Celanese Chemicals Ltd. v. U.S.

Full title: CELANESE CHEMICALS LTD., Plaintiff, v. UNITED STATES, Defendant, and E.I…

Court: U.S. Court of International Trade

Date published: Jan 29, 2007

Court No. 04-00594, Slip Op. 07-16 (Ct. Int'l Trade Jan. 29, 2007)

Opinion

Court No. 04-00594, Slip Op. 07-16.

January 29, 2007

Greenberg Traurig, LLP (Philippe M. Bruno, Jeffrey S. Neely, Eric Aronson, and Rosa S. Jeong), for Plaintiff.

James M. Lyons, General Counsel, U.S. International Trade Commission (Mark B. Rees); for Defendant.

Crowell Moring LLP (Matthew P. Jaffe, Jeffrey L. Snyder,Alexander H. Schaefer, and Erin E. Mikita), for Defendant-Intervenors.

[Plaintiff's Motion for Judgment Upon Agency Record granted in part; Negative Preliminary Determination remanded to International Trade Commission.]

OPINION

DELISSA RIDGWAY, District Judge

In this action, plaintiff Celanese Chemicals Ltd. contests the negative preliminary determination reached by the U.S. International Trade Commission in an antidumping duty investigation of polyvinyl alcohol imports from Taiwan. See Polyvinyl Alcohol From Taiwan, 69 Fed. Reg. 63,177 (Oct. 29, 2004) (notice of preliminary determination).¹ Pending before the Court are Plaintiff's Motion for Judgment Upon the Agency Record and supporting briefs, in which Celanese contends that the Commission's determination in this matter was "arbitrary, capricious, an abuse of discretion and not in accordance with law," and which requests that the matter be remanded to the agency. See generally Brief in Support of Plaintiff's Motion for Judgment Upon the Agency Record ("Pl.'s Brief"); Reply Brief in Further Support of Plaintiff's Motion for Judgment Upon the Agency Record ("Pl.'s Reply Brief").

1 The Commission's determination, public versions of the views (majority and dissenting), and the staff report in this investigation are published, with business proprietary information redacted, in Polyvinyl Alcohol from Taiwan, Inv. No. 731-TA-1088 (Preliminary), USITC Pub. 3732 (Oct. 2004) ("Preliminary Determination") ( available at https://eofpub.usitc.gov/edis-efile/app as document number 217696). The administrative record consists of three sections, designated "Public" (P.D. ___), "Confidential" (C.D. ___), and "Privileged," respectively. The "Public" section includes copies of documents from the "Confidential" section with all confidential information redacted.
The publicly available version of the Commission's views included in the preliminary determination is cited as the "Preliminary Determination." The confidential version of the Commission's views included in the administrative record as C.D. 65 is cited as "Conf. Comm. Views." And the confidential staff report that accompanied the Commission's findings, which is included in the administrative record as C.D. 34 (original), as amended by C.D. 35 (Mem. INV-BB-127 (Oct. 18, 2004)) and C.D. 43 (Mem. INV-BB-130 (Oct. 21, 2004)), is cited as the "Staff Report."

The Commission and Defendant-Intervenors — E.I. DuPont de Nemours Co. and Chang Chun PetroChemical Co., Ltd. — oppose Celanese's motion, and maintain that the Commission's determination should be sustained in all respects. See generally Memorandum of Defendant United States International Trade Commission in Opposition to Plaintiff's Motion for Judgment on the Agency Record ("Def.'s Brief"); Memorandum of Points and Authorities in Opposition to Motion for Judgment on Agency Record ("Def.-Ints.' Brief").

For the reasons set forth below, Plaintiff's Motion for Judgment Upon the Agency Record is granted in part, and this action is remanded to the Commission for further proceedings not inconsistent with this opinion.

I. The Facts of The Case

As the Commission now sees it, the antidumping investigation of imports of polyvinyl alcohol ("PVA")² from Taiwan at issue here "pitted . . . two . . . [leading] domestic producers — indeed the only producers that supply to the U.S. commercial market — against one another." Def.'s Brief at 2. In its September 2004 petition triggering that investigation, Celanese alleged that the domestic industry — which Celanese sought to define to include only itself — was materially injured or threatened with material injury by reason of imports of PVA from Taiwan.

2 Polyvinyl alcohol ("PVA") is a water-soluble synthetic polymer. As the Commission staff explains: "PVA is used primarily as an intermediate product in the production of PVB [polyvinyl butyral], which is an adhesive used in the manufacture of automotive safety glass and load-resistant architectural glass. PVA is also used in the textile and paper industries in sizing formulations; as a binder in adhesive and soil binding formulations; and as an emulsion or polymerization aid in colloidal suspensions, water-soluble films, cosmetics, and joint compounds." Staff Report at I-6.

A. Prior Antidumping Investigations

PVA had been the subject of antidumping investigations before. The first antidumping petitions against PVA imports were filed in 1995, by Air Products and Chemicals, Inc. (the predecessor of Celanese). The Commerce Department subsequently published antidumping duty orders against PVA imports from China, Japan, and Taiwan, in late March 1996. However, those orders were revoked effective May 14, 2001, due to domestic producers' failure to participate in the five-year sunset reviews. Staff Report at I-3.

In early September 2002, Celanese and DuPont filed a second series of antidumping petitions — this time, against PVA imports from China, Germany, Japan, and Korea. The 2002 petitions resulted in the issuance of an antidumping duty order against PVA imports from Japan in early July 2003. Antidumping duty orders against PVA imports from China and Korea followed three months later. Staff Report at I-3 to I-4. The Commission incorporated "certain public factual findings and analysis from these previous investigations . . . including information about the product, purchasing behavior, the domestic and foreign producers, and other conditions of competition . . . into the record of this investigation." Preliminary Determination at 7; Conf. Comm. Views at 10.

In this action, Celanese makes much of the fact that the Period of Investigation ("POI") for the proceeding here at issue (full calendar years 2001, 2002, and 2003, and the first half of 2004) overlaps for two years with the 2003 investigations (which covered full calendar years 2000, 2001, and 2002). Pl.'s Brief at 6. The Commission and Defendant-Intervenors, in turn, emphasize the differences between the two investigations — the year and a half that do not overlap, as well as the different subject countries. Def.'s Brief at 61-64; Def.-Ints.' Brief at 45-49.

B. The Investigation At Issue

The instant action involves an antidumping petition against PVA³ imports from Taiwan, which Celanese filed in early September 2004, and which fellow domestic producer DuPont opposed. As the initial step in a preliminary injury investigation, the Commission first defines the "domestic like product" and the "industry." The Commission declared a "single domestic like product defined coextensively with [the] scope of the investigation" to consist of "all PVA hydrolized in excess of 80 percent, whether or not mixed or diluted with commercial levels of defoamer or boric acid, but not including PVA in fiber form." Preliminary Determination at 7-8, 11; Conf. Comm. Views at 10-11, 15.

The Commission defined the domestic industry as all producers of the domestic like product: petitioner Celanese; E.I. du Pont de Nemours and Co. (DuPont); and a third producer, Solutia Inc. (Solutia). Preliminary Determination at 11; Conf. Comm. Views at 10, 16. Only one firm, defendant-intervenor Chang Chun PetroChemical Co., Ltd. ("CCPC"), is known to have produced PVA in Taiwan during the period of investigation. Preliminary Determination at 8; Conf. Comm. Views at 10-11.

The Commission did not exclude any party under the related parties provision of the Act, 19 U.S.C. § 1677(4)(B) (2000). Preliminary Determination at 12-14; Conf. Comm. Views at 17-20. The Commission considered whether to exclude DuPont who, in addition to producing PVA domestically, also imported subject PVA from Taiwan during the period of investigation. Celanese argued that DuPont should be considered a related party due to its imports of PVA from Taiwan and its relationship with CCPC. However, while the Commission found DuPont was a related party by virtue of its imports of PVA from Taiwan, it did not find any direct or indirect control relationship between DuPont and CCPC. Preliminary Determination at 13; Conf. Comm. Views at 18. In the end, the Commission concluded it was not appropriate to define the domestic industry to exclude DuPont, based on, inter alia, the company's "share of domestic production, the ratio of its imports to such production, and the impact of its importing on its domestic production operations." Def.'s Brief at 11 ( citing Conf. Comm. Views at 17-20).

Of the three known producers of PVA in the United States, plaintiff Celanese is the [_____] domestic producer of PVA; defendant-intervenor DuPont is the [_____] U.S. producer; and Solutia produces PVA as part of a vertically integrated production process of PVB, a downstream product of PVA. Pl.'s Brief at 9 ( citing Conf. Comm. Views at 20 n. 85); see also Preliminary Determination at 14 n. 85.

In the course of its investigation, the Commission staff collected questionnaire responses from all the domestic producers. Preliminary Determination at 7; Conf. Comm. Views at 10. The staff also collected data from importer questionnaire responses (which they estimate covered more than 90% of PVA imports from Taiwan), as well as from questionnaire responses from CCPC, the Taiwanese producer. Staff Report at I-2, Tables C-1 and C-2. Indeed, the only information the Commission claims it was unable to obtain is certain financial information from Celanese, notwithstanding repeated requests. Preliminary Determination at 4; Conf. Comm. Views at 5; Staff Report at VI-4 to VI-7. Celanese disputes that claim, asserting that it responded to all requests for information. Pl.'s Reply Brief at 24 n. 71. In any case, Celanese says its quarrel is with the Commission's analysis of the information it has on record, and that it has never claimed the lack of any cost information as a reason to continue the investigation. Id.

In brief, the Commission found that the record as a whole contained clear and convincing evidence that the domestic PVA industry was not materially injured or threatened with material injury. Of the three domestic producers, only two supply the commercial market: Celanese, the plaintiff; and Defendant-Intervenor DuPont, who opposes the petition and imports certain PVA from Taiwan. Preliminary Determination at 4-5; Conf. Comm. Views at 5-7.

The Commission found that the volume of subjects imports was significant, but found no evidence to support significant price effects, adverse impact, or threat. Although subject imports increased over the POI, so did domestic consumption. The Commission also found that the domestic industry's share of the merchant and total U.S. PVA markets increased, and — throughout most of the POI — non-subject imports accounted for a larger market share than subject imports. According to the Commission, the record also indicated no significant price underselling by subject imports; nor was any significant price depression evident.⁴ Moreover, despite evidence of an industry cost-price squeeze, the Commission found no significant price suppression. Preliminary Determination at 4-5; Conf. Comm. Views at 5-7.

4 The Commission based its depression finding on the fact that the domestic industry had been able to raise prices in the more recent period despite increases in the volume of subject imports. Conf. Comm. Views at 6.

In finding no evidence of adverse impact, the Commission noted that there had been declines and improvements in the domestic industry's performance factors. "Between 2002 and 2003, when subject imports . . . experienced their largest relative volume increases, domestic producers gained some market share, increased production, increased their capacity utilization, increased U.S. shipments, continued to experience declining inventories, and experienced increased domestic unit sales values." Preliminary Determination at 5; Conf. Comm. Views at 6-7.

