed. Cir. 1995) (requiring
    correction of typing errors); Timken U.S. Corp. v. United States, 
    434 F.3d 1345
    ,
    1353 (Fed. Cir. 2006) (expanding the holding in NTN to “any type of importer
    error—clerical, methodology, substantive, or one in judgment—…provided that
    the importer seeks correction before Commerce issues its final results and
    adequately proves the need for the requested corrections”).
    843 F.3d at 1384.
    Court No. 18-00162                                                                        Page 22
    Federal Circuit found that Commerce abused its discretion by rejecting NTN’s corrections.
    Moreover, as discussed above, cascading effects of errors -- such as the effect of CONNUMs here
    -- do not transform a correctible error into new factual information.
    The cases provided by Defendant-Intervenor and the Government involving untimely new
    factual information -- as opposed to correctible importer error -- are easily distinguishable from
    the instant case, and in fact further illuminate how Goodluck’s submission differs from
    impermissible new factual information. For example, in Mukand Ltd. v. United States, the Federal
    Circuit found that Commerce had permissibly refused the respondent’s untimely new information
    because Commerce had asked Mukand “for size-specific cost information on five separate
    occasions” and Mukand was only “suddenly willing and able to provide” the data after Commerce
    had preliminarily resorted to adverse facts available. 767 F.3d at 1305. By contrast, Goodluck
    demonstrated no such withholding of requested information from Commerce. Instead, Goodluck
    complied with Commerce’s requests for the underlying data, but improperly applied product
    characteristic codes, with respect to one of nine physical characteristics. Pl.’s Br. at 40. While
    Mukand’s submission was clearly an attempt to fill “[a] gap caused by [Mukand’s] failure to
    provide a questionnaire response,” and therefore constituted untimely new factual information, see
    Fischer, 
    700 F. Supp. 2d at 1377
    , Goodluck’s submission was merely corrective information to
    rectify reporting errors in its earlier submission.13
    13
    For its part, the Government cited to Stupp Corp. v. United States, 43 CIT __, 
    359 F. Supp. 3d 1293
    , 1301 (2019) for the proposition that manipulating existing record evidence to derive
    corrected data necessarily constitutes new factual information. The court disagrees with this
    interpretation as applied to the facts of this case. In Stupp, the respondent, SeAH, manipulated
    existing record evidence to create two new databases, not for the purpose of correcting its own
    coding error, but rather to rebut evidence placed on the record by Commerce. Such rebuttal
    evidence falls expressly within the definition of “factual information” under 
    19 C.F.R. § 351.102
    (b)(21)(iv) (“evidence submitted by any interested party to rebut . . . evidence placed on
    the record by the Department”), with corresponding regulatory deadlines. By contrast
    Court No. 18-00162                                                                             Page 23
    For these reasons, the court concludes that Goodluck’s submission was correctible
    importer14 error and not untimely new factual information. Goodluck sought to rectify reporting
    mistakes contained in its previous submission to the record, and not to supply additional underlying
    data to fill gaps caused by an omission or withholding of requested information from Commerce.
    While the Government nevertheless contended that Goodluck’s mistakes should be held to
    “reach[] the threshold for new factual information pursuant to 
    19 C.F.R. § 351.301
    (c)(5),”
    Commerce’s proffered reasons for doing so are either unsupported by case law or would entail
    Commerce arbitrarily treating similar situations differently, Consolidated Bearings, 
    348 F.3d at
    1007 (citing RHP Bearings, 
    288 F.3d at 1347
    ), amounting to an abuse of discretion. Finally, while
    the Government contended that verification was simply too late in the investigation process to
    correct Goodluck’s errors, as discussed above, the Federal Circuit has found correcting analogous
    errors disclosed in a similar timeframe to be feasible and appropriate. See NTN, 
    74 F.3d at 1208
    (“Preliminary determinations are ‘preliminary’ precisely because they are subject to change. Thus,
    the tension between finality and correctness simply did not exist at the time NTN requested
    correction.”). In sum,
    [D]raconian penalties are [not] appropriate for the making of clerical errors in order
    to insure submission of proper data. Clerical errors are by their nature not errors in
    judgment but merely inadvertencies. While the parties must exercise care in their
    submissions, it is unreasonable to require perfection. [Commerce’s] refusal to
    consider [plaintiff’s] request for correction of clerical errors in this case constituted
    an abuse of discretion.
