Nippon Steel Corp. v. United States 350 F. Supp. 2d 1186

OPINION

RESTANI, Chief Judge.

Before the court is the United States International Trade Commission’s (“Commission” or “ITC”) second remand determination concerning tin- and chromium-coated steel sheet (“TCCSS”) imports from Japan. In its original determination, the Commission concluded that the United States TCCSS industry was materially injured by reason of TCCSS imports from Japan (“subject imports”) that were sold at less than fair value (“LTFV”). Tin- and Chromium-Coated Steel Sheet From Japan, 65 Fed.Reg. 50,005, USITC Pub. 3300, Inv. No. 731-TA-860 (final determ.) (Aug.2000) (A.R.2-148) [hereinafter Final Determination ]. Although the court *1188found the Commission’s conclusions with respect to subject .import volume supported by substantial evidence, the court ordered the Commission to reevaluate its analysis of the effect of subject imports on domestic pricing as well as its conclusions with respect to causation. Nippon Steel Corp. v. United States, 182 F.Supp.2d 1830, 1356 (CIT 2001) (“Nippon I”).

On remand, the Commission again determined that the domestic industry was materially injured by reason of subject imports. Tin- and Chromium-Coated Steel Sheet from Japan, Inv. No. 731-TA-860 (final determ.) (March 2002) (A.R.2261R) [hereinafter First Remand Determination ]. After reviewing. the Commission’s explanations and the evidence, the court found otherwise. Nippon Steel Corp. v. United States, 223 F.Supp.2d 1349 (CIT 2002) (“Nippon II”). The court held that uncontested evidence established that LTFV subject imports did not have a material effect on domestic prices and that there was no valid reason to discount non-price factors or non-subject imports as the predominant cause of material injury. Id. Furthermore, the court found that a remand for reconsideration or recalculation was not necessary because the Commission had “demonstrated an unwillingness or inability to address the substantial claims made by the respondents or the concerns expressed by the court in Nippon I ...” Id. at 1371-72. Instead, the court vacated the affirmative injury finding and directed the Commission to enter a negative material injury determination. Id. at 1372.

The Commission appealed the court’s decision in Nippon II. The Court of Appeals for the Federal Circuit held that this court abused its discretion by not returning the case to the Commission for further analysis.' Nippon Steel Corp. v. International Trade Com’n, 345 F.3d 1379, 1381 (Fed.Cir.2003) (“Nippon III”). The Federal Circuit explained that “to the extent that the Court of International Trade engaged in refinding the. facts (e.g., by determining witness credibility), or interposing its own determinations on causation and material injury itself, [it] ... exceeded its authority.” Id. The Federal Circuit vacated the court’s decision in Nippon II and remanded the case to the Commission to “attend to all the points made by the Court of. International Trade, especially those of [Nippon II] which the Commission [had] not yet had the opportunity to address.” Id. at 1382.

Therefore, the Commission considered the case on a second remand. In its second remand determination, the Commission determines that the domestic industry was materially injured by reason of Japanese imports. Tin- and Chromium-Coated Steel Sheet from Japan, Inv. No. 731-TA-860 (Feb.2004) (A.R.2-263R) [hereinafter Second Remand Determination]. 1 Nippon Steel Corporation, NKK Corporation, Kawasaki Steel Corporation, and Toyo Kohan Co., Ltd., (collectively “Nippon” or “Plaintiffs”), challenge this *1189determination on the grounds that the Commission’s findings of price effects and causation remain unsupported by substantial evidence. For the reasons set forth below, the Commission’s affirmative material injury determination is remanded with instructions to issue a negative material injury determination.

JURISDICTION AND STANDARD OF REVIEW

The court has jurisdiction pursuant to 28 U.S.C. § 1581(c) (2000). The court will uphold the Commission’s final determination in an antidumping investigation unless it is “unsupported by substantial evidence on the record, or is otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i) (2000).