After antidumping duty orders were imposed on PVA imports from China, Korea, and Japan in mid to late 2003, "the domestic industry's performance for interim 2004 was at levels that were better than or similar to levels in interim 2003 for many of these same factors" despite the increasing subject imports from Taiwan. Preliminary Determination at 5; Conf. Comm. Views at 6-7.

Celanese charges that the Commission "miscalculated the subject and non-subject import volumes" in such a way as to overstate the volume of non-subject imports, while understating the volume of subject imports. Pl.'s Brief at 12. Celanese also takes issue with the data the Commission relied on to find: a) no significant pattern of underselling; b) attenuated competition between subject imports and domestically produced PVA; and c) no price effects. Id. Celanese asserts that a) the data for underselling was not credible; b) the attenuated competition finding was based on a misreading of customer questionnaire responses; and c) the price effects finding was based on the above mentioned faulty volume calculations and flawed pricing information. Id.

Further, Celanese argues that the Commission applied an improperly high threshold of causation in its injury analysis. And, finally, Celanese argues that the Commission "improperly relied on, inter alia, self-contradictory and unexplained evidence" from the Taiwanese producer, CCPC, to find no threat of material injury. Pl.'s Brief at 13.

The Commission and Defendant-Intervenors, in turn, vigorously dispute Celanese's allegations, characterizing them as an attempt to get the Court to reweigh the evidence in a manner more favorable to Celanese. See Def.-Ints.' Brief at 17-19; Def.'s Brief at 3 (describing Celanese's claims as "largely factual in nature, raising the sort of evidence-weighing questions appropriate for consideration by the Commission in its capacity as fact-finder but not for this Court in its review function"), 47, 72.

II. The Procedural History of The Case

In its petition filed with the U.S. Department of Commerce and with the Commission, Celanese alleged that an industry in the United States was materially injured and threatened with further material injury by reason of less than fair value imports of polyvinyl alcohol ("PVA") from Taiwan. Complaint ¶ 5; see also Polyvinyl Alcohol from Taiwan, 69 Fed. Reg. 55,653 (Sept. 15, 2004) (notice of receipt of petition and initiation of investigation); Polyvinyl Alcohol from Taiwan, USITC Pub. 3732, Inv. No. 731-TA-1088 (Prelim.) (Oct. 2004) ("Preliminary Determination").

Upon receipt of Celanese's petition, the Commission initiated its investigation, Polyvinyl Alcohol From Taiwan, Inv. No. 731-TA-1088 (Preliminary). See Polyvinyl Alcohol from Taiwan, 69 Fed. Reg. 55,653 (Sept. 15, 2004). Within a matter of weeks, the Commission held a preliminary conference, after which parties to the proceeding filed post-conference briefs with the Commission. Pl.'s Brief at 4.

Upon completion of the preliminary investigation, the Commission determined by a 3-2 vote that there was no reasonable indication that an industry was being materially injured or was threatened with material injury within the meaning of 19 U.S.C. § 1673b(a) by reason of subject imports of PVA from Taiwan. See Polyvinyl Alcohol from Taiwan, 69 Fed. Reg. 63, 177 (Oct. 29, 2004) (notice of preliminary determination). Accordingly, the agency investigations were terminated. See 19 U.S.C. § 1673b(a)(1) ("If the Commission finds that imports of the subject merchandise are negligible or otherwise makes a negative determination under this paragraph, the investigation shall be terminated.").

III. Standard of Review

In reviewing a challenge to a negative preliminary injury determination by the Commission in an antidumping case, "[t]he court shall hold unlawful any determination, finding, or conclusion found . . . to be arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law. . . ." 19 U.S.C. § 1516a(b)(1)(A); see also Co-Steel Raritan, Inc. v. United States, 357 F.3d 1294, 1309 (Fed. Cir. 2004). The "arbitrary and capricious" standard requires that the Commission "examine the relevant data and articulate a satisfactory explanation for its action including a `rational connection between the facts found and the choice made.'" Mot. Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983) ( citing Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168 (1962)).

IV. Analysis

Clear and Convincing Evidence Standard

In a preliminary injury determination, the Commission "has consistently viewed the statutory `reasonable indication' standard as one requiring that it issue a negative determination . . . only when (1) the record as a whole contains clear and convincing evidence that there is no material injury or threat of such injury; and (2) no likelihood exists that contrary evidence will arise in a final investigation." Am. Lamb Co. v. United States, 785 F.2d 994, 1001 (Fed. Cir. 1986). Material injury is defined as "harm which is not inconsequential, immaterial, or unimportant." 19 U.S.C. § 1677(7)(A) (1996).

Celanese disputes a number of the Commission's individual findings as not supported by clear and convincing evidence. See Pl.'s Brief at 28 (arguing that Commission's findings on pricing for certain products "are not supported by clear and convincing evidence on the record as a whole"); see also Pl.'s Brief at 27, 59, 63; Pl.'s Reply Brief at 2-5, 17, 19, 20-23, 28-29. However, Celanese misstates the standard for a negative preliminary determination. It is not that each piece of evidence must be based upon clear and convincing evidence, but — rather — that "the record as a whole contain clear and convincing evidence that there is no material injury." Conn. Steel Corp. v. United States, 18 CIT 313, 315-316, 852 F. Supp. 1061, 1064 (1994) ( citing Am. Lamb Co., 785 F.2d at 1001).

Celanese also identifies numerous findings which it asserts are based on less than comprehensive evidence or for which contradictory evidence exists, and argues that those findings inherently fail to meet the clear and convincing evidence standard and, thus, require further investigation. See Pl.'s Brief at 27-28, 34, 67; Pl.'s Reply Brief at 17 (arguing that any finding based on evidence "twice contradicted . . . cannot meet the clear and convincing standard);⁵ see also id. at 3-5, 7, 11-13, 28-30. Leaving aside the fact that not every piece of evidence must meet the clear and convincing standard, it is assuredly within the Commission's authority to "weigh all the evidence and resolve conflicts in the evidence." Conn. Steel Corp., 18 CIT at 315, 852 F. Supp. at 1064; see also Am. Lamb Co., 785 F.2d at 1002-04.

5 Celanese is referring to information about lost sales/lost revenue described by the Commission as "consistent with other evidence offered by DuPont." Pl.'s Brief at 17 ( citing Def.'s Brief at 47, quoting Conf. Comm. Views at 12). Celanese claims that it is unclear to what evidence the Defendant's Brief is referring, and suggests the reference may be to either testimony by Kathy McCord at the preliminary conference (which it claims would then be contradicted by her testimony in 2003) or to the transaction report offered by DuPont as an appendix to its Post-Conference Brief (which Celanese claims supports its contention that price is "the major competitive factor and thus thoroughly contradicts McCord's conference testimony"). Id.
A closer reading suggests that it is not so difficult as Celanese suggests to determine the evidence to which the Commission is referring. Defendant's Brief cites this comment to the Conf. Comm.'s Views at 40 which lists examples of such evidence offered by DuPont regarding non-price factors including its questionnaire responses and specific Post-Conference Brief Exhibits which include its transactions reports among other sources. Def.'s Brief at 47 ( citing Conf. Comm. Views at 40 n. 179 ( citing Staff Report at II-13 to II-14, V-12 to V-14; Preliminary Determination at II-8 to II-9; DuPont Post-Conference Brief, C.D. 19, at Exhibits 5, 6, 7)). This issue is addressed more fully below.

Causation

In a preliminary injury determination, the Commission is required to determine, based on the information available to it at the time of the determination, whether there is a reasonable indication that an industry in the United States is (i) materially injured or (ii) threatened with material injury, by reason of imports of the subject merchandise. 19 U.S.C. § 1773b(a). Celanese would argue that this means that unless the effect of subject imports is found to be tangential, incidental or trivial, the Commission must find causation. Pl.'s Brief at 60 ( citing Nippon Steel Corp. v. United States, 345 F.3d 1379, 1381 (Fed. Cir. 2003) ("{D}umping need not be the sole or principal cause of injury. As long as its effects are not merely incidental, tangential, or trivial, the foreign like product meets the causation requirement.")).

Celanese adds that the legal standards applied by the Commission are "directly contrary" to the Statement of Administrative Action ("SAA") accompanying the Trade Agreements Act of 1979, which specifically states that "the Commission is not to `weigh causes.'" Pl.'s Brief at 61 ( citing Statement of Administrative Action, Trade Agreements Act of 1979, S. Rep. 96-249 at 75 (July 1979)).

Celanese is correct that it does not need to show that the subject imports are the sole, or even the major, cause of injury. However, its selective quotations from the Commission's Preliminary Determination — for example, "`while price is an important factor, other factors were important' . . . `factors other than price were relevant' . . . factors other than subject imports `contributed materially'" — do not show that the Commission failed to apply the proper standard of causation, much less that the Commission "raised the threshold for finding material injury . . . far beyond that required by law" by "improperly attributing too much importance to factors other than the substantial volumes of subject imports that were being sold at less-than-fair-value." Pl.'s Brief at 60 ( quoting Preliminary Determination at 32, 39, 47).

Moreover, Celanese's reliance on Nippon is misplaced. Nippon itself cites Gerald Metals, Inc. v. United States, 132 F.3d 716, 721-23 (Fed. Cir. 1997), which points out that the Federal Circuit has never adopted the SAA "non-de minimus" methodology and that, in fact, by-reason-of analysis may require the weighing of factors. See Gerald Metals, 132 F.3d at 722 (finding that "evidence of de minimis causation of injury does not reach the causation level required under the statute"). Contrary to Celanese's assertion, the Commission does not have to state outright that it found subject imports to be "incidental, tangential, or trivial." See Pl.'s Reply Brief at 28.

A. Volume of Imports

Under the statutory regime, the Commission must determine the significance of the volume of subject imports from Taiwan, including any increase in that volume, both in absolute terms and relative to production and consumption in the United States. 19 U.S.C. § 1677(7)(C)(i). Although the Commission did find the actual volume of subject imports and the increase in that volume over the period of investigation to be significant, it nonetheless found volume did not have a significant impact on the domestic market. Preliminary Determination at 22; Conf. Comm. Views at 33. In summary, the Commission found 1) U.S. market consumption increased over the POI; 2) the domestic industry's share of the market increased; and 3) non-subject imports of PVA held a larger market share than subject imports from Taiwan for most of the POI but lost market share after becoming subject to the 2003 antidumping orders. Preliminary Determination at 22; Conf. Comm. Views at 33-34. Moreover, the Commission found that despite the decline in non-subject imports towards the end of the POI, "subject imports did not seem to take advantage of this opportunity." Preliminary Determination at 23; Conf. Comm. Views at 34. To support this statement, the Commission explained that market share levels for subject imports in interim 2004 and interim 2003 were similar and not much higher than their market share in 2001, "when they were covered by an antidumping duty order for the first five months of the year." Preliminary Determination at 23; Conf. Comm. Views at 34.