    “Commerce’s regulations . . . do not provide a time frame within which [importer error] corrections
    should be submitted.” Deacero, 353 F. Supp. at 1307.
    14
    In light of the court’s finding that Goodluck’s errors are minor, it need not address the issue as
    to whether Commerce is only obligated to correct clerical errors versus methodological ones
    beyond the discussion found in footnote 12.
    Court No. 18-00162                                                                             Page 24
    Id. at 1208–09. The court thus remands to Commerce to consider Goodluck’s corrected data
    submission.15
    II.      Commerce’s did not adequately explain its export subsidy cash deposit offset
    determination.
    Commerce used the export subsidy rate of another respondent, TPI, to calculate
    Goodluck’s export subsidy cash deposit offset, rather than the rate specifically calculated for
    Goodluck in the companion CVD investigation. Goodluck argued that doing so was contrary to
    Commerce’s general practice of adjusting antidumping margins -- even those calculated through
    AFA -- by the company-specific export subsidy rate calculated in the companion CVD case, and
    was otherwise unreasonable. Pl.’s Br. at 43. The Government counters that because 19 U.S.C. §
    1677a(c)(1)(C)16 does not otherwise define how Commerce should calculate a cash deposit rate in
    an investigation, “Commerce has discretion to establish a reasonable practice.” Def.’s Br. at 26.
    15
    As discussed above, Goodluck also challenged Commerce’s application of facts available and
    adverse inferences. Although Commerce has discretion when applying facts available and adverse
    inferences, “this discretion is not without limits. The appropriate rate ‘will depend upon the facts
    of a particular case,’ cannot be ‘punitive, aberrational, or uncorroborated,’ includes ‘some built-in
    increase’ to deter non-compliance, and . . . reflects the seriousness of the non-cooperating party’s
    misconduct.” BMW of N. Am. LLC v. United States, 
    926 F.3d 1291
    , 1301 (Fed. Cir. 2019) (citing
    Nan Ya Plastics Corp. v. United States, 
    810 F.3d 1333
    , 1347 (Fed. Cir. 2016) and F.lli De Cecco
    Di Filippo Fara S. Martino S.p.A. v. United States, 
    216 F.3d 1027
    , 1032 (Fed. Cir. 2000)).
    However, because the court finds that Goodluck’s submission involved minor correctible importer
    error, at this time, it need not address Commerce’s use of facts available, as the gap alleged by
    Commerce may be resolved by Commerce’s consideration of Goodluck’s submission.
    Additionally, because a proper facts available determination is a prerequisite for use of adverse
    facts available, the court correspondingly does not reach Commerce’s reliance on adverse
    inferences and particularly total AFA.
    16
    19 U.S.C. § 1677a(c)(1)(C) provides:
    (c) Adjustments for export price and constructed export price
    The price used to establish export price and constructed export price shall
    be—
    (1) increased by—
    Court No. 18-00162                                                                         Page 25
    Commerce indeed has discretion to establish a reasonable practice, but it must also explain
    the reasons for deviating from that practice. See, e.g., Consolidated Bearings, 
    348 F.3d at 1007
    (holding that if “Commerce acted differently in this case than it has consistently acted in similar
    circumstances without reasonable explanation, then Commerce’s actions will have been
    arbitrary””).   Here, Goodluck cited several determinations in which “Commerce adjusted
    respondents’ [AFA] ADD margins . . . by the export subsidies the company received in the
    companion CVD case.” Pl.’s Reply at 16–18. See, e.g., Carbazole Violet Pigment 23 from India:
    Final Results of Antidumping Duty Administrative Review; 2015-2016, 
    83 Fed. Reg. 15,788
    ,
    15,789 (Apr. 12, 2018); Fine Denier Polyester Staple Fiber From India: Final Affirmative
    Determination of Sales at Less Than Fair Value, 
    83 Fed. Reg. 24,737
     (May 30, 2018). At Oral
    Argument, the Government acknowledged that Fine Denier indeed represents Commerce’s most
    common export subsidy cash deposit offset practice, see Oral Argument, July 11, 2019, ECF No.