OVERVIEW

TCCSS is a tin-coated flat-rolled steel product, primarily used in the production of containers for the food processing industry.2 The domestic TCCSS industry is characterized by unique conditions of competition. For example, there are a relatively small number of buyers and sellers: Seven domestic suppliers (“suppliers” or “producers”), two dozen importers, and six major U.S. purchasers.3 In addition, the majority of U.S. suppliers are located in the Eastern and Midwestern United States and typically supply facilities in those areas.4 Although Japanese suppliers compete more heavily in the West, they supply purchasers throughout the United States. Id. at II — 7, Table II — 1. Non-subject imports, on the other hand, compete only in the East and Midwest, and during the period of investigation entered the U.S. market in larger volumes than Japanese imports. Id. at II — 7, IV-5. Id. Nonetheless, domestic mills account for the majority of U.S. consumption.5 Furthermore, TCCSS is almost always sold pursuant to annual contracts that establish fixed prices and target volumes. Prior to entering into a contract, however, the majority of purchasers require suppliers to demonstrate an ability to make reliable deliveries, supply high-quality product and specialty items, and provide quality service. Id. at 11-10. These non-price considerations are important factors to TCCSS purchasers. Id.

DISCUSSION

In the final phase of an antidumping duty investigation, the Commission determines whether an industry in the United States is materially injured by reason of the imports under investigation. 19 *1190U.S.C. § 1673d(b). Material injury is defined as “harm which is not inconsequential, immaterial, or unimportant.” 19 U.S.C. § 1677(7)(A). An affirmative material injury determination requires the Commission to find that the volume, price effects, and impact of the subject imports are significant, and that the material injury was by reason of the subject imports. Id. § 1677(7)(B); see also Gerald Metals, Inc. v. United States, 22 CIT 1009, 1012-13, 27 F.Supp.2d 1351, 1354-55 (1998). In this case, the court previously upheld the Commission’s determination of a small but significant volume, and Plaintiffs did not challenge the Commission’s findings on impact. Nippon I, 182 F.Supp.2d at 1340. Thus, the court now reviews the Commission’s findings regarding (I) price effects, and (II) causation.

I. Price Effects

In evaluating the effect of subject imports on domestic prices, the Commission must consider whether there has been “significant price underselling by the imported merchandise,” and whether the effect of such imports “otherwise depresses prices to a significant degree or prevents price increases, which otherwise would have occurred, to a significant degree.” 19 U.S.C. § 1677(7)(C)(ii). In addition, the Commission must evaluate price effects within the context of the “conditions of competition that are distinctive to the affected industry.” Id. § 1677(7)(C)(iii). In Nippon II, the court held that the Commission’s finding of significant price effects was unsupported by substantial evidence. 223 F.Supp.2d at 1351-52. Specifically, the court found that the Commission (A) failed to explain its selection and compilation of price comparison data, (B) did not support its finding of significant underselling, (C) ignored evidence contradicting a finding of domestic price depression, and (D) did not consider the industry’s unique conditions of competition. Id. The Commission addresses these concerns in its Second Remand Determination, however, the record as a whole continues to demonstrate that subject imports did not have a significant effect on domestic prices.

A. Selection and Compilation of Price Comparison Data

The Commission has “broad discretion in analyzing and assessing the significance of the evidence on price undercutting.” Copperweld Corp. v. United States, 12 CIT 148, 161, 682 F.Supp. 552, 565 (1988). It must, however, “provide a reasonable explanation as to why it chose the evidence used to support its findings.” Bratsk Aluminum Smelter v. United States, No. 03-00200, 2004 WL 1385848, *10 (CIT 2004). In Nippon II, the court held that the Commission failed to provide a reasonable explanation as to why it (1) separated out Purchaser A’s different facilities and product types, (2) considered only bids that ultimately culminated in final sales, and (3) analyzed underselling data for only 1999.

1. Considering Purchaser A’s Data Separately

The court previously held that the Commission failed to explain its decision to keep Purchaser A’s separate facilities 'and product types in disaggregated form. Nippon II, 223 F.Supp.2d at 1355. As a result, the Commission now calculates a single aggregate unit price for Purchaser A’s facilities and products and then includes that data in its price comparison tables.6 Plaintiffs challenge the Commission’s aggregation of this data with respect *1191to the bid and volume comparison table, Table Second Remand 3 (“Table 3”).7 Plaintiffs point out that after the Commission aggregated Purchaser A’s data, the number of subject underselling bids decreased but the volumes associated with those bids increased. PL’s Resp. at 4. Therefore, Plaintiffs argue that the Commission must have incorrectly calculated or erroneously listed volumes associated with these underselling bids. Id. The Commission responds that it accurately prepared Table 3 by aggregating the purchaser’s data, and listing the average annual Japanese bids, and the volumes associated with those bids, in the appropriate column. ITC Reply at 5.