Celanese disputes the data on which the Commission relied to calculate the volume of subject and non-subject imports and their respective market shares, as well as the domestic industry's market share performance, asserting that the Commission both understates the volume of subject imports and overstates the volume of non-subject imports. See generally Pl.'s Brief at 16-24. Celanese argues that properly calculated the data indicates subject imports overtook non-subjects imports toward the end of the investigation (after the antidumping duty order was imposed on China, Korea and Japan in mid to late 2003) and that U.S. producer share did not increase but in fact remained static. In the view of Celanese the "corrected" data, discussed infra, indicates subject imports did in fact take advantage of the opportunity presented by the antidumping orders for China, Japan, and Korea in mid to late 2003 and the U.S. producer share remained static — a very different trend than that cited by the Commission to support its finding that volume of subject imports did not have a significant effect on the domestic industry. Pl.'s Brief at 16-24.

1. Subject Imports

To determine the volume of subject imports, the Commission used importer questionnaire data. Celanese claims that the Commission erred by calculating the volume of subject imports (that is, imports from Taiwan) using importer questionnaire data that was "less than complete." See generally Pl.'s Brief at 2, 12, 18-19; Pl.'s Reply Brief at 5-6, 9-10. Celanese asserts that the importer questionnaire data "understated [subject imports] by about 10 percent," and contends that the Commission should have used official Census statistics or data from CCPC's foreign producer questionnaire response instead.⁶ Pl.'s Brief at 18-19; Pl.'s Reply Brief at 9-10.

6 Celanese reasons that, because CCPC is the only known Taiwanese producer of subject merchandise, CCPC's shipment data represented 100% of subject imports. See Pl.'s Brief at 19.

According to Celanese, the Commission's use of the importer questionnaire data to calculate the volume of subject imports — together with the agency's use of unadjusted data to calculate the volume of non-subject imports (discussed in section A.2, below) — had a decisive impact on some of the Commission's key findings. See generally Pl.'s Brief at 2-3, 12, 19-24, 48, 64-65; Pl.'s Reply Brief at 5-6, 9-10, 30.

As a threshold matter, there is no presumption favoring the use of official government import statistics such as Census data, or — for that matter — any other set of data. The use of importer questionnaire data to calculate subject import volumes is a well-established and accepted practice. See, e.g., Allura Red Coloring from India, Inv. Nos. 701-TA-433 (Preliminary) and 731-TA-1029 (Preliminary), Pub. No. 3595 at 10 (April 2003), aff'd, Sensient Techs. Corp. v. United States, 28 CIT ___, 2004 WL 2030258 (2004) (noting Commission reliance on importer questionnaire data in import volume calculations); Blast Furnace Coke from China and Japan, Inv. Nos. 731-TA-951-952 (Preliminary) (Remand), Pub. No. 3619 at 19-20 (Aug. 2003) ( citing Blast Furnace Coke from China and Japan, Inv. Nos. 731-TA-951-952 (Preliminary), Pub. No. 3444, Tables IV-2, IV-6, and C-1 (Aug. 2002), aff'd, Comm. for Fair Coke Trade v. United States, 28 CIT ___, ___, 2004 WL 1615600 at * 16 (2004) (sustaining Commission reliance on importer questionnaire data in import volume calculations). There is thus nothing inherently wrong with the Commission's reliance on importer questionnaire data in its subject import volume calculations in this case. See generally Def.-Ints.' Brief at 26-27; see also Def.'s Brief at 25.

Moreover, Celanese overstates its case when it claims that the importer questionnaire data "understated [subject imports] by about 10 percent." See Pl.'s Brief at 19. In fact, the Staff Report is clear that the importer questionnaire data actually "account[ed] for well over 90 percent" of the imports from Taiwan. See Staff Report at IV-1 (emphasis added); see also id. at I-2. Thus, the gap is not as great as Celanese tries to suggest. See generally Def.-Ints.' Brief at 25-26; Def.'s Brief at 26.⁷

7 The Government finds it telling that "Celanese opts not to show the Court the actual scope of coverage" for the importer questionnaire data as compared to the official Census Bureau statistics. The Government emphasizes that the questionnaire data overstated official Census statistics in some years and understated them in others, and that — over the relevant Period of Investigation — "[t]he average for coverage of official statistics by questionnaire data" was much closer to 100% than to 90%. See Def.'s Brief at 26.

Even so, Celanese correctly notes that the coverage varied from [_____] in 2001 to [_____] in interim 2004 and thus the understatement was most pronounced in the period after imposition of the 2003 antidumping duty on imports from China, Korea, and Japan. Pl.'s Reply Br. at 9-10 ( citing Def.'s Brief at 26). Moreover, Celanese asserts that this fact, together with the asserted error in calculating non-subject imports discussed in section A.2 below, is significant enough to reverse some of the Commission's key findings. Id.

Even if it were no greater than 90% (as Celanese has claimed), the questionnaire data's scope of coverage might well be sufficiently comprehensive. See, e.g., Comm. for Fair Coke Trade, 28 CIT at ___ n. 24, ___, 2004 WL 1615600 at * 14 n. 24, ** 15-16 (sustaining Commission's negative preliminary determination where import volume findings were based on importer questionnaire responses accounting for approximately 80% of imports from China and virtually all imports from Japan); Torrington Co. v. United States, 16 CIT 220, 222-23, 790 F. Supp. 1161, 1166 (1992), aff'd, 991 F.2d 809 (Fed. Cir. 1993) (sustaining Commission's negative preliminary determination where impact findings were based on responses of 25 of 51 firms, which accounted for 74% of total shipments by quantity and 68% by value). See generally Def.'s Brief at 26-27; Def.-Ints.' Brief at 27-28.

At bottom, Celanese's challenge to the Commission's calculation of the volume of subject imports is predicated on the premise that, confronted with three somewhat different sets of data, the Commission irrationally ignored two perfect sets of data and instead chose to use the one set of data that was flawed. However, the record indicates that the Commission found none of the three sets of data to be perfect. As Celanese emphasizes, the Commission candidly acknowledged that the scope of coverage of the importer questionnaire data was less than 100%. The Commission nevertheless used those data in calculating subject import volume, based on its determination that the importer questionnaire data were more reliable than the other data that were available (and which Celanese favors). See Staff Report at I-2 n. 5, IV-1 n. 2; see generally Def.'s Brief at 3-4, 27; Def.-Ints.' Brief at 25.

For example, in its opening brief, Celanese contends that — in lieu of the importer questionnaire data — the Commission should have used data from CCPC's foreign producer questionnaire. See generally Pl.'s Brief at 19. But, contrary to Celanese's implication, there were problems with that data as well. Unlike other import data sets, foreign producer questionnaire data do not account for volume in terms of value. And, in terms of quantity, there were only "comparatively minor variances" between the importer questionnaire data that the Commission used and the foreign producer questionnaire data that Celanese advocates. See Def.'s Brief at 27-28. It is therefore perhaps no surprise that Celanese's reply brief appears to abandon its advocacy for use of the foreign producer questionnaire data.

Celanese's principal contention is that the Commission should have calculated subject import volume using Census import statistics, rather than importer questionnaire data. See generally Pl.'s Brief at 19; Pl.'s Reply Brief at 9-10. But the Census statistics, too, were not without problems. As the Staff Report explained, Celanese itself specifically cautioned against using Census data to measure subject import volume in value terms; and DuPont had warned that the Census data reflected clerical errors by its Customs broker (which it was taking steps to correct). See Staff Report at I-2 n. 5, IV-1 n. 2; see generally Def.'s Brief at 27 ( citing Petition, C.D. 1, at 27, 37 n. 104; DuPont's Post-Conference Brief, C.D. 19, at 18-20).

Celanese does not dispute the existence of the problems cited by the Commission. Instead, Celanese asserts that, to the extent relevant, those problems had been cured by the time that post-conference briefs had been filed with the agency. See Pl.'s Brief at 19 n. 45; Pl.'s Reply Brief at 10.

To be sure, the Commission has broad statutory discretion to choose the data on which it bases its determinations. See, e.g., Comm. for Fair Coke Trade, 28 CIT at ___, 2004 WL 1615600 at * 16 (noting discretion accorded to Commission in choosing data sets);American Bearing Mfrs. Ass'n v. United States, 28 CIT ___, ___, 350 F. Supp. 2d 1100, 1108-09 (2004) ("[W]hether to rely primarily on quantity data, value data, or both, to measure the significance of import volume is precisely the type of decision that Congress has entrusted the ITC to make in light of the facts and circumstances of each particular case."). And the Court generally is not at liberty to substitute its judgment for that of the Commission. See, e.g.,Torrington Co., 16 CIT at 226, 790 F. Supp. at 1167 ("[I]t is not the Court's function to decide that it would have made another decision on the basis of the evidence."); Comm. for Fair Coke Trade, 28 CIT at ___, 2004 WL 1615600 at * 3-4; Sensient Techs. Corp., 28 CIT at ___, 2004 WL 2030258 at * 8; Ranchers-Cattlemen Action Legal Found. v. United States, 23 CIT 861, 878, 74 F. Supp. 2d 1353, 1368-69 (1999) (court "cannot substitute its judgment . . . for that of the Commission," but only "ascertain[s] whether there was a rational basis for the [Commission's] determination").

Here, however, there has been no proffer to respond to Celanese's assertion that corrected Census data were available to the Commission at the time it made its preliminary determination, and should have been used to calculate subject import volume. Nowhere, for example, has the Government (much less the agency itself) disputed Celanese's assertion that the errors in the Census data were corrected before the Commission reached its determination. Nor has there been any attempt to justify the Commission's continued reliance on the importer questionnaire data, if indeed corrected Census data were available for use. Cf. Calabrian Corp. v. United States, 16 CIT 342, 351, 794 F. Supp. 377, 386 (1992) (Commission must consider "the `best information available' contained in the record at the time of its determination") ( cited in Def.'s Brief at 21).

Generally, the Commission's use of importer questionnaire data to calculate volume of subject imports would be well within agency's discretion. See Def.'s Brief at 25-28; Def.-Ints.' Brief at 26-28. However, it is not clear that the Commission here relied on the best information available. Calabrian Corp., 16 CIT at 351, 794 F. Supp. at 386 (the "best information available" standard includes all information, even information not in the staff report, in the record at the time of Commission's determination). The Commission has not denied it had the corrected Census data at the time of its decision. On remand the Commission shall explain why the questionnaire responses remained the best information available in light of the apparent corrections to the errors cited as reasons for not using Census data in the first instance.

2. Non-subject Imports

The Commission used official import data reported by the U.S. Census Bureau (Census) for Harmonized Tariff Schedule of the United States (HTSUS) subheading 3905.30.00 to determine the volume of non-subject imports, i.e., imports from countries other than Taiwan. Staff Report at II-9. See generally Pl.'s Brief at 16-17. That subheading includes PVA of all hydrolysis levels, however, while the scope of the current investigation includes only PVA with a hydrolysis level of 80% or higher. Pl.'s Brief at 16. Based on the unadjusted Census data, the Commission found non-subject imports were the second largest source of supply for the U.S. domestic PVA market in 2003 after domestic producers. Preliminary Determination at 17.

Celanese argues that by using data that included out-of-scope products the Commission overstated the volume of non-subject imports which contributed to understating the impact of subject imports from Taiwan on the domestic market. Pl.'s Brief at 16-18. Celanese points out that in 2003, while conducting investigations into PVA from China, Japan and Korea, the Commission adjusted the Census data to exclude non-subject PVA from its calculation of non-subject import volumes. Pl.'s Brief at 17. The Commission did not make any such adjustment in the instant case which Celanese claims distorts the data relied on by the Commission for its negative preliminary injury determination. Id.

The Commission contends that "the total volume of out-of-scope PVA for all non-subject countries was relatively small" and would not have changed its determination. Def.'s Brief at 32. Moreover, it contends that differences in the respective scopes of the investigations justified the adjustment in the 2003 investigations but do not require the adjustment here. Def.'s Brief at 29-31. The Commission notes that the scope of the 2003 investigations included a list of 14 additional specific exclusions making it much narrower. Id. at 29.

Celanese argues that the Commission's reliance on the difference in the scope of the investigations fails on two counts. First, Celanese asserts that the argument is a post hoc rationalization, not mentioned in the Commission's Views. Pl.'s Reply Brief at 9. Second, Celanese claims that the distinctions drawn by the Commission are largely illusory. Celanese points out that the list of 14 specific exclusions consisted entirely of products hydrolized in excess of 80 percent while the adjustment for "out-of-scope" PVA excluded PVA hydrolized at less than 80 percent. Pl.'s Reply Brief at 8. While the Court can imagine reasons why narrowing the scope of an investigation might justify excluding additional out-of-scope products which would otherwise be lumped under an HTSUS subheading, as in the earlier investigation in question, the Commission did not go that extra step and articulate its reasoning here. Bowman Transp., Inc. v. Ark.-Best Freight Sys., 419 U.S. 281, 285-86 (1974) (court may not "supply a reasoned basis for the agency's action that the agency itself has not given") (citations omitted).

Celanese similarly challenges the Commission's claim that the out-of-scope product was negligible and excluding it would not have affected its determination. Pl.'s Brief at 16-18. Comparing the adjusted data for 2001 and 2002 (the overlapping years of the investigation) indicates, according to Celanese, that the volume of non-subject imports for PVA is overstated by approximately [ ] percent and [ ] percent respectively. Pl.'s Brief at 17. Celanese also takes issue with the Commission's contention that it did in fact consider the issue, noting that the only evidence the Commission points to in support of this contention is a single sentence in the Staff Report.⁸ Pl.'s Reply Brief at 6 ( citing Def.'s Brief at 32). The Staff Report does not in fact elaborate on how the quantity of out-of-scope merchandise included in the Census data was estimated. Celanese similarly emphasizes that, in the 2003 investigations, the Commission excluded out-of-scope product from two countries, and argues that out-of-scope product should have been excluded here as well. See Pl.'s Brief at 19-20; Pl.'s Reply Brief at 6. The Commission notes that the relevant data from the 2003 investigations was business proprietary information, and has since been destroyed. See Def.'s Brief at 32. But, logically, that fact has no bearing on whether the out-of-scope product at issue here was negligible, as the Commission claims.

8 The Staff Report states: "An undetermined but relatively small share of imports from all other sources consists of PVA with a hydrolysis level of 80 percent or lower." Staff Report at Table IV-1 n. 1.

Celanese argues that excluding the out-of-scope product would show that subject imports had replaced non-subject imports in 2003 and, thus, "undermine the Commission's finding that non-subject imports were the second largest source of supply in 2003." Pl.'s Reply Brief at 7.

The Commission asserts that Celanese conceded in the 2003 investigations that its injuries were not caused by dumped imports from Taiwan. See Def.'s Brief at 46, 52. In actuality, what Celanese said was that it had "no evidence PVA from Taiwan was being dumped in the U.S. market." Conf. Comm. Views at 25 n. 113; see also Pl.'s Reply Brief at 5 n. 14. Such a statement is different from a positive assertion that it was not being injured. More importantly, one of Celanese's main claims is that subject imports from Taiwan were taking advantage of the antidumping order of 2003 to expand their market in the United States at the expense of the domestic industry's ability to raise prices. Pl.'s Brief at 20 ( citing Celanese Post-Conference Brief, C.D. 24, at 3-9; Petition, C.D. 1, at 38-39). It would seem then that a careful analysis of the volume of subject imports compared to non-subject imports would be essential.

Whether the exclusion of out-of-scope non-subject imports would have ultimately affected the Commission's analysis is not clear, but the Commission failed to provide a satisfactory explanation for its decision to exclude them. On remand the Commission will a) demonstrate precisely how the exclusion of certain products hydrolyzed in excess of 80 percent from the investigation's original scope is related to adjusting for products hydrolyzed at less than 80 percent, already out of the original scope by definition, in such a way as to preclude it in the present investigation; and b) provide support for its conclusion that the volume of out-of-scope, non-subject imports included in the Census data is negligible and would not affect the outcome of the determination.

B. Price Effects

In addition to analyzing volume, the statute also mandates that the Commission consider the "effect of imports . . . on prices in the United States for domestic like products." 19 U.S.C. § 1677(7)(B)(i)(II). In evaluating the price effects⁹ of subject imports as part of a preliminary injury determination, the Commission is required to consider whether there has been significant price underselling by the subject imports, and whether such imports have the effect of depressing or suppressing domestic prices to a significant degree. 19 U.S.C. § 1677(7)(C)(ii).¹⁰ Notably, "{u}nderselling alone is legally insufficient to support an affirmative injury determination." Bic Corp. v. United States, 21 CIT 448, 458, 964 F. Supp. 391, 401 (1997).

In the determination under review here, the Commission found generally that there was no pattern of significant underselling of subject imports, and that subject imports had no significant price depressing or suppressing effects. Conf. Comm. Views at 35-44; Preliminary Determination at 23-27. The quarterly weighted-average unit prices for domestic and imported PVA showed no pattern of significant underselling during the POI, and there was "attenuated competition" between domestically-produced and imported PVA. Preliminary Determination at 23-27; Conf. Comm. Views at 35-44.¹¹

11 While the Commission does not clearly define "attenuated competition," it appears to refer to a lack of overlap in major purchasers of subject imports and of domestically produced PVA, and, thus, a lack of head-to-head competition between subject imports and domestically-produced PVA. See Preliminary Determination at 23-27; Conf. Comm. Views at 35-44.

Celanese contests the Commission's negative findings on underselling, price depression and price suppression. Celanese first argues that the Commission's finding of no significant price underselling contradicts other Commission findings, and is not based on clear and convincing evidence.¹² Specifically, in regards to pricing information data,¹³ Celanese asserts that the Commission erred in (a) using quarterly weighted-average unit prices for products 1 through 3 to examine instances of underselling despite "inconsistencies" in pricing data; and (b) affording less weight to evidence of underselling for products 4 and 5. Pl.'s Brief at 25-35.

12 The Commission concluded that underselling was not significant due to three main findings: (1) underselling by subject imports from Taiwan was not at times or for products that were significant; (2) there was no meaningful overlap in the larger customers served by both the domestic industry and subject imports from Taiwan, nor was there evidence of significant underselling where there was overlap; and (3) the importance of factors other than price. Conf. Comm. Views at 40; Preliminary Determination at 25.

13 To evaluate the price effects of subject imports, the Commission requested U.S. producers and importers of PVA to provide quarterly net U.S. f.o.b. selling value and quantity data for sales to unrelated U.S. customers for five non-specialty PVA products (products 1 through 5) suggested by petitioner that are produced in the United States and imported from Taiwan. Preliminary Determination at 23. Both Celanese (who recommended the pricing products) and DuPont agreed that the products were representative of both the domestic and subject imported PVA products. Id. The Commission calculated quarterly weighted average unit prices for the largest importers and compared them to weighted-average unit prices of commercial sales of U.S. producers for the same quarters. Preliminary Determination at 23.

Celanese also challenges the Commission's attenuated competition finding, on which the Commission relied as further support for its conclusion that underselling was not significant. According to Celanese, that finding was based on an incomplete inquiry. Celanese further argues that the inquiry represented an arbitrary switch from traditional practice, with results directly at odds with lost sales and lost revenue evidence. Pl.'s Brief at 37-38.

1. Underselling

The Commission enjoys great discretion in its selection of a methodology to analyze and assess the evidence of underselling, as long as the choice is supported by substantial evidence. U.S. Steel Group v. United States, 18 CIT 1190, 1218, 873 F. Supp. 673, 699 (1994); see also Hynix Semiconductor, Inc. v. United States, 30 CIT ___, ___, 431 F. Supp. 2d 1302, 1311 (2006) ("{T}he U.S. Court of International Trade . . . has specifically held that the {Commission} possesses broad discretion in selecting methodologies to analyze price effects in particular.") (citations omitted). Specifically, while the statute requires the Commission to analyze underselling, the statute does not stipulate how the Commission is to calculate the price of the imported merchandise or the domestic like product. The Commission thus has broad discretion to choose the manner in which it calculates the prices, and how it compares them. Nitrogen Solutions Fair Trade Comm. v. United States, 29 CIT ___, ___, 358 F. Supp. 2d 1314, 1324 (2005) ( citing Nucor Corp. v. United States, 28 CIT ___, ___, 318 F. Supp. 2d 1207, 1257 (2004)); see also Nippon Steel Corp. v. United States, 19 CIT 450, 466 (1995) (noting the Commission's broad discretion to analyze underselling).

The statute further requires that the Commission consider the domestic market as a whole. Calabrian Corp., 16 CIT at 350, 794 F. Supp. at 386 ("Congress intended the ITC determine whether or not the domestic industry (as a whole) has experienced material injury due to the imports. This language defies the suggestion that the ITC must make a disaggregated analysis of material injury."). The Commission asserts that using weighted-average pricing analysis allowed it to consider the market as a whole.¹⁴ Def.'s Brief at 39. Moreover, the Commission contends that it actually considered disaggregated price data for products 1 through 3. Def.'s Brief at 39. The investigation included measuring "all statutory factors based on any permutation in argument raised by the parties." Def.'s Brief at 40 ( citing Staff Report at Tables E-1 to E-3).

14 The Commission explains weighted-average price as "a measure of a single market price that takes account of differing sales volumes of two or more firms reporting price data for the same product and period. The weighted-average price is construed for a particular product and time period (calendar quarter) from price and quantity data reported by two or more firms in response to questionnaire requests. Arithmetically, it is calculated by summing the reported sales value (price X quantity) of each responding firm for a particular product and quarter, and dividing this aggregate sales value by the sum of the reported quantity of each responding firm for this product and quarter. Conceptually, the calculation averages prices of several reporting firms weighed by the quantities reported by each firm, such that the price associated with a higher-quantity sale will have a greater weight (influence) in the resulting weighted-average price than the price of a lower-quantity sale." Def.'s Brief at 38.

As even Celanese concedes, the substantial deference accorded the Commission in its selection of methodologies to analyze price effects is even greater when, in regards to weighted-average price analysis, the methodology selected by the Commission represents its normal practice. See Hynix Semiconductor, 30 CIT at ___, 431 F. Supp. 2d at 1311; Def.'s Brief at 37-40; Pl.'s Brief at 29 (Commission's reliance on "quarterly weighted-average unit prices . . . was consistent with the Commission's normal practice").

Further, the Commission's use of weighted-average price has been previously sustained by the court. Def.'s Brief at 39 ( citing Nippon Steel Corp., 19 CIT at 466). Celanese nevertheless contends that — normal practice or not — the Commission erred in its use of weighted-average unit prices in the case at bar, because of anomalies in the data. Pl.'s Brief at 25-34; Pl.'s Reply Brief at 11-14.

Comparing domestic and subject import pricing data, using a weighted-average price analysis, the Commission found a "mixed pattern of underselling and overselling by subject imports" during the POI, with an "even number of instances of under- and overselling." Conf. Comm. Views at 36; Preliminary Determination at 23-24. The Commission found that underselling was not significant, noting, among other findings, that the pattern was one of underselling in the earlier part of the period, giving way to overselling in the more recent portion.¹⁵ Conf. Comm. Views at 36; Preliminary Determination at 23-24.

15 Based on a price comparison for five non-specialty PVA products (products 1 through 5), the Commission found that, for products 1 through 3, there were actually more instances of overselling than underselling. Conf. Comm. Views at 36-37 ( citing Staff Report at Tables V-1 to V-6); Preliminary Determination at 24. In addition, for product 1, the Commission noted that early underselling gave way to overselling at the same time that the volume of subject imports was increasing. Products 2 and 3 both had more incidences of overselling than underselling as well. However, while most of the underselling was during the earlier portion of the period of investigation, product 2 was [_____]. Conf. Comm. Views at 36-37; Preliminary Determination at 24.

a. Weighted-Average Prices i. Products 1-3

Celanese claims that, in calculating quarterly weighted-average unit prices of imported PVA for products 1 through 3, the Commission relied on importer data that was inconsistent with the Commission's own finding that PVA is a fungible product, as well as the importer's admission that prices generally do not vary significantly. Pl.'s Brief at 32. Celanese claims such inconsistencies should have led the Commission to use disaggregated data or, "at a minimum," to conduct further inquiry. Pl.'s Brief at 33.

As an initial matter, it is worth noting — in the context of Celanese's challenge to the reliability of information provided by another party — that the credibility of sources is largely a matter within the province of the Commission, as the trier of fact. Al Tech Specialty Steel Corp. v. United States, 27 CIT 1791, 1802 (2003); see also Cleo Inc. v. United States, 30 CIT ___, ___, 2006 WL 2685080 at * 12 (2006) ( citing Nippon Steel Corp. v. United States, 458 F.3d 1345, 1357 (Fed. Cir. 2006)).

It is true the Commission is to use "the best information available" when applying the reasonable indication standard. Am. Lamb Co., 785 F.2d at 1002. That means that the Commission is to "investigate the allegations in as thorough a manner as possible using the information available within that time period, and . . . provide interested parties a reasonable opportunity to present their views." Id. (citing H.R. Rep. No. 317, 96^th Cong., 1^st Sess. 61 (1979)).

Here, the Commission did take steps to confirm the data that Celanese questions. What Celanese discounts as a "routine check," the Commission describes as a "careful review" conducted "per its normal procedure," where it specifically questioned the importer about the points raised by Celanese, in addition to inquiries on several points where the Commission had its own pricing concerns. Staff Report at V-16 n. 53.

The Commission is charged with determining whether there is a reasonable indication of material injury, "not a reasonable indication of the need for further inquiry." Am. Lamb Co., 785 F.2d at 1001. The purpose of the preliminary injury determination is to "eliminate unnecessary and costly investigations which are an administrative burden and an impediment to trade." Am. Lamb Co., 785 F.2d at 1002-03 ( citing S. Rep. No. 1298, reprinted in 1974 U.S.C.C.A.N. 7186, 7308).

There is no reason to assume that the Commission's normal procedures are cursory. Determining whether certain data are "inconsistent" is a factual analysis well-suited for the expertise of the Commission. In the circumstances of this case, Celanese's request that the Court remand on the grounds that data are inconsistent is, in effect, a request that the Court reweigh the evidence, which is decidedly beyond the Court's domain.Timken Co. v. United States, 12 CIT 955, 962, 699 F. Supp. 300, 306 (1988).

Further, the finding that PVA is a fungible commodity is not inherently inconsistent with price differentiations. See Preliminary Determination at 29; Def.-Ints.' Brief at 35. The analysis of interchangeability for like product substitutability may be different from that used in price effects analysis. See Ranchers-Cattlemen Action Legal Found., 23 CIT at 881, 74 F. Supp. 2d at 1371 (upholding, based on the importance of context, Commission decision to treat interchangeability differently in its cumulation analysis than in its like product analysis).

ii. Products 4 5

Celanese contends that the Commission failed to articulate a rational basis for its decision to afford less weight to evidence of underselling for products 4 and 5, and that its decision is therefore arbitrary and capricious, an abuse of discretion and not in accordance with the law. The Commission found that, [_____], product 4 was in the paper sector (which involved less marketing overlap), and the [_____] of product 4 sold was less significant than the products where there was more overselling. Def.'s Brief at 42-43. Celanese further asserts that the Commission wrongfully relied on the "fact" that DuPont had not complained of any adverse price effects to determine that underselling was not significant with respect to product 5. Pl.'s Brief at 34-37; Pl.'s Reply Brief at 14.

Celanese's challenges to the Commission's pricing analysis are without merit. Generally, the Commission has the discretion to weigh evidence, resolve contradictory evidence, and analyze prior findings in light of new circumstances. Comm. for Fair Beam Imports v. United States, 27 CIT 932, 955 (2003). Further, the Commission is presumed to have considered all evidence in the record in making its determination, absent some showing to the contrary. Ranchers-Cattlemen Action Legal Found., 23 CIT at 891, 74 F. Supp. 2d at 1379 ( citing Conn. Steel Corp., 18 CIT at 316, 852 F. Supp. at 1065) (citations omitted).

In this case, in reaching its conclusions as to product 5, the Commission did not rely solely on DuPont's statements, but also specifically considered the "undisputed fact that prices for product 5 were rising in the most recent period." Def.'s Brief at 44 ( citing Conf. Comm. Views at 38-40). Finally, Congress intended that the Commission determine whether or not the domestic industry as a whole has experienced material injury due to the imports. Nucor Corp. v. United States, 414 F.3d 1331, 1336 (Fed. Cir. 2005). The Commission is not required to focus on one portion of the industry by making a disaggregated analysis of material injury. Calabrian Corp., 16 CIT at 350, 794 F. Supp. at 385.

iii. Conclusion

It is worth noting that industry views may be considered as an economic factor. Allegheny Ludlum Corp. v. United States, 287 F.3d 1365, 1375 (Fed. Cir. 2002) ("{T}he industry best knows its own economic interests and, therefore, its views can be considered an economic factor.") ( quoting Suramerica de Aleaciones Laminadas, C.A. v. United States, 44 F.3d 978, 984 (Fed. Cir. 1994)). The statute "authorizes broad evaluation of any relevant factors." Allegheny Ludlum, 287 F.3d at 1375 ( quoting 19 U.S.C. § 1677(7)(B)(ii), the Commission "may consider such other economic factors as are relevant to the determination regarding whether there is material injury by reason of imports"). The Commission therefore was well within its discretion to consider the views of DuPont.

Moreover, "a finding of material injury requires a causal, not merely temporal, connection" between underselling and material injury. Comm. for Fair Coke Trade, 28 CIT at ___, 2004 WL 1615600 at * 20 ( citing Gerald Metals, 132 F.3d at 720). The Commission acknowledged that, "in isolation," the underselling could be viewed as significant. Conf. Comm. Views at 36; Preliminary Determination at 24. However, the Commission found underselling not significant based on the record as a whole, including consideration of non-price factors and the lack of underselling in areas of significant competition overlap. Conf. Comm. Views at 40. Finally, the significance of price effects is not based solely on evidence of underselling. Comm. for Fair Beam Imports, 27 CIT at 952.

In the instant case, the Commission's choice of a weighted-average pricing methodology was reasonable and deserves deference because, inter alia, the Commission routinely uses weighted-average pricing; the Commission's use of weighted-average pricing has been previously sustained on judicial review; and the Commission considered and rejected the use of disaggregated data. On the record before the Court, it cannot be said that the Commission has failed to articulate a rational basis for its pricing analysis.

b. Attenuated Competition Finding

Celanese next asserts that the Commission's finding of attenuated competition between domestic and imported PVA was based on a departure from normal practice, is contradicted by other Commission findings, and is not corroborated by evidence on the record. Specifically, Celanese claims that the finding was based on an incomplete inquiry into customer overlap, and is contradicted by the lost sales and lost revenue evidence. Pl.'s Brief at 37-38.

i. Normal Practice

Celanese first objects that the Commission "inexplicably departed" from its traditional practice of examining evidence of lost sales and revenues to determine the extent of competition, and focused instead on the extent to which imported PVA from Taiwan was sold to the same major customers to which the domestic industry sells. Based on a comparison of the top ten customers reported by domestic producers and subject importers, the Commission found attenuated competition between the domestic market and subject imports.¹⁶ The Commission then relied on that finding as support for its determination that underselling by subject imports was not significant. Celanese argues that, besides reflecting an abrupt departure from routine, the Commission's finding is (a) directly at odds with the lost sales and lost revenue evidence, and (b) based on insufficient representation. Pl.'s Brief at 37-38.

16 The Commission found that evidence showed that, of Celanese's top ten customers, [_____] bought imported PVA from Taiwan, indicating "attenuated competition" and supporting the Commission's finding of a lack of any significant underselling. Preliminary Determination at V-5; Staff Report at V-6.

There is a general rule that an "agency must either conform itself to its prior decisions or explain the reasons for its departure." Citrosuco Paulista, S.A. v. United States, 12 CIT 1196, 1209, 704 F. Supp. 1075, 1088 (1988) ( citing Sec'y of Agric. v. United States, 347 U.S. 645 (1954)) (additional citations omitted). An action by the Commission "becomes `agency practice' when a uniform and established procedure exists that would lead a party, in the absence of notification of change, reasonably to expect adherence to the established practice or procedure." Ranchers-Cattlemen Action Legal Found., 23 CIT at 885, 74 F. Supp. 2d at 1374.

Celanese offers no direct support for the proposition that the Commission's analysis in this case represented an abrupt departure from routine. However, the Commission likewise offers no support for its contention that it "routinely solicits information respecting the major customers of suppliers of products under investigation." Def.'s Brief at 45.

The Commission points to language in the Preliminary Determination indicating that "{c}ompetition among PVA suppliers and its impact on their selling prices can also be affected by the extent to which they sell to the same customers." Id. But it is not clear how that constitutes evidence of routine. The Commission fails to point to any language analyzing customer overlap information for the 10 largest customers of domestic producers and importers in the 2003 investigations, but admits that in 2003 it did consider lost sales and lost revenue evidence in conjunction with evidence of widespread underselling and other factors. Here, the Commission discounts lost sales and revenue allegations as non-dispositive anecdotal evidence. Id. at 45 n. 73, 47-48. But it was precisely such evidence of lost sales and revenue that the Commission relied on in 2003 as "evidence of direct head-to-head price competition." Pl.'s Brief at 42-43 ( citing Preliminary Determination at 40).

While the scope of judicial review of a preliminary injury determination by the Commission is narrow, the agency must nonetheless explain the reasons behind its determination. See Bowman Transp., 419 U.S. at 285-86; Mot. Vehicle Mfrs. Ass'n, 463 U.S. at 43 (citations omitted). In the case at bar, it is impossible to determine from the briefing and the existing record whether the Commission in fact has a normal practice, much less whether it departed from that practice in the proceeding at issue (and, if so, whether the departure was warranted). Remand is warranted to afford the Commission an opportunity to amplify its rationale to further address Celanese's claim.

ii. Lost Sales

Celanese argues that it did in fact lose sales due to price effects of subject imports. See Pl.'s Brief at 41-43. The Commission, however, discounts these instances as attributable to other factors. See Def.'s Brief at 46-50. According to the Commission, in one instance, the purchaser indicated that there were reasons other than price for its decision to buy from a different supplier; and a lost sale to a second purchaser occurred in [_____], when (according to the Commission) the volume of subject imports from Taiwan was declining. Def.'s Brief at 46.

The Commission is affirmatively obligated to undertake a "thorough" investigation based on the information available to it prior to making its preliminary determination. Am. Lamb Co., 785 F.2d at 1003. The requirement to use the best information available obliges the Commission to reasonably seek all information that is accessible or obtainable from whatever source. Id. However, there is no statutory provision requiring the Commission to analyze lost sales in a particular manner. The Commission "may make such an evaluation on whatever rational basis it chooses." Maine Potato Council v. United States, 9 CIT 293, 302, 613 F. Supp. 1237, 1245 (1985).

Nonetheless, a review of the record shows that the first purchaser specifically replied "yes" on the questionnaire when asked whether price was the reason for switching. Pl.'s Brief at 39 ( citing Staff Report at V-38). It is difficult to reconcile that fact with the Commission's assertion that price was not the relevant factor in that lost sale. As for the second lost sale, Celanese asserts that, although subject import volume might have declined for [_____], it did not decrease during [_____] when the sale at issue was lost. Pl.'s Brief at 40. The Commission does not address this charge in its brief; the cited section of the Commission's Views does not break down the data by quarters. Def.'s Brief at 46; see Staff Report at IV-1.

iii. Customer Overlap

Celanese asserts that the Commission's finding of attenuated competition — based on lack of extensive overlap among the top ten customers — was non-representative, and is similarly at odds with the record as a whole. According to Celanese, the Commission's methodology masked the significant competition in the remaining commercial transactions which constituted a substantial portion of Celanese's customer base. Pl.'s Reply Brief at 16. Celanese points out that the survey data show significant competition between domestic and imported PVA beyond the top ten customers focused on by the Commission. Pl.'s Brief at 40. Four of the 17 purchasers surveyed stated they had shifted for price reasons, and 10 reported that their U.S. source had reduced its price to compete with lower priced subject imports. Pl.'s Brief at 40.

iv. Conclusion

Remand is appropriate where an agency has not articulated a rational connection between the facts found and the choices made. See Burlington Truck Lines, 371 U.S. at 168. Here, the Commission has not adequately articulated its reasons for finding that competition between the subject imports from Taiwan and domestic PVA is attenuated; indeed, as noted above, it is not entirely clear from the Preliminary Determination how the Commission is defining "attenuated competition."

On remand, the Commission shall further explain how it defines attenuated competition, as well as the relevant analytical methodology. The Commission shall state with specificity the factors underlying its finding of attenuated competition in this proceeding, and explain specifically why it is reasonable to discount the significant competition outside the top 10 customers, particularly since 14 of the 17 purchasers surveyed cited price competition as important.

c. Non-price Factors

Celanese also argues that the Commission provided no "non-price related explanation" as to why purchasers would choose subject imports over domestic PVA. Pl.'s Brief at 27. In addition, Celanese asserts that the Commission's finding on the "importance of price in the context of analyzing . . . substitutability" contradicts its analysis of the importance of non-price factors. Pl.'s Brief at 44.

However, the Commission did point to specific factors — such as diversity of suppliers, quality, and prequalification — as non-price considerations that purchasers took into account. Preliminary Determination at 25. The Commission also pointed to evidence that "the PVA industry has become increasingly global in nature and that PVA prices have converged across different regions and applications as large multinational firms have greater access to price information and are able to secure global contracts for their PVA needs." Id. at 26 ( citing CCPC Post-Conference Brief, C.D. 21, at 6). Moreover, the Commission never said substitutability was complete.¹⁷ It was well within the discretion of the Commission to assign different weights to different pieces of evidence. Comm. for Fair Beam Imports, 27 CIT at 955.

17 In fact, the Commission stated: "The degree of substitution in demand between PVA produced in the United States and that imported from Taiwan depends upon such factors as relative prices, types of customers, conditions of sales, purchaser supply requirements, and product differentiation. Product differentiation depends on factors such as the range of products, quality, availability, reliability of supply, and the market perception of these latter three factors. Based on the reported information in this investigation, we find there is substitutability in demand between the PVA produced domestically and that imported from Taiwan, but some reported product differentiation and other differences may limit the degree of this demand substitution." Preliminary Determination at 20.

Celanese notes evidence provided by DuPont and evidence from the 2003 investigations (which the Commission claims corroborates its finding on the importance of factors other than price), and argues that the evidence in fact shows price to be the most important factor. Pl.'s Brief at 45.

DuPont submitted a number of internally prepared "transaction reports" documenting [_____]. DuPont Post-Conference Brief, C.D. 19. In the vast majority of those transactions, DuPont attributed the lost sale to price. Celanese claims that the transaction reports contradict the testimony of DuPont's witness concerning the trend toward diversifying supply sources as accelerating and not price-driven. Pl.'s Brief at 46 ( citing Conference Transcript, P.D. 39, at 71). Celanese also claims this testimony is undermined by the same witness's statement under oath in the 2003 investigations that "{b}y far the single most important factor is price and then price and then price again. Once we meet the subject import price, these other non-price factors may become a tie-breaker in a buying decision but if we don't meet the prices, we lose the business." Pl.'s Brief at 47 ( citing Polyvinyl Alcohol from China and Korea, Inv. No. 731-TA-1014 and 1017 (Final), USITC Pub. 3634 ("PVA from China and Korea") at 32 (Sept. 2003)). Celanese further emphasizes that, while the Commission in its 2003 investigation noted the non-price factors on which it now relies, the Commission found price to be much more important then. Id.

As a general matter, however, the Commission has the discretion to weigh evidence, resolve contradictory evidence and analyze prior findings in light of new circumstances. Conn. Steel Corp., 18 CIT at 315, 852 F. Supp. at 1064; see also Am. Lamb Co., 785 F.2d at 1002-04. "It is not within the Court's domain either to weigh the adequate quality or quantity of the evidence for sufficiency or to reject a finding on grounds of a differing interpretation of the record." Timken Co., 12 CIT at 962, 699 F. Supp. at 306. Celanese's challenges to the Commission's findings on non-price factors must therefore be rejected.

2. Price Depression

In the proceeding at issue, the Commission found no price depression. The Commission explained that, while there were declines in PVA prices in the U.S. market over the POI, those declines largely occurred in the earlier portion of the period (between 2001 and 2002) when subject imports from Taiwan were declining. Preliminary Determination at 26. The Commission further noted that "{g}enerally prices began to increase or stabilized during the latter part of the period of investigation, notwithstanding an increase in the volume of subject imports between 2002 and 2003." Id. ¹⁸

18 The Commission also noted that domestic producers had been able to increase prices in the recent period. Def.'s Brief at 51 ( citing Conf. Comm. Views at 41-42). Celanese claims this finding is contradicted by evidence showing that prices have not risen, and that the Commission does not dispute the fact that Celanese, unlike DuPont, had been unable to raise prices. Pl.'s Reply Brief at 19. The evidence in the record is mixed. It appears that Celanese was able to raise prices [_____], but not enough to cover increased costs. Preliminary Determination at 26-27; Conf. Comm. Views at 43; see also Pl.'s Brief at 51-52, 56. Based on the evidence available in the record, the importance of such mixed ability to increase prices to the Commission's finding that there was no price depression is unclear. On remand, the Commission shall explain in greater detail how the evidence supports its finding. Staff Report at V-17 to V-26.

Celanese contends that, in making its finding, the Commission relied on the same flawed questionnaire data concerning subject imports discussed in section A.1 above, and asserts that Census data show a much less pronounced decline. Pl.'s Brief at 49. Celanese has a compelling argument regarding the price trends found by the Commission. While the Commission apparently viewed prices as generally stabilizing or increasing toward the end of the investigation, the evidence seems to indicate that they were flat or even decreasing.¹⁹ On remand, the Commission shall elaborate on its view of the relevant evidence, and explain any apparent contradictions and inconsistencies.

19 For example, [_____]. Staff Report at V-17 to V-26.

3. Price Suppression

Celanese also objects to the Commission's finding of no significant price suppression. While the Commission did find evidence of a cost-price squeeze, it did not find reasonable evidence that it was a result of cost suppression by subject imports from Taiwan.²⁰ Preliminary Determination at 27; Conf. Comm. Views at 44. Rather, in addition to its analysis of [_____],²¹ the Commission also focused on evidence that overselling increased toward the end of the period of investigation, when subject import volume was rising, and that factors other than subject imports had an adverse impact on prices. Preliminary Determination at 27; Conf. Comm. Views at 44.

20 A cost-price squeeze occurs when the cost of goods sold ("COGS") exceeds price and the producer is unable to raise the price — that is, when the producer is unable to sell the good for more than it costs to produce it. Nippon Steel Corp., 458 F.3d at 1354.

21 Celanese takes strong exception to the Commission's entire analysis of its cost structure, and asserts that many of Commission's findings regarding its differences are not supported by the record — [_____]. But, as discussed above, the Commission has broad discretion in choosing methodologies, and Celanese merely protests the outcome. Celanese does not contend that the methodology used by the Commission was arbitrary or a surprise. Pl.'s Brief at 55. Evidence on the record shows the Commission fully considered the relevant issues of fact. Def.'s Brief at 55-57.

Celanese disputes the Commission's treatment of cost information, asserting that the Commission failed to articulate a rational connection between its findings on cost information and the importance to production of natural gas (the price of which was rising). Pl.'s Brief at 54. Celanese further asserts that, although the Commission purports to have considered the cost of natural gas in its cost-price squeeze analysis, it actually did not do so. Pl.'s Reply Brief at 21. But the record does not bear out that charge. The Commission specifically took note of the impact of natural gas prices in its price suppression analysis.²² Preliminary Determination at 27; Conf. Comm. Views at 43. And, in any event, the Commission is presumed to have considered all of the evidence on the record, absent a showing to the contrary.Ranchers-Cattlemen Action Legal Found., 23 CIT at 887, 74 F. Supp. 2d at 1376; Conn. Steel Corp., 18 CIT at 316, 852 F. Supp. at 1065. Celanese has failed to make such a showing.

22 Specifically, the Commission noted: "Cost of goods sold as a ratio to sales declined between 2001 and 2002 as natural gas prices fell during a time of declining PVA prices, but the ratio of cost of goods sold to sales increased between 2002 and 2003 and continued to increase between interim 2003 and interim 2004, as increases in prices in the U.S. market were unable fully to keep pace with increasing costs." Preliminary Determination at 27.

Celanese argues that the Commission's findings as to Celanese's [_____], "are simply a function of the methodology the Commission itself chose for accounting for those revenues." Pl.'s Brief at 55. As discussed elsewhere above, the Commission has broad discretion to choose a methodology for analyzing price effects. However, the Commission has not addressed several of Celanese's arguments concerning the rational connection between certain cost structure findings made by the Commission and its finding that subject imports did not contribute to the cost-price squeeze.²³ Pl.'s Brief at 52-56; Def.'s Brief at 54-57.

23 Celanese argues that [_____] — both noted as components of Celanese's cost structure differences — are symptoms evidencing injury, not a cause of injury. Celanese further asserts that the fact that it had [_____] compared to other producers does not discount the fact that those revenues showed [_____] than [_____]. Pl.'s Reply Brief at 22.

The Commission, of course, is not required to make explicit findings with respect to all factors considered. Calabrian Corp., 16 CIT at 350, 794 F. Supp. at 385. However, no court can review what the Commission has not expressed. Altx, Inc. v. United States, 26 CIT 1425, 1426 (2002). This matter must be remanded to allow the Commission to further explain the connection between cost structure factors and its finding of no price suppression. The Commission shall also reconsider its findings on price suppression and depression, in light of any new conclusions it may reach on volume.

C. Impact

Celanese's chief complaint about the Commission's impact analysis is the Commission's decision to consider and compare the [_____] of domestic producers [_____].²⁴ Pl.'s Brief at 56-59. Celanese argues that, because the Commission did not consider cost structure differences in its causation analysis in its 2003 investigation, it may not do so here absent a rational explanation. Pl.'s Brief at 58; Pl.'s Reply Brief at 23-27. Specifically, Celanese asserts that the "Commission's decision to disregard, without explanation or justification, its prior findings was arbitrary." Pl.'s Brief at 58.

24 In sum, the Commission found that the domestic industry's [_____] were largely a result of Celanese's [_____], and that Celanese's lack of profitability was largely attributable to [_____]. Preliminary Determination at 29; Conf. Comm. Views at 47-48.

As a threshold matter, it is unclear to precisely what "findings" Celanese refers. Celanese makes no argument, nor is there any evidence, that the Commission made an affirmative finding in 2003 that cost structure differences were not a cause of injury.²⁵ There is no indication that the Commission changed either its analytic framework or its factual findings.

25 Celanese cites to Usinor for the proposition that "the Commission may not disregard previous findings of a general nature that bear directly upon the current review." Usinor v. United States, 26 CIT 767, 792 (2002). But Usinor is distinguishable on its facts from the situation in this case. The findings at issue in Usinor were broader and more affirmative, in that they referred to a claim that European countries were generally export-oriented — a statement which the Commission had subsequently retracted without explanation. See Comm. for Fair Beam Imports, 27 CIT at 947-48 (discussing the "general nature" standard in Usinor).

"{A}n agency action is arbitrary when the agency offers insufficient reasons for treating similar situations differently." Consolidated Bearings Co. v. United States, 348 F.3d 997, 1007 (Fed. Cir. 2003) (citations omitted). Thus, it is generally true that an agency must conform with prior decisions or explain its departure. Citrosuco, 12 CIT at 1209, 704 F. Supp. at 1087. However, the statute directs the Commission to "evaluate all relevant economic factors," and states that its analysis shall include (though it is not necessarily limited to) a list of certain factors. 19 U.S.C. § 1677(7)(C)(iii). Furthermore, the Commission is required to consider these factors within the context of the "business cycle and conditions of competition that are distinctive" to the industry. Id. Because the statute is silent on how much weight to assign each of these factors, and because the Commission is charged with administering the statute, the Commission's construction of the statute is entitled to Chevron deference. Nucor Corp., 414 F.3d at 1336 ( citing Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984)).

While the Commission's discretion is not absolute, Celanese's argument is largely without merit. It is well-established that the Commission's material injury determinations are sui generis. Nucor Corp., 414 F.3d at 1340 ("Each injury investigation is sui generis, involving a unique combination and interaction of many economic variables; and consequently, a particular circumstance in a prior investigation cannot be regarded by the {Commission} as dispositive of the determination in a later investigation.") (citations omitted). Nor are "findings" in a prior related determination of injury generally dispositive for purposes of subsequent investigations.Citrosuco, 12 CIT at 1217, 704 F. Supp. at 1094 (whether earlier findings of threat materialized or not was irrelevant to the investigation at issue).

The causation analysis in an earlier investigation thus does not set a precedent for causation analysis in a latter investigation of the same product. Caribbean Ispat Ltd. v. United States, 29 CIT ___, ___, 366 F. Supp. 2d 1300, 1305 (2005), rev'd on other grounds, 450 F.3d 1336 (Fed. Cir. 2006). Indeed, due to the "dynamic" nature of the global economy, "the agency's findings and determinations are necessarily confined to a specific period of investigation with its attendant, peculiar set of circumstances." Comm. for Fair Beam Imports, 27 CIT at 944.

As noted above, in examining the impact of imports on the domestic industry, the Commission is required to evaluate all relevant economic factors which have a bearing on the state of the industry in the United States, within the context of the particular period of investigation. 19 U.S.C. § 1677(7)(C)(iii).²⁶ Whether or not the Commission considered cost structure differences in the 2003 investigations is irrelevant. Equally irrelevant is whether any of the parties in the present case argued that Celanese's cost structure differences were critical.²⁷ See Pl.'s Brief at 25-26.

26 Specifically, when examining the impact of imports on the domestic industry, the Commission is required to evaluate all relevant economic factors which have a bearing on the state of the industry in the United States, including, but not limited to: (I) actual and potential decline in output, sales, market share, profits, productivity, return on investments, and utilization of capacity (II) factors affecting domestic prices, (III) actual and potential negative effects on cash flow, inventories, employment, wages, growth, ability to raise capital, and investment, (IV) actual and potential negative effects on the existing development and production efforts of the domestic industry, and (V) in an antidumping duty proceeding, the magnitude of the margin of dumping. The factors must be evaluated within the context of the business cycle and conditions of competition that are distinctive to the affected industry. 19 U.S.C. § 1677(7)(C)(iii)(I)-(V).

27 Celanese disputes the Commission's claim that "no party argued that the cost structure of one of the domestic producers vitiated the entire industry's injury argument in the {2003} investigations," and cites evidence and argument in the 2003 investigations urging the Commission to consider Celanese's "excessive costs." Pl.'s Reply Brief at 25 ( citing Def.'s Brief at 63). However, that evidence merely proposed Celanese's "excessive costs" as one factor to consider. See, e.g., Pl.'s Reply Brief, Att. 3 (Post-Conference Brief of Wego Chemical Mineral Corp. (Oct. 2, 2002)) at 1.

The fact that the Commission decided to analyze cost structure differences in the current investigation, even if it did not in 2003, does not undermine its finding here, because it is within the Commission's discretion to determine the significance of a particular factor based on a reasonable interpretation of the evidence, and because each injury investigation is sui generis. Maine Potato Council, 9 CIT at 300, 613 F. Supp. at 1244 ( citing S.Rep. 249, 96^th Cong., 1^st Sess. at 74-75 (1979), reprinted in 1979 U.S.C.C.A.N. 381, 460-61); Armstrong Bros. Tool Co. v. United States, 84 Cust. Ct. 102, 115, 489 F. Supp. 269, 279 (1980).

A factor that was less important in an earlier investigation may, in the context of a new investigation covering overlapping but different years and a different subject country, become more important. Here, the Commission did not change its analytical framework or discount an earlier finding, but instead weighed the evidence in context and under the unique circumstances of the case as it is required to do.

Celanese cites Citrosuco for the for the proposition that prior decisions "need not be followed if `new arguments or facts are presented that support different conclusions'" but that "the Commission may not act arbitrarily and must explain reasons for any departure." Pl.'s Brief at 58 ( citing Citrosuco, 12 CIT at 1209, 704 F. Supp. at 1087-88). But Celanese has not made the requisite showing. Indeed, the Commission has demonstrated that this case is not factually identical to the prior investigation that Celanese cites.

The current investigation overlapped with the 2003 investigations for the years 2001 and 2002. But it did not overlap during the latter half of the period of investigation, from 2003 through interim 2004.²⁸ Even if the factors were exactly the same for the two overlapping years, the Commission would be well within its discretion to focus on the more recent period. Nucor Corp., 414 F.3d at 1337 ("since the Commission has broad discretion to choose the most appropriate period of time for its investigation, it would be nonsensical to hold that once the Commission had chosen an investigation period, it is required to give equal weight to imports throughout the period it has selected."). Moreover, factors were not exactly the same — for example, the subject countries were different. See Comm. for Fair Beam Imports, 27 CIT at 947 (discussing the importance of different subject countries as a factor in distinguishing material injury investigations).

28 Celanese argues that the different time period is irrelevant because the cost structure was the same in both time periods. However, the Commission is specifically mandated to evaluate factors within the context of the business cycle and conditions of competition that are distinctive to the affected industry. 19 U.S.C. § 1677(7)(c)(iii). Due to changes in industry circumstances, different factors may become more or less important.

In addition, the Commission pointed to differences in volume, price effects, timing, and other economic factors as reasons for differences between its conclusions in the current investigation compared to the 2003 investigations. Preliminary Determination at 30; Conf. Comm. Views at 48-49. Although the Commission did not specify the other economic factors considered beyond a general reference to those "discussed above," so long as the path of the Commission's reasoning may be "reasonably discerned," it is to be sustained. Ceramica Regiomontana, 810 F.2d at 1139 ( citing Bowman Transp., 419 U.S. at 286).

Celanese dismisses as post hoc rationalization the Commission's claim that the "factual records" of the investigations were sufficiently different to justify differences in its analysis. Pl.'s Brief at 23-27. But the Commission's Preliminary Determination specifically enumerated multiple factors distinguishing the investigations. Preliminary Determination at 30; Conf. Comm. Views at 48-49. That the Commission made no mention of the divide in the industry in the specific section of the Preliminary Determination dealing with impact does not nullify the other factual differences pointed out. See Def.'s Brief at 63. The Commission clearly appreciated the divided nature of the industry, which was addressed throughout the Commission's determination.

Celanese also contends that the Commission failed to respond to Celanese's argument regarding other negative profitability trends in the domestic industry.²⁹ Pl.'s Reply Brief at 26-27. The Commission similarly failed to explain the connection between its negative impact finding and the fourth factor regarding exports from the U.S. cited in the Preliminary Determination.³⁰ See Pl.'s Brief at 57 n. 177. On remand, the Commission shall elaborate on these matters, explaining, inter alia, (a) why it found the profitability trend to be not significant, and (b) the significance of the exports noted.

29 Specifically, Celanese argues that, while DuPont might have been profitable, DuPont's [_____]. Pl.'s Brief at 59; Pl.'s Reply Brief at 26-27. Celanese also denies that it failed to provide the Commission with financial information that it requested. Pl.'s Reply Brief at 24.

30 The Commission noted the fact that the domestic industry exported large volumes of PVA at average unit values [_____]. Conf. Comm. Views at 47.

To the extent that the Commission's impact analysis was based on cost structure differences, it is sustained. To the extent that it relied on volume and price effect, the Commission shall reconsider its finding in light of its reconsidered volume and price effect findings on remand.

D. Threat

As a final step in its analysis, the Commission is required to determine whether the domestic industry is threatened with material injury by reason of subject imports.³¹ 19 U.S.C. § 1677(7)(F). The Commission is obligated to consider certain enumerated factors, among other relevant factors, as a whole in assessing the possible threat of imminent injury. But no single factor is necessarily determinative. Moreover, the Commission's determination must not be based on "mere conjecture or supposition." Id.

31 In a threat of material injury determination, the Commission must find whether "further dumped or subsidized imports are imminent and whether material injury by reason of imports would occur unless an order is issued." 19 U.S.C. § 1677(7)(F)(ii). The relevant factors outlined in the statute for a finding of threat of material injury include, but are not limited to: (II) any existing unused production capacity or imminent, substantial increase in production capacity in the exporting country indicating the likelihood of substantially increased imports of the subject merchandise into the United States, taking into account the availability of other export markets to absorb any additional exports, (III) a significant rate of increase of the volume or market penetration of imports of the subject merchandise indicating the likelihood of substantially increased imports, (IV) whether imports of the subject merchandise are entering at prices that are likely to have a significant depressing or suppressing effect on domestic prices, and are likely to increase demand for further imports, (V) inventories of the subject merchandise, (VIII) the actual and potential negative effects on the existing development and production efforts of the domestic industry, and (IX) any other demonstrable adverse trends that indicate the probability that there is likely to be material injury by reason of imports of the subject merchandise (whether or not it is actually being imported at the time). 19 U.S.C. § 1677(7)(F)(i).
Further, the Commission must evaluate the statutory factors "as a whole" in making a threat determination; and no particular factor will "necessarily give decisive guidance with respect to the determination." 19 U.S.C. § 1677(7)(F)(ii).
As the Commission noted here, statutory threat factor (I) is inapplicable, because no countervailable subsidies are involved; statutory threat factor (VI) is inapplicable, as there is no evidence of production facilities in Taiwan that are currently being used to produce other products that can be used to produce the subject merchandise; and statutory threat factor (VII) is inapplicable, because no imports of agricultural products are involved. Preliminary Determination at 31 n. 206; Conf. Comm. Views at 50 n. 206.

Here, the Commission made explicit findings on five relevant mandatory factors: (1) production capacity; (2) volume or market penetration of subject imports; (3) the suppressing or depressing effects of subject imports on domestic prices; (4) inventories of the subject merchandise; and (5) negative effects on development and production of the domestic industry. Def.'s Brief at 67-72 ( citing Conf. Comm. Views at 50-53). Celanese challenges the Commission's findings on capacity, volume, price effects, and inventories.

1. Production Capacity

Under the statute, the Commission was required to assess any latent production capacity of PVA in Taiwan, and the possibility of substantially increased imports to the United States, considering the availability of other export markets to absorb any additional exports. 19 U.S.C. § 1677(7)(F). The Commission found that the evidence on the production capacity of Chang Chun PetroChemical Co. Ltd ("CCPC"), the only firm known to have produced PVA products in Taiwan during the period of investigation, did not indicate a threat.³² Def.'s Brief at 67.

32 The Commission noted that CCPC had [_____] production capacity, and [_____]. Def.'s Brief at 67.

Celanese challenges the Commission's finding on subject country production capacity, claiming that it is based on contradictory evidence that fails to meet the clear and convincing standard, and asserts that the issue requires further investigation. Pl.'s Brief at 61-64. Specifically, Celanese charges that the Commission should have rejected CCPC's data as not credible.³³ Pl.'s Brief at 62. But Celanese's argument is without merit. As discussed above, the Commission has great discretion to weigh the probative value of contradictory evidence. Conn. Steel Corp., 18 CIT at 315, 852 F. Supp. at 1064; see also Am. Lamb Co., 785 F.2d at 1002-04. There is no indication that the Commission was arbitrary in its analysis here.

33 Celanese emphasizes CCPC's last minute revisions. Pl.'s Brief at 62.

Celanese's argument as to the alleged contradictory nature of CCPC's evidence regarding [_____] likewise fails. Celanese offers nothing more than its own speculation that CCPC will not be able to do as it plans.³⁴ And a threat determination cannot be based on a "mere conjecture or supposition." 19 U.S.C. § 1677(7)(F)(ii). Moreover, even firm evidence of increased capacity does not necessarily imply increased exports to the United States. Am. Spring Wire Corp. v. United States, 8 CIT 20, 28, 590 F. Supp. 1273, 1281 (1984).

34 CCPC indicated that it [_____]. See Pl.'s Brief at 64 ( citing Staff Report at VII-2).

2. 3. Volume and Price Effects

In its volume and price effects findings regarding threat, the Commission specifically referred to its findings on the conditions of competition as muting the impact of earlier increases in volume in determining no evidence of substantially increased imports in the near future.³⁵ Preliminary Determination at 31-32; Conf. Comm. Views at 52. The Commission further referred to the above findings and its findings on price effects, supra, to support its finding of no evidence of suppressing or depressing price effects. Additionally, the Commission found no evidence that the conditions of competition in the U.S. market would change in such a way as to make any increase more significant in the future. Because each of these findings may be subject to change on remand, judicial review of the Commission's volume and price effects findings would be inappropriate at this time.

35 While the Commission found that the volume of subject imports had increased, it weighed this against the conditions of competition and the lack of price effect. It also noted that subject import volumes from Taiwan did not increase significantly during the period of investigation or during the most recent period when they had opportunities to do so, such as shortly after the termination of the antidumping duty order on PVA imports from Taiwan in 2001 or after the imposition of the antidumping duty orders on PVA imports from China, Korea, and Japan in July 2003 (Japan) and October 2003 (China and Korea). Preliminary Determination at 31 n. 208; Conf. Comm. Views at 51 n. 208.

4. Inventory

The Commission found that end-of-period inventories of subject PVA in Taiwan declined throughout the period of investigation and were projected to continue declining in 2004 and 2005. Preliminary Determination at 31; Conf. Comm. Views at 51. In addition, the Commission found that importers' end-of-period inventories of subject PVA in the United States were relatively stable throughout the period of investigation and were projected to remain stable in 2004 and 2005. Preliminary Determination at 32; Conf. Comm. Views at 52-53.

Celanese challenges the Commission's inventory analysis as based on data from importers that Celanese contends "lacks credibility on its face." Pl.'s Brief at 67; Pl.'s Reply Brief at 30. Celanese targets the inventory data of one importer,³⁶ disparaging it as a mere estimate rather than confirmed values. However, the Commission regularly instructs companies to provide estimates when unable to supply information "in exactly the form requested." Conn. Steel Corp. v. United States, 30 CIT ___, ___, 462 F. Supp. 2d 1322, 1331 (2006).

36 Celanese singles out [_____].

Celanese similarly disputes the information supplied by a second importer as incomplete and lacking credibility. Pl.'s Brief at 67. Celanese asserts that the importer's claim regarding the limits of its data are unbelievable, and argues that the Commission failed to thoroughly investigate (as Celanese argues it was required to do). Id. According to Celanese, this "lack of data . . ., combined with the likelihood that additional data could be obtained by the time of the final determination, requires the Commission to continue the case." Id.

Celanese is again arguing that the Commission should have arrived at a different conclusion based on determinations as to the credibility of facts that underlie the Commission's analysis. As in a material injury determination, in making its threat determination the Commission "is afforded discretion in interpreting the data, and the court does not weigh the evidence." U.S. Steel Group, 18 CIT at 1224, 873 F. Supp. at 703 (citations omitted). In other words, judicial review "does not extend beyond determining whether the Commission has acted within its delegated authority and has correctly interpreted and applied the law." Bando Chem. Indus., Ltd. v. United States, 17 CIT 798, 802 (1993).

As to Celanese's claim that "contrary evidence" with respect to inventories would likely arise in a final investigation, Celanese does not suggest with any particularity what that evidence might be. Indeed, throughout its briefs, Celanese repeatedly asserts the need for "additional evidence." However, as established byAm. Lamb Co., the standard for a preliminary injury determination requires the Commission to determine whether "(1) the record as a whole contains clear and convincing evidence that there is no material injury or threat of such injury; and (2) no likelihood exists that contrary evidence will arise in a final investigation." Am. Lamb Co., 785 F.2d at 1001 (emphasis added). Moreover, the Commission must make its decision based on the evidence available at the time. Co-Steel Raritan, 357 F.3d at 1313 ("The Commission will investigate the allegations in the petition in as thorough a manner as possible using the information available within that time period . . .") ( citing Am. Lamb Co., 785 F.2d at 1003 ( quoting H.R. Rep. No. 96-317, at 61 (1979))).

5. Development and Production

Finally, the Commission found no reasonable indication that subject imports would have an actual or potential negative effect on the domestic industry's existing development and production efforts.³⁷ Preliminary Determination at 31; Conf. Comm. Views at 53. The statute requires a causal link between a threat of material injury finding and subject imports. 19 U.S.C. § 1673d(b)(1). The record before the Commission reflects a number of favorable conditions pertaining to the domestic industry (in addition to some negative trends). Based on the record as whole, there is no indication that the Commission's determination on this issue was in any way unreasonable or factually unsupportable.

37 The Commission noted that the domestic industry [_____]; and, although Celanese reported actual and anticipated [_____], [_____]. Preliminary Determination at 31-32; Conf. Comm. Views at 53.

V. Conclusion

For the reasons set forth above, Plaintiff's Motion for Judgment Upon the Agency Record is granted in part, and this action is remanded to the Commission for further action in accordance with this opinion.

ORDER

In accordance with the opinion of the Court issued this date in this matter, it is hereby

ORDERED that this matter is remanded to the U.S. International Trade Commission for further proceedings not inconsistent with that opinion; and is further

ORDERED that the Commission shall file the results of this remand with the Court no later than March 30, 2007; and it is further

ORDERED that any comments on those results shall be filed no later than April 30, 2007; and it is further

ORDERED that the Commission's response to any comments shall be filed no later than 21 days after the filing of those comments.