Slip Op. 23-
UNITED STATES COURT OF INTERNATIONAL TRADE
MARMEN INC., MARMEN ÉNERGIE INC., and MARMEN ENERGY CO.,
Plaintiffs, and
WIND TOWER TRADE COALITION,
Consolidated Plaintiff, Before: Jennifer Choe-Groves, Judge v.
Consol. Court No. 20-00169 UNITED STATES,
Defendant, and
WIND TOWER TRADE COALITION, MARMEN INC., MARMEN ÉNERGIE INC., and MARMEN ENERGY CO.,
Defendant-Intervenors.
OPINION AND ORDER
[Sustaining the U.S. Department of Commerce’s remand redetermination in the 2018–2019 antidumping duty investigation of utility scale wind towers from Canada.]
Dated: March 20, 2023
Jay Charles Campbell, Allison Jessie Gartner Kepkay, and Ron Kendler, White & Case, LLP, of Washington, D.C., for Plaintiffs and Defendant-Intervenors Marmen, Inc., Marmen Energy Co., and Marmen Energie Inc.
Alan Hayden Price, Derick G. Holt, Elizabeth Seungyon Lee, John Allen Riggins, Laura El-Sabaawi, Maureen Elizabeth Thorson, Robert Edward DeFrancesco, III, Tessa Victoria Capeloto, and Theodore Paul Brackemyre, Wiley Rein, LLP, of Washington, D.C., for Consolidated Plaintiff and Defendant-Intervenor Wind Tower Trade Coalition.
Reginald T. Blades, Jr., Assistant Director, and Joshua Ethan Kurland, Senior Trial Attorney, Commercial Litigation Branch, Civil Division, U.S. Department of Justice, of Washington, D.C., for Defendant United States. With them on the brief were Brian M. Boynton, Principal Deputy Assistant Attorney General, and Patricia M. McCarthy, Director. Of counsel on the brief was Jesus N. Saenz, Attorney, Office of the Chief Counsel for Trade Enforcement and Compliance, U.S. Department of Commerce.
Choe-Groves, Judge: Before the Court is the U.S. Department of Commerce’s (“Commerce”) remand redetermination in the antidumping duty investigation of utility scale wind towers from Canada, filed pursuant to the Court’s Remand Order in Marmen Inc. v. United States (“Marmen I”), 45 CIT __, 545 F. Supp. 3d 1305 (2021). See Final Results of Redetermination Pursuant to Court Remand (“Remand Redetermination”), ECF Nos. 61, 62; see also Utility Scale Wind Towers from Canada (“Final Determination”), 85 Fed. Reg. 40,239 (Dep’t of Commerce July 6, 2020) (final determination of sales at less than fair
3
value and final negative determination of critical circumstances; 2018–2019), accompanying Issues and Decision Mem. for the Final Affirmative Determination in the Less-Than-Fair-Value Investigation of Utility Scale Wind Towers from Canada, ECF No. 18-5 (June 29, 2020) (“Final IDM”).
In Marmen I, the Court remanded for Commerce to reconsider the rejection of the cost reconciliation information of Plaintiffs Marmen Inc., Marmen Energy Co., and Marmen Energie Inc. (collectively, “Marmen”) and Commerce’s use of the differential pricing average-to-transaction (“A-to-T”) method to calculate Marmen’s dumping margin. Marmen I, 45 CIT at __, 545 F. Supp. 3d at 1315–20. On remand, Commerce reconsidered the additional cost reconciliation information and the use of the Cohen’s d test in light of Stupp Corp. v. United States, 5 F.4th 1341 (Fed. Cir. 2021). See generally, Remand Redetermination. Marmen filed comments in opposition to the Remand Redetermination. Pls.’ Comments Opp’n Final Results of Redetermination Pursuant to Court Remand (“Pls.’ Cmts.”), ECF Nos. 66, 67. Defendant United States (“Defendant”) responded to Plaintiffs’ Comments. Def.’s Resp. Pls.’ Comments Commerce’ Remand Redetermination (“Def.’s Resp.”), ECF Nos. 70, 71 (superseded by ECF Nos. 79, 80). Defendant- Intervenor Wind Tower Trade Coalition (“Defendant-Intervenor”) filed comments in support of the Remand Redetermination. [Def.-Interv.’s] Comments Supp.