The Commission fails to address why aggregation of Purchaser A’s data altered Table 3’s total volume calculation.8 Although the Commission notes that the actual number of bids decreased due to aggregation, it fails to explain why the volumes associated with low bids increased, while those with high bids decreased. Indeed, the Commission’s assertion that “concerns over over-representation arise only with respect to the number of bids, and not with respect to the volume of sales won by the Japanese suppliers,” seems to suggest that aggregation would not alter total volume calculations. Second Remand Determ, at 24 n.97. Because the Commission has failed to explain this volume discrepancy in Table 3, it is unreasonable for it to rely on this data in its underselling analysis. See infra n. 13.

2. Considering Only Bids Resulting in Final Sales

In addition, the court previously held that the Commission’s consideration of only those bids that ultimately resulted in final sales might give skewed results. Nippon II, 223 F.Supp.2d at 1355. The Commission now explains that consideration of only final bids will not lead to skewed results because only three purchasers submitted initial bid data, and even those were superseded by final bids.9 In addition, the Commission explains that it is long-standing Commission practice to rely on weighted averaging data to assess underselling margins, and weighted averaging — which gives greater prominence to higher volume sales, and thus requires *1192that prices have some associated volume— cannot evaluate initial bids that were not awarded any volume.

Because the Commission provides reasonable explanations for its consideration of only final bids and its reliance on weighted average pricing data to assess underselling margins, the court upholds the Commission’s use of these methodologies.10 See Nucor Corp. v. United States, 318 F.Supp.2d 1207, 1257 (CIT 2004) (quoting Mitsubishi Elec. Corp. v. United States, 12 CIT 1025,1050, 700 F.Supp. 538, 558 (1988)) (holding that it is within the Commission’s “discretion to select a particular methodology and as long as substantial evidence supports that choice, the [c]ourt reviewing such methodology will sustain the [agency’s] decision”).

3. Analyzing Underselling Data for Only 1999

The court also held that the Commission had dispensed with the use of a trend analysis of underselling data, and was instead relying only on evidence of underselling in 1999. Nippon II, 223 F.Supp.2d at 1355. The court found that this focus on underselling data for only 1999 was inconsistent with the Commission’s analysis of volume trends and domestic pricing patterns for the entire period of investigation. See id. (“the Commission may not rely, as it has done in this case, on trends in subject import market share and domestic pricing to substantiate the significance of its one-year data on underselling”). As a result, the Commission now relies on weighted average pricing and bid comparison data in its Second Remand Determination to support its finding of a trend of significant underselling. Therefore, contrary to Plaintiffs’ assertion, the Commission addresses the court’s concern by assessing underselling trends over the period of investigation. See discussion infra at Section I.B.I.

B. Significance of Underselling

The statute requires the Commission to determine whether “there has been significant price underselling by the imported merchandise as compared with the price of domestic like products ...” 19 U.S.C. § 1677(7)(C)(ii)(I). In determining the significance of underselling, “[i]t is within the Commission’s discretion to make reasonable interpretations of the evidence ...” Maine Potato v. United States, 9 CIT 293, 300, 613 F.Supp. 1237, 1244 (1985). It must, however, “articulate a ‘rational connection between the facts found and the choice made.’ ” Bowman Transp. Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) (quoting Burlington Truck Lines v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962)). In this case, the Commission fails to articulate a rational connection between the record evidence and its findings with regards to underselling. In particular, its findings that (1) there was a trend of increased underselling over the period of investigation, (2) underselling was significant in 1999, and (3) underselling margins were not attributable to the domestic industry’s lead time advantage, are not supported by substantial evidence.

1. The ITC’s Finding of a Trend of Increased Underselling is not Supported by Substantial Evidence

As discussed above, the Commission relies on weighted average pricing and bid *1193comparison data to support its finding that there was a trend of significant underselling over the period of investigation. Neither of these data, however, supports the Commission’s conclusion.

First, the Commission contends that the weighted average underselling data in Table Second Remand 1 shows increasing levels of subject underselling over the period of investigation. In fact, only two purchasers listed in that table experienced widening underselling margins over the period of investigation.11 Two other purchasers experienced decreasing margins, two had mixed patterns of underselling and overselling, and two only reported data for 1999.12 Two purchasers’ data out of eight is not substantial evidence of a trend of increased underselling over the period of investigation.