    43, but asserted that because Goodluck was uncooperative in the companion countervailing duty
    investigation, the lowest possible export subsidy cash deposit offset adjustment was appropriate.17
    Def.’s Br. at 27.    On the other hand, Defendant-Intervenors offer a different rationale for
    (C) the amount of any countervailing duty imposed on the
    subject merchandise under part I of this subtitle to offset an
    export subsidy
    17
    Goodluck contends that it fully cooperated in the companion countervailing duty investigation.
    Pl.’s Reply at 16. The court notes that although AFA was applied in the companion countervailing
    duty investigation, it was on the basis of the Government of India’s and another respondent’s
    failure to cooperate. See Certain Cold-Drawn Mechanical Tubing of Carbon and Alloy Steel From
    India: Final Affirmative Countervailing Duty Determination, 
    82 Fed. Reg. 58,172
     (Dec. 11, 2017),
    and accompanying Issues and Decision Memorandum at 27 (noting that, regarding the application
    of an AFA rate to Goodluck, “[t]he CAFC has upheld the Department’s application of AFA to a
    non-cooperating government even if it subjects a cooperating respondent to the ‘collateral effects’
    of the adverse inference”).
    Court No. 18-00162                                                                        Page 26
    Commerce’s decision: that, “[c]ontrary to Goodluck’s claim that Commerce ‘inexplicably’ offset
    Goodluck’s antidumping rate using the lowest export subsidy rate, the Department’s normal
    practice is to offset an antidumping duty rate based on AFA with the lowest export subsidy rate in
    the companion countervailing proceeding.” Def.-Inters.’ Br. at 34–35 (emphasis original).
    Nowhere in the record does Commerce, itself, offer either explanation. In the Final
    Determination, Commerce merely states that:
    Commerce normally adjusts cash deposits for estimated antidumping duties by the
    amount of export subsidies countervailed in a companion countervailing duty
    (CVD) proceeding, when CVD provisional measures are in effect. Accordingly,
    where Commerce made an affirmative determination for countervailable export
    subsidies, Commerce has offset the estimated weighted- average dumping margin
    by the appropriate CVD rate.
    Final Determination, 83 Fed. Reg. at 16,297.       In response to Goodluck’s ministerial error
    allegations, Commerce reiterated this language and detailed that “Commerce deducted the lowest
    calculated export subsidy rate determined for any party in the companion CVD investigation from
    Goodluck’s antidumping margin, which was based on total adverse facts available, (AFA), for
    antidumping cash deposit purposes, as intended.” U.S. Department of Commerce Ministerial Error
    Allegation Memorandum, (May 16, 2018), P.R. 318. This statement describes what Commerce
    did, but does not explain why Commerce used TPI’s rate from the companion CVD investigation
    rather than the typical practice of using Goodluck’s. Thus, because Commerce has not provided a
    sufficient explanation on the record for departing from its usual practice of using the companion
    CVD rate in Goodluck’s case, the court remands to Commerce to reconsider or to provide a more
    comprehensive explanation.
    CONCLUSION
    In sum, the court finds that Goodluck’s revised data submission should be categorized as a
    correctible importer mistake as opposed to untimely new factual information and remands to
    Court No. 18-00162                                                                            Page 27
    Commerce for consideration of Goodluck’s corrected submission. Correspondingly, the court
    does not reach whether Commerce’s reliance on facts otherwise available and adverse inferences
    is supported by substantial evidence and in accordance with law at this time. The court further
    remands to Commerce to better explain or reconsider its approach to calculating the export subsidy
    cash deposit offset. Commerce shall file with this court and provide to the parties its remand
    results within 120 days of the date of this order; thereafter, the parties shall have 30 days to submit
    briefs addressing the revised final determination to the court and the parties shall have 15 days
    thereafter to file reply briefs with the court.
    SO ORDERED.
    /s/ Gary S. Katzmann
    Gary S. Katzmann, Judge
    Dated: $XJXVW
    New York, New York