4
Remand Redetermination (“Def.-Interv.’s Cmts.”), ECF Nos. 72, 73. For the following reasons, the Court sustains the Remand Redetermination.
BACKGROUND
The Court presumes familiarity with the underlying facts and procedural history of this case and recites the facts relevant to the Court’s review of the Remand Redetermination. See Marmen I, 45 CIT at __, 545 F. Supp. 3d at 1311– 12. In August 2019, Commerce initiated an antidumping duty investigation into wind towers from Canada for the period covering July 1, 2018 through June 30, 2019. Utility Scale Wind Towers from Canada, Indonesia, the Republic of Korea, and the Socialist Republic of Vietnam, 84 Fed. Reg. 37,992, 37,992–93 (Dep’t of Commerce Aug. 5, 2019) (initiation of less-than-fair-value investigations). Commerce selected Marmen, Inc. and Marmen Energie Inc. as mandatory respondents. See Decision Mem. for the Prelim. Determination in the Less-Than- Fair-Value Investigation of Utility Scale Wind Towers from Canada (Feb. 4, 2020) (“Prelim. DM”) at 1–2, PR 146.1 In the Final Determination, Commerce assigned weighted-average dumping margins of 4.94 percent to Marmen, Inc. and Marmen
1 Citations to the administrative record reflect public record (“PR”) and public remand record (“PRR”) document numbers filed in this case, ECF Nos. 46, 75.
5
Energie Inc.2 Final Determination, 85 Fed. Reg. at 40,239. Commerce determined the all-others weighted average dumping margin of 4.94 percent based on Marmen’s dumping margin. Id.
In the Final Determination, Commerce determined that Marmen’s steel plate costs did not reasonably reflect the costs associated with the production and sale of the products and weight-averaged Marmen’s reported steel plate costs. Final IDM at 4–6. Commerce rejected a portion of the supplemental cost reconciliation information submitted by Marmen as untimely, unsolicited new information. Id. at 7–9. Commerce applied a differential pricing analysis using the Cohen’s d test and determined that there was a pattern of export prices that differed significantly. Id. at 10–11. As a result, Commerce calculated Marmen’s weighted-average dumping margin by using the alternative average-to-transaction method. Id.
The Court remanded for Commerce to explain its use of the Cohen’s d test in light of Stupp Corp. v. United States, 5 F.4th 1341 (Fed. Cir. 2021), and for Commerce to further explain or consider Marmen’s supplemental cost reconciliation information. Marmen I, 45 CIT at __, 545 F. Supp. 3d. at 1317–21.
2 The Court notes that, although Marmen Energy Co. was not included as a mandatory respondent alongside Marmen, Inc. and Marmen Energie Inc., comments and questionnaire responses were submitted collectively by the three Plaintiffs during Commerce’s investigation. The Court herein refers to their assigned weighted-average dumping margins collectively as “Marmen’s dumping margin.”
6
On remand, Commerce accepted the previously rejected information from Marmen. Remand Redetermination at 411. Commerce examined the additional cost reconciliation information together with other information on the record, and Commerce determined that the purported corrections were already reflected in Marmen’s audited financial statements. Id. Commerce did not adjust Marmen’s cost of manufacturing or cost of production. Id. Commerce also reconsidered the differential pricing analysis and determined that the assumptions of normality and roughly equal variances at issue in Stupp were not relevant to Commerce’s application of the Cohen’s d test on remand. Id. at 1250.