Second, the Commission asserts that Table 3 (a bid and volume comparison) shows a pattern of increased underbidding over the period of investigation. Table 3 reveals that the number of Japanese bids “below all U.S. bids” increased from 1997 to 1999, but decreased from 1999 to 2000. Second Remand Determ, at 24. The Table also shows that in 1997 and 1998 there were more Japanese bids “within the range of all U.S. bids” than “below all U.S. bids.”13 Therefore, Table 3 does not support the Commission’s finding of a trend of increased underbidding over the period of investigation.

2. The ITC’s Finding of Significant Underselling in 1999 is not Supported by Substantial Evidence

In its First Remand Determination, the Commission focused its underselling analysis on the margins of several large purchasers in 1999, the only year in which it found generally lower Japanese prices. First Remand Determ, at 12-13. Specifically, the Commission evaluated responses to Question IV-8, in which purchasers estimated the price differential that would induce them to switch suppliers, and found that 1999 underselling margins were “generally near or at the ranges found by responding purchasers to be significant....” Id.

The court held that the Commission’s finding was unsupported by substantial evidence because the wording of Question IV-8 was not relevant to underselling, and the Commission ignored explanatory responses and individual purchasing histories in assessing purchasers’ estimates of price differentials. Nippon II, 223 F.Supp.2d at 1351. In its Second Remand Determination, although the Commission provides a reasonable explanation for its continuing reliance on Question IV-8 in assessing determinative price differentials, its eonclu*1194sion that underselling margins in 1999 were significant remains unsupported by substantial evidence.

a. The ITC’s Continuing Reliance on Question IV-8 for some Purposes is Reasonable

Question IV-8 asked purchasers to estimate “during 1999, approximately how much higher would the price for the imported product have to have been before [they] would have purchased U.S.-produced [TCCSS] instead.” Purchaser Questionnaires at IV-8. The court noted that this question was not phrased to elicit responses relevant to underselling, and proposed that a better question would ask purchasers “how much lower [subject] prices would have to be before [they] would switch from a domestic to a Japanese producer.” Nippon II, 223 F.Supp.2d at 1357 (emphasis in original).

In response to the court’s concern, the Commission reopened the record on remand and asked purchasers a rephrased question. Notice and Scheduling of Remand Proceedings, 69 Fed.Reg. 2361 (Jan. 15, 2004). The Commission asked purchasers “whether [they were] willing to pay more for TCCSS from one source, versus the other during the [period of investigation];” and “if yes, how much more ... and for what reason or reasons .... ” Memo from De Filippo to the Commission (Jan. 23, 2004), (A.R.2-262R), at 3, ITC App., Tab 17, at 3. This questioning, however, failed to generate any substantive responses.14

Although Plaintiffs argue that this rephrased question, like the original, does not address how much lower subject merchandise would have to be to induce purchasers to switch from domestic suppliers, it is unlikely that any question would have generated substantive responses, considering that the relevant information is four years old. As a result, it is not unreasonable for the Commission to continue to rely on responses to Question IV-8 to assess purchaser estimates of price differentials in 1999. See 19 U.S.C. § 1677e(a)(l) (authorizing the Commission to rely on facts otherwise available when necessary information is missing from the record). The Commission also .points out that the court in Acciai Speciali Terni, S.p.A. v. United States, 19 CIT 1051, 1059 (1995), found that purchaser responses to an almost identical question were relevant to assessing the importance of price differentials between subject and domestic merchandise on purchase decisions in the market. Because the Commission was unable to obtain more relevant information, and because the Commission’s reliance on a similar question was previously upheld by this court, the Commission’s continued reliance on Question IV-8 for some purposes is reasonable.

b. Responses to Question IV-8 do not show that Underselling Margins in 1999 were Significant

Purchaser responses to Question IV-8 varied. Six purchasers provided a numerical estimate of the price differential at which they would switch suppliers, one gave an estimate of zero, and three gave explanatory responses indicating that price was not their most important purchasing consideration. Based on the numerical responses, the Commission derived an overall range — two to six percent — of a price differential that would induce a switch of *1195suppliers for 1999, and concluded that underselling margins were generally within this range and therefore significant. First Remand Determ, at 12-13. The court in Nippon II, however, criticized this finding because the Commission ignored explanatory responses giving context to purchaser estimates, and failed to determine the extent to which purchasers’ numerical responses were actually borne out by individual purchasing history. Nippon II, 223 F.Supp.2d at 1351.