JURISDICTION AND STANDARD OF REVIEW The Court has jurisdiction under 19 U.S.C. § 1516a(a)(2)(B)(i) and 28 U.S.C. § 1581(c), which grant the Court authority to review actions contesting the final determination in an antidumping duty investigation. The Court shall hold unlawful any determination found to be unsupported by substantial evidence on the record or otherwise not in accordance with the law. 19 U.S.C. § 1516a(b)(1)(B)(i). The Court also reviews determinations made on remand for compliance with the Court's remand order. Ad Hoc Shrimp Trade Action Comm. v. United States, 38 CIT __, __, 992 F. Supp. 2d 1285, 1290 (2014), aff'd, 802 F.3d 1339 (Fed. Cir. 2015).
7
DISCUSSION
I. Commerce’s Rejection of Marmen’s Additional Cost Reconciliation Information
In order to determine whether certain products are being sold at less than fair value in the United States, Commerce compares the export price, or constructed export price, with normal value. 19 U.S.C. § 1673. Export price and constructed export price are the price at which the subject merchandise is being sold in the U.S. market, while normal value is the price at which a “foreign like product” is sold in the producer’s home market or in a comparable third-country market. Id. §§ 1677a(a)–(b), 1677b(a)(1)(B). Before calculating a dumping margin, Commerce must identify a suitable “foreign like product” with which to compare the exported subject merchandise. See § 1677b(a)(1)(B). A “foreign like product,” in order of preference, is:
(A) The subject merchandise and other merchandise which is identical in physical characteristics with, and was produced in the same country by the same person as, that merchandise.
(B) Merchandise —
(i) produced in the same country and by the same person as the subject merchandise,
(ii) like that merchandise in component material or materials and in the purposes for which used, and (iii) approximately equal in commercial value to that merchandise.
(C) Merchandise —
(i) produced in the same country and by the same person and of the same general class or kind as the subject merchandise,
8
(ii) like that merchandise in the purposes for which used, and (iii) which the administering authority determines may reasonably be compared with that merchandise.
19 U.S.C. § 1677(16); NSK Ltd. v. United States, 26 CIT 650, 657–58, 217 F. Supp. 2d 1291, 1299–1300 (2002).
When determining costs of production, 19 U.S.C. § 1677b states that:
costs shall normally be calculated based on the records of the exporter or producer of the merchandise, if such records are kept in accordance with the generally accepted accounting principles [“GAAP”] of the exporting country (or the producing country, where appropriate) and reasonably reflect the costs associated with the production and sale of the merchandise.
19 U.S.C. § 1677b(f)(1)(A). The statute requires that “reported costs must normally be used only if (1) they are based on the records . . . kept in accordance with the GAAP and (2) reasonably reflect the costs of producing and selling the merchandise.” See Dillinger France v. United States, 981 F.3d 1318, 1321 (Fed. Cir. 2020) (emphasis in original) (internal quotation marks and citations omitted). Commerce is not required to accept the exporter’s records. Thai Plastic Bags Indus. Co. v. United States, 746 F.3d 1358, 1365 (Fed. Cir. 2014). Commerce may reject a company’s records if it determines that accepting them would distort the company’s true costs. See Am. Silicon Techs. v. United States, 261 F.3d 1371, 1377 (Fed. Cir. 2001). Commerce is directed to consider all available evidence on the proper allocation of costs. 19 U.S.C. § 1677b(f)(1)(A). Physical characteristics
9
are a prime consideration when Commerce conducts its analysis. Thai Plastic Bags, 746 F.3d at 1368. If factors beyond the physical characteristics influence the costs, however, Commerce will normally adjust the reported costs in order to reflect the costs that are based only on the physical characteristics. See id.
To determine whether the subject merchandise wind towers from Canada were sold in the United States at less than fair value under section 733 of the Tariff Act of 1930, Commerce first considered all products produced and sold by Marmen in Canada during the period of investigation for the purpose of determining the appropriate product comparisons to U.S. sales. Prelim. DM at 13. Commerce determined that there were no sales of identical merchandise in the ordinary course of trade in Canada that could be compared to U.S. sales. Id.
Commerce did not dispute whether Marmen’s records were kept properly, noting that “the record is clear that the reported costs are derived from the Marmen Group’s normal books and records and that those books are in accordance with Canadian GAAP.” Final IDM at 5; see also Marmen’s Utility Scale Wind Towers from Canada: Response to Question 14.g of the Supplemental Section Questionnaire (Dec. 13, 2019) at 2–4, PR 123–25. Commerce focused on the second prong of 19 U.S.C. § 1677b(f)(1)(A), calling into question whether Marmen reasonably reflected the costs of producing and selling the merchandise. Final IDM at 5.
10
In Marmen I, this Court remanded for Commerce to reconsider the rejection of Plaintiffs’ cost reconciliation information. Marmen I, 45 CIT at __, 545 F. Supp. 3d at 1315–17. On remand, Commerce accepted and reconsidered Marmen’s cost reconciliation information that Commerce had previously rejected. See Remand Redetermination at 4–11. Commerce explained that on remand it evaluated the information provided by Plaintiffs and determined that one portion of the information should be rejected because the information adjusted for amounts already accounted for in the costs that were reported to Commerce. Id. Commerce determined that Plaintiffs’ overall cost reconciliation difference remained outstanding and attributed the amount to Marmen’s cost of production. Id. Defendant-Intervenor supports Commerce’s determination. See Def.-Interv.’s Cmts. at 10–16.
Marmen argues that Commerce’s rejection of the information was unreasonable because the information was a “minor correction to Marmen Inc.’s cost reconciliation worksheet based on incorrect and confused claims that are unsupportable.” Pls.’ Cmts. at 2. Marmen challenges Commerce’s characterization that the information would double count an exchange rate adjustment already reflected in the audited cost of goods sold and reported cost of production. Id.
11
A party may submit factual information to rebut, clarify, or correct questionnaire responses. 19 C.F.R. § 351.301(c). The regulations state that
[i]f the factual information is being submitted to rebut, clarify, or correct factual information on the record, the submitter must provide a written explanation identifying the information which is already on the record that the factual information seeks to rebut, clarify, or correct, including the name of the interested party that submitted the information and the date on which the information was submitted.
Id. § 351.301(b)(2).
Commerce has a duty “to determine dumping margins as accurately as possible.” See NTN Bearing Corp. v. United States, 74 F.3d 1204, 1208 (Fed. Cir. 1995) (internal quotation marks and citation omitted). “[A]ntidumping laws are remedial not punitive.” Id. (citation omitted). The U.S. Court of Appeals for the Federal Circuit (“CAFC”) has stated that “Commerce is obliged to correct any errors in its calculations during the preliminary results stage to avoid an imposition of unjustified duties.” Fischer S.A. Comercio, Industria & Agricultura v. United States, 471 Fed. App’x 892, 895 (Fed. Cir. 2012) (citation omitted). Further, “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 United States Corp. v. United States (“Timken”), 434 F.3d 1345, 1353
12
(Fed. Cir. 2006). The Court reviews whether Commerce abused its discretion when rejecting submitted information. See Papierfabrik August Koehler SE v. United States, 843 F.3d 1373, 1384 (Fed. Cir 2016) (“Commerce abused its discretion in refusing to accept updated data when there was plenty of time for Commerce to verify or consider it.”) (citations omitted). When reviewing Commerce’s determination to reject corrective information, this Court may consider factors such as Commerce’s interest in ensuring finality, the burden of incorporating the information, and whether the information will increase the accuracy of the calculated dumping margins. Bosun Tools Co. v. United States, 43 CIT __, __, 405 F. Supp. 3d 1359, 1365 (2019) (citations omitted).
On remand, Commerce accepted and considered the numerous revisions presented by Marmen. Remand Redetermination at 4–11, 38–46. Marmen argues that the information submitted consisted of minor corrections and not new information. See Pls.’ Cmts. at 2. Commerce agreed that several of the revised reconciliations were “minor errors,” such as cell formatting errors and other small clerical errors, which Commerce accepted because they did not alter the data presented in the audited financial statements. Remand Redetermination at 6–7. Commerce stated in the Remand Redetermination, however, that “there was one non-clerical revision that Marmen explained it found while reviewing its records for purposes of preparing the revised cost reconciliations. This revision resulted
13
from an alleged discovery of certain expenses that Marmen claims were not converted from [U.S. dollars] to [Canadian dollars].” Id. at 7 (citing Marmen’s Utility Scale Wind Towers from Canada: Second Supp. Section D Resp. (Feb. 7, 2020) (“Marmen’s Second Supplemental Section D Response”) at 14, PR 151–54). Commerce determined that:
In short, the increase to the [cost of manufacturing] (i.e., the increase in the unreconciled difference) driven by the restatement of the audited financial statements was offset by this new change to Marmen’s cost reconciliation. According to Marmen, this new reconciling item represents non-booked exchange losses that Marmen Inc. incurred on purchases of wind tower sections from affiliate Marmen Energie. This explanation is parallel to the adjusting entry to restate Marmen Inc.’s other purchases to the [Canadian dollar] equivalent values, as discussed above, as an auditor amendment to the financial statements.
Id. (citing Marmen’s Utility Scale Wind Towers from Canada: Request for Additional Information Concerning Second Supp. Section D Resp. (Dec. 8, 2021) (“Marmen’s Second Supplemental Remand Section D Response”) at Attachment 1), PRR 2. Commerce rejected Marmen’s cost reconciliation information because “Marmen did not further explain how, if at all, this error and correction related to the restated financial statements, or whether it was one of the adjustments brought up by the external auditor, Deloitte. The record does not provide any actual support that this new change is required, nor that it is not already accounted for within Marmen’s normal books.” Id.
14
Defendant asserts that the new cost reconciliation information had the effect of duplicating the adjustments for exchange gains and losses already reflected in Marmen’s financial statements. Def.’s Resp. at 24. Defendant contends that Commerce correctly determined that the information in the cost reconciliation spreadsheet, viewed in conjunction with Marmen’s representations regarding its auditor’s adjustments, indicated that Marmen’s auditor had already made any necessary adjustment in restating Marmen’s financial statements that produced the cost of goods sold figure used in the reconciliation. Id.
In support of its determination that the new cost reconciliation information was already accounted for in Marmen’s costs, Commerce cited record evidence comparing an Excel spreadsheet in the Supplemental Remand Section D Response at Attachment 1 with Marmen’s Initial Section D Response at pages D-15 and D- 33 and Exhibit D-3.3 Remand Redetermination at 8–9; see Marmen’s Second Supplemental Remand Section D Response at Attachment 1; Marmen’s Utility Scale Wind Towers from Canada: Sections B, C, and D Response (Oct. 11, 2019)
3 Exhibit D-3 to Marmen’s Initial Section D Response is not included in the record before the Court. Exhibit Supp. D-3 to Marmen’s December 6, 2019 Supplemental Section D Response appears to correspond to the information referenced by Commerce in the Remand Redetermination. See Marmen’s Utility Scale Wind Towers from Canada: Supplemental Section D Response (Dec. 6, 2019) (“Marmen’s December 6, 2019 Supplemental Section D Response”) at Ex. Supp. D-3, PR 114–19.
15
(“Marmen’s Initial Section D Response”) at D-15, D-33, PR 89–97; Marmen’s Utility Scale Wind Towers from Canada: Supplemental Section D Response (Dec. 6, 2019) (“Marmen’s December 6, 2019 Supplemental Section D Response”) at Ex. Supp. D-3, PR 114–19. Marmen’s Initial Section D Response reviewed by Commerce shows that Marmen recorded amounts in its normal books and records in its home currency of Canadian dollars using an alternative exchange rate. Remand Redetermination at 8–9 (citing Marmen’s Initial Section D Response at Exhibit D-3). Citing Marmen’s Initial Section D Response at page D-15, for example, Commerce determined that for purchases in U.S. dollars, Marmen reported that its normal books reflected a cost system conversion from U.S. dollar purchases to Canadian dollars at specific conversion rates. Id. at 8–9. Commerce cited Marmen’s December 6, 2019 Supplemental D Response at D-17 and D-18 to support its determination that Marmen’s auditors periodically adjusted the already converted purchases, and that in preparing Marmen’s original 2018 audited financial statements, the auditors had already made adjustments to reflect actual exchange rates during 2018. Id. at 9; see Marmen’s December 6, 2019 Supplemental Section D Response at D-17–D-18. Based on its review of these record documents, Commerce determined that Marmen’s prior statements and reported calculations established that the exchange gains and losses were already accounted for in Marmen’s costs. Remand Redetermination at 9, 38–46. Thus,
16
Commerce determined that “the record evidence thereby demonstrates that the reported costs, including those of the sections purchased from Marmen Energie, were, in fact, already correctly inclusive of exchange rate differences, and it would be inappropriate to adjust them again for those exchange gains and losses.” Id. at 11.
Because record evidence, including Marmen’s Initial Section D Response with exhibits, Marmen’s December 6, 2019 Supplemental Section D Response with exhibits, and Marmen’s Second Supplemental Remand Section D Response at Attachment 1, shows that Marmen’s auditors already adjusted the reported costs to account for exchange rate differences, the Court concludes that Commerce’s determination that another adjustment would be inappropriate is supported by substantial evidence. The Court holds that Commerce did not abuse its discretion by rejecting Marmen’s proposed corrective information, recognizing that Commerce has an interest in ensuring finality and increasing the accuracy of the calculated dumping margins. Bosun Tools, 43 CIT at __, 405 F. Supp. 3d at 1365.
II. Commerce’s Use of the Cohen’s d Test In Stupp, the CAFC directed the Court to remand Commerce’s use of the Cohen’s d test for further explanation because the data Commerce used may have violated the assumptions of normality, sufficient observation size, and roughly equal variances. 5 F.4th at 1357–60. Before the CAFC, Commerce argued that
17
concerns of normality and population were misplaced because, unlike sampling data used in determining probability or statistical significance, Commerce’s review considered a complete universe of data. Id. at 1359–60. The CAFC expressed concern with Commerce’s explanation because it failed to “address the fact that Professor Cohen derived his interpretive cutoffs under the assumption of normality.” Id. at 1360.
On remand, Commerce reconsidered the use of the Cohen’s d test in light of Stupp as this Court directed in Marmen I. See Remand Redetermination at 12–37, 46–50; Marmen I, 45 CIT at __, 545 F. Supp. 3d at 1320. The standard of review for considering Commerce’s differential pricing analysis is reasonableness. Stupp, 5 F.4th at 1353. The CAFC and the U.S. Court of International Trade have held the steps underlying the differential pricing analysis as applied by Commerce to be reasonable. See e.g., Mid Continent Steel & Wire, Inc. v. United States, 940 F.3d 662, 670–74 (Fed. Cir. 2019) (discussing zeroing and the 0.8 threshold for the Cohen’s d test); Apex Frozen Foods Priv. Ltd. v. United States, 40 CIT __, __, 144 F. Supp. 3d 1308, 1314–37 (2016) (discussing application of the A-to-T method, the Cohen’s d test, the meaningful difference analysis, zeroing, and the “mixed comparison methodology” of applying the A-to-A method and the A-to-T method when 33–66% of a respondent’s sales pass the Cohen’s d test), aff’d, 862 F.3d 1337 (Fed. Cir. 2017); Apex Frozen Foods Priv. Ltd. v. United States, 862 F.3d
18
1322, 1330–34 (Fed. Cir. 2017) (affirming zeroing and the 0.5% de minimis threshold in the meaningful difference test); Stupp Corp. v. United States, 47 CIT __, __, 2023 WL 2206548, at *6–10 (2023) (discussing the reasonableness of the Cohen’s d test as one component of Commerce’s differential pricing analysis). However, the CAFC has stated that “there are significant concerns relating to Commerce’s application of the Cohen’s d test . . . in adjudications in which the data groups being compared are small, are not normally distributed, and have disparate variances.” Stupp, 5 F.4th at 1357.
The Cohen’s d test is “a generally recognized statistical measure of the extent of the difference between the mean of a test group and the mean of a comparison group.” Apex Frozen Foods, 862 F.3d at 1342 n.2. The Cohen’s d test relies on assumptions that the data groups being compared are normal, have equal variability, and are equally numerous. See Stupp, 5 F.4th at 1357. Applying the Cohen’s d test to data that do not meet these assumptions can result in “serious flaws in interpreting the resulting parameter.” See id. at 1358.
Commerce determined on remand that “the assumptions of normality and roughly equal variances” are not relevant to Commerce’s application of the Cohen’s d test. Remand Redetermination at 18. Commerce explained that its dumping analysis in this case assessed the pricing behavior of Marmen in the entire United States market, stating:
19
The U.S. sale price data on which this analysis is based constitute the entire population of sales data and are not a sample of a respondent’s sales data (i.e., the data are for all sales in the United States of subject merchandise by a company during the period of investigation or review). The basis for this analysis is the respondent’s U.S. sales of the subject merchandise for a given period of time. By definition, these U.S. sales comprise the universe of sales on which the respondent’s weighted-average dumping margin depends. The Differential Pricing Analysis examines all sales to determine whether the A-to-A method is the appropriate approach on which to base this calculation. Therefore, in the context of the calculation of the weighted-average dumping margin, the data used are not a sample, but rather constitute the entire population of a respondent’s sales of subject merchandise during the period under examination for the calculation of the weighted-average dumping margin.
Id. at 22.
Commerce determined on remand that the statistical criteria, such as the number of observations, a normal distribution, and approximately equal variances, are related to the statistical significance of sampled data and establish the reliability of an estimated parameter based on the sample data. Id. at 23. Commerce explained further that:
However, for the Cohen’s d test applied in the context of the Differential Pricing Analysis, there is no estimation of the parameters (i.e., mean, standard deviation, and effect size) of the test group or of the comparison group as the calculation of these parameters is based on the complete universe of sale prices to the test and comparison groups.
Unlike with a sample of data where the estimated parameters will change with each sample selected from a population, each time these parameters would be calculated as part of Commerce’s Cohen’s d test, the exact same results would be found because the calculated parameters are the parameters of the entire population and not an estimate of the parameters based on a sample. Accordingly, the means,
20
standard deviations, and Cohen’s d coefficients calculated are not estimates with confidence levels or sampling errors as would be associated with sampled data, but, rather, are the actual values which describe a company’s pricing behavior. Consequently, the statistical significance of the results of the Cohen’s d test is not relevant in Commerce’s application of the differential pricing analysis, which measures practical significance.
Id. at 23–24. Commerce determined, therefore, that:
[i]n Commerce’s application of the Cohen’s d test, such additional analysis is not relevant because the data in both the test group and the comparison group use the full population of sales in each group and are not determined based on controlled random and independent samples of the population. Rather, the results of the Cohen’s d test are based on the entire population of sale price data for comparable merchandise for the test and comparison groups.
Id. at 26.
The Court concludes that Commerce’s use of a population, rather than a sample, in the application of the Cohen’s d test sufficiently negates the questionable assumptions about thresholds that were raised in Stupp. Based on Commerce’s explanation, this Court concludes that Commerce’s application of the Cohen’s d test to determine whether there was a significant pattern of differences was reasonable because Commerce applied the Cohen’s d test to a population rather than a sample. Because Commerce adequately explained how its methodology is reasonable, the Court holds that Commerce’s use of the Cohen’s d test applied as a component of its differential pricing analysis is in accordance with law.
21
CONCLUSION
For the foregoing reasons, Commerce’s remand results are supported by substantial evidence, are in accordance with law, and comply with the Court’s Order, Oct. 22, 2021, ECF No. 51, and are therefore sustained. Judgment will enter accordingly.
/s/ Jennifer Choe-Groves
Jennifer Choe-Groves, Judge
Dated: March 20, 2023 New York, New York