i. Explanatory Responses

Purchasers B, D, and A provided explanatory responses to Question TV-8 suggesting that they did not switch suppliers in 1999 based on price. See Purchaser B Questionnaire Resp.15 (stating that it “[chose] steel suppliers based on 1) quality, 2) service, and 3) price, in that order of importance ... unique manufacturing capabilities of some off-shore sources drive us to purchase from them irrespective of their prices which, in most cases, are higher than U.S. producer prices”); Purchaser D Questionnaire Resp. 16 (stating that it “did not select the Japanese [suppliers] based on price, but on quality performance, and the[ ] ability to demonstrate this quality to domestic suppliers, thereby proving that [Purchaser D’s] expectations for quality could be achieved”); Purchaser A Questionnaire Resp. 17 (stating that “pricing is not the main factor. [Purchaser A] is developing long-term global partners that have lowest cost and highest performance. This decision can be more geographical driven than price driven.”). Nonetheless, the Commission concludes that these explanatory responses do not undermine its finding of significant underselling in 1999 because other record evidence shows that these purchasers considered price to be an important factor in their purchasing decisions.18 Even without consideration of these explanatory responses, however, the Commission fails to point to substantial evidence that Purchaser B, D, or A considered the margin at which they bought subject imports in 1999 to be significant.

First, the Commission contends that Purchaser B’s purchasing history shows a correlation between underselling margins and purchasing decisions. Second Remand Determ, at 39. The Commission notes that in 1999, Purchaser B increased its purchases of Japanese imports at the same time prices decreased. A simple correlation between underselling and purchasing, however, is insufficient to establish that underselling margins were significant. See Nippon I, 182 F.Supp.2d at 1342 (“a rapid rate of increase nonetheless may be immaterial if, for example, the margin never goes above a price differential that would cause purchasers to change suppliers to any significant degree”) (emphasis added). In fact, notwithstanding lower Japanese prices in 1999, Purchaser B continued to obtain the vast majority of its TCCSS from U.S. suppliers.19 Therefore, *11961999 underselling margins were not to such an extent to induce Purchaser B to change suppliers to any significant degree. This evidence shows that Purchaser B did not consider 1999 margins to be significant.

Second, the Commission asserts that Purchaser D’s purchasing patterns show that price was an important part of its purchasing decisions because it purchased subject merchandise in 1999 at a price that was significantly lower than the lowest domestic bid. Sécond Remand Determ, at 32. Plaintiffs argue that Purchaser D’s purchasing history shows an increase in purchases from domestic suppliers over the period of investigation. PI. ’s Resp. at 8. The issue is whether Japanese underselling margins in 1999 were significant. In 1999, Purchaser D paid approximately $70 less for Japanese imports than for U.S. merchandise.20 Notwithstanding this price difference, however, Purchaser D obtained nearly ninety-nine percent of its TCCSS from U.S. suppliers that year. See id. at Revised Table TCCSS-1. . Similar to Purchaser B’s data, this evidence indicates that Japanese underselling was not significant enough to induce Purchaser D to substantially switch suppliers in 1999. It shows that Purchaser D did not consider 1999 underselling margins to be significant.

Third, the Commission asserts that Purchaser A considered 1999 underselling margins to be significant because it bought more TCCSS from domestic suppliers that year in response to narrowing underselling margins. Plaintiffs argue that the purchasing history contradicts the Commission’s underselling analysis. From 1998 to 1999, although Japanese prices fell and U.S. prices rose, Purchaser A bought less Japanese merchandise and more U.S. product.21 Contrary to the Commission assertion, this purchasing data indicates that Purchaser A shifted volume in 1999 for reasons other than price. It does not support a finding of significant underselling that year.

Therefore, notwithstanding the explanatory responses, purchasing histories show that Purchasers B, D, and A did not consider the price at which Japanese imports undersold domestic TCCSS in 1999 to be significant.

ii. Numerical Estimates

Six purchasers responded to Question IV-8 with numerical price differentials.22 Of these six, the underselling margins for two purchasers fell within their estimated price differentials, indicating that they considered the margin at which they purchased Japanese imports in 1999 to be *1197significant.23 Two supportive responses out of nine is not substantial evidence showing that 1999 underselling margins were significant. Rather, it indicates the opposite.

3. The ITC Fails to Substantially Support its Finding that Premiums Paid for Superior Domestic Lead Times were Minimized by other Factors

Evidence of underselling has been found to be less significant where there are price premiums for superior domestic lead times. See Committee for Fair Beam Imports v. United States, No. 02-00531, 2003 WL 21555105, *14 (CIT 2003) (noting that “the ITC may discount incidences of underselling on account of [a] price premium, where appropriate, as this price premium mitigates underselling that is observed”); Timken Co. v. United States, 20 CIT 76, 87-88, 913 F.Supp. 580, 589-90 (1996) (discussing how ITC “concluded that factoring in the price premium for domestic [products] made the relatively small margins of underselling even less significant”); Trent Tube Div., Crucible Materials Corp. v. United States, 14 CIT 386, 402, 741 F.Supp. 921, 935 (1990) (according less weight to consistent underselling based on domestic price premium due to customer preferences and lead time differences); Roses, Inc. v. United States, 13 CIT 662, 665-66, 720 F.Supp. 180, 183 (1989) (finding 62 out of 110 instances of underselling insignificant because price premiums for locally-grown roses based on freshness and an ability to supply the flowers on a short-term need basis).

In Nippon I, the court held that the Commission failed to analyze whether the undisputed lead-time advantage held by the domestic industry translated into price premiums over imports, and accounted for the margin of overselling. 182 F.Supp.2d at 1342. In its First Remand Determination, the Commission responded that although the domestic lead time advantage “can give purchasers more flexibility in modifying their purchase orders on shorter notice, which can translate into a price premium,” it was diminished by other factors, and thus not responsible for underselling margins. First Remand Determ. at 13. The court held, however, that even if the price premium were somewhat diminished, it “may still eclipse the underselling margin.” Nippon II, 223 F.Supp.2d at 1360. Without assessing the extent of the price premium in relation to underselling margins, the court held that the Commission’s findings were unsupported by substantial evidence. Id.

In response to the court’s concerns, the Commission reopened the record on its second remand to obtain additional information regarding the existence and size of any such premium.24 As discussed above, however, it was unsuccessful in obtaining any substantive responses. See discussion supra Section I.B.2.a. As a result, the Commission continues to rely on the original record to support its finding that a premium, if it did exist, was offset by premiums paid for superior Japanese quality or other factors.25 Second Remand *1198 Determ. at 50-56. Plaintiffs argue that the Commission again fails to assess the price premium in relation to underselling margins. PI. ’s Resp. at 14.

Without evidence of the extent or size of price premiums, the Commission has no support for its belief that premiums paid for faster domestic lead times were minimized by premiums paid for superior Japanese quality or other factors, and thus.did not translate into underselling margins.26 An affirmative determination cannot be supported by speculation. The Commission may rely on the less than perfect available information, but it must be actual information.

C. Domestic Price Depression or Suppression

An affirmative injury determination also requires a finding that subject imports “depress[ ] prices to a significant degree or prevent[ ] price increases, which otherwise would have occurred, to, a significant degree.” 19 U.S.C. § 1677(7)(C)(ii)(II). The Commission here bases its finding of domestic price depression and suppression on the following evidence: (1) The domestic industry’s cost-price squeeze, (2) a lost sale allegation, and (3) purchaser pricing data. This evidence, however, does not substantially support, the Commission’s finding of significant price depression or suppression.

1. The ITC’s Cost-Price Squeeze Analysis does not Substantially Support a Finding of Price Depression or Suppression

An industry is in a cost-price squeeze situation if it is unable to raise prices to cover the cost of goods sold. Here, the Commission finds that the domestic industry experienced a cost-price squeeze over the period of investigation because the industry’s overall cost of goods sold increased in relation to its net sales. Second Remand Detem. at 26-27. Plaintiffs challenge this finding on the grounds that the two domestic mills that compete most directly with Japanese imports reported positive operating margins during the period of investigation. PI. ’s Resp. at 6.

Evidence of a cost-price squeeze generally supports a finding of domestic price suppression. See Gerald Metals, Inc., 22 CIT at 1023, 27 F.Supp.2d at 1362 (discussing how an ITC Commissioner could not find “clear evidence of a cost-price squeeze, [and thus] conclude[d] that price suppression could not be attributed to the LTFV imports to a significant degree”). In this case, however, because of the particular conditions of competition, the court previously instructed the Commission to assess individual purchaser data to determine whether a correlation between subject imports and domestic pricing existed. Nippon I, 182 F.Supp.2d at 1344. Therefore, although- the record shows that the domestic industry generally may have been experiencing a cost-price squeeze,27 *1199the Commission’s finding of a correlation between subject import pricing and U.S. pricing must be supported by individual pricing, data. Id.; see also Altx, Inc. v. United States, No. 00-08-00477, 2002 WL 1560884, *13 (CITrade July 12, 2002) (“It is true that the Commission need not find an exact correlation between subject import pricing and domestic pricing. Nevertheless, the Commission’s cursory look at general net declines over the POI is insufficient.”).

2. The ITC’s Reliance on a Lost Sale Allegation is Improper

The Commission also finds that evidence of a lost sale indicates that the domestic industry’s prices were suppressed. A lost sales allegation refers to the situation in which the domestic industry is unable to make a sale because of the presence of lower priced imports. Copperweld Corp., 12 CIT at 169 n. 15, 682 F.Supp. at 572 n. 15. Although evidence of a lost sale may be probative of price suppression, the lack of such evidence ordinarily will not vitiate a Commission’s determination. USX Corp. v. United States, 11 CIT 82, 86, 655 F.Supp. 487, 491 (1987). Where the Commission chooses to rely on findings of lost sales, however, it must support such findings with substantial evidence. Nippon II, 223 F.Supp.2d at 1364-65.

In Nippon II, the court held that the Commission improperly relied on the lost sale allegation because it was unsupported by the record and undermined by evidence that the sale was lost for non-price reasons. Id. at 1352. In its Second Remand Determination, the Commission again cites this lost sale allegation as evidence of domestic price suppression. . Second Remand Determ, at 81. Plaintiffs argue that the lost sale is unsupported by substantial evidence. PL’s Resp. at 23. Because the record (a) does not correspond with the volume and pricing data of this alleged lost sale, and (b) indicates that the lost sale, if it did occur, was due to non-price reasons, the Commission’s reliance on this lost sale is improper.

a. Purchaser A’s Data does not Support the Alleged Lost Sale

The lost sale at issue involves an allegation by a particular domestic producer, claiming that it lost a sale in October 1998 to Purchaser A because of a lower Japanese bid.28 On its second remand, the Commission finds that the lost sales allegation is confirmed by Purchaser A’s 2000 record data. Second Remand Determ, at 81. Plaintiffs correctly point out, however, that the lost sales allegation is not reflected in Purchaser A’s volume or pricing data.29

First, Purchaser A’s volume data does not correspond to that of the alleged lost sale. In 2000, Purchaser A bought approximately 6,000 short tons more of Japanese TCCSS than was alleged lost by the *1200U.S. purchaser at the relevant facility.30 Contrary to the Commission’s assertion, these volumes are not “fully consistent.” Second, Purchaser A’s pricing data does not correspond to that of the lost sale. The Commission concedes that the U.S. producer’s bid at the relevant facility was “well above” the alleged price per ton.31 Second Remand Determ, at 83. As a result, it alternatively points to a bid that the U.S. producer made at a different facility, which is closer in price to the one allegedly lost. Id. The Commission fails to note that rather than losing that bid, the U.S. producer was awarded that sale.32 The record simply does not support the volume and price values of the alleged lost sale.

b. Purchaser A’s Comments Undermine the Commission’s Finding that the Lost Sale was due to Lower Japanese Prices

Apparently because the record does not support the alleged lost sale, the Commission emphasizes the fact that Purchaser A agreed with the allegation. Although Purchaser A agreed with the lost sale allegation, the court in Nippon II held that other statements by Purchaser A indicated that the sale was lost because of non-price reasons. 223 F.Supp.2d at 1365-66. Specifically, the court noted that Purchaser A(l) emphasized non-price factors when agreeing with the lost sale allegation, (2) testified that it had long-term commitments with its Japanese supplier at the relevant facility, and (3) noted that the particular U.S. producer did not bid seriously for its West Coast business. Id. Although the Commission asserts otherwise, these statements continue to show that if the sale was lost, it was because of non-price factors.

First, when Purchaser A agreed with the lost sale allegation, it emphasized that “price is not the only factor in contract selection; quality and delivery time are also critical, and whether the firm can supply globally.” Staff Report at V-25. The Commission interprets this statement to mean that Purchaser A considered price to be a significant factor in its purchasing decisions, and as a result, the sale was lost to lower priced Japanese imports. Second Remand Determ, at 86. The Commission’s interpretation of this statement stretches it beyond recognition. Rather it appears that Purchaser A bought Japanese imports instead of this particular U.S. purchaser’s merchandise because of quality, delivery, and global considerations.

Second, Purchaser A noted that it had long-term commitments with its Japanese supplier.33 The Commission argues that *1201because Purchaser A also indicated that the Japanese suppliers were willing to offer more attractive prices, the sale was lost because of lower-priced Japanese imports. On the contrary, however, this statement clearly indicates that Purchaser A awarded Japanese suppliers with tonnage because of deepening commitments and plans to develop new products.

Third, Purchaser A stated that the particular U.S. supplier did not bid seriously for its West Coast business, suggesting that the sale would have been lost to another domestic supplier in the absence of Japanese competition.34 The Commission argues that purchasing history shows that the U.S. supplier bid consistently and competitively on Purchaser A’s business at that facility throughout the period. Purchaser A’s purchasing history reveals, however, that the U.S. supplier’s bids were significantly higher than other U.S. bids, which were ultimately accepted.35 Purchaser A’s statement as well as its purchasing history do not support the Commission’s conclusion that the lost sale was due to lower Japanese prices.

Therefore, Purchaser A’s statements undermine the Commission’s finding that the alleged lost sale was due to lower Japanese prices.

3. Purchaser Pricing Data does not Substantially Support a Finding of Domestic Price Depression or Suppression

In its Final Determination, the Commission finked a general decline in domestic prices to overall underselling trends in Japanese pricing. Final Determ, at 15. As discussed above, the court held that although the Commission is not required in all cases to determine the relationship between subject imports and domestic prices on an individual purchaser basis, “where the other data is mixed and where data is available to determine whether such a correlation existed for particular purchasers, and is relied on by respondents, the Commission must address the individual purchaser data in some manner.” Nippon I, 182 F.Supp.2d at 1344. Nonetheless, in its First Remand Determination, the Commission failed' to do so. Nippon II, 223 F:Supp.2d at 1358-59. In its Second Remand Determination, the Commission evaluates the pricing data of four of the six large purchasers and finds a correlation between Japanese imports and the price at which they bought domestic TCCSS. Second Remand Determ, at 42-50. Plaintiffs argue that purchasers’ pricing data reveal no such correlation. Pl.’s Resp. at 13.

The four purchasers whose pricing data was evaluated by the Commission — Pur*1202chasers A, C, D, and B — did not buy Japanese imports at significant margins of underselling.36 See discussion supra Section I.B. These four purchasers apparently operated to some degree on the West Coast.37 As discussed below, the record as a whole indicates that Japanese imports did not have a significant effect on domestic prices over the period of investigation.

a. Purchaser A

The court in Nippon II held that the Commission failed to explain its finding of significant price depression in light of the fact that Purchaser A, which accounted for the bulk of the instances of underselling, paid increasing domestic prices between 1997 and 1999. 223 F.Supp.2d at 1358. In response, the Commission evaluates the pricing data and finds that Purchaser A did not pay increasing domestic prices throughout the period, and that Purchaser A’s volume levels correlated with Japanese underselling margins over the period of investigation. Second Remand Detenn. at 44. Plaintiffs argue that the pricing data shows that Japanese prices did not affect domestic prices. PI. ’s Resp. at 12.

Purchaser A’s pricing data indicates that U.S. suppliers were able to raise prices notwithstanding lower Japanese prices. Specifically, Purchaser A paid more for domestic TCCSS and less for Japanese TCCSS over the period of investigation.38 Although the Commission argues that Purchaser A bought higher volumes of Japanese imports in response to lower prices, the issue here is price effects. The Commission’s evaluation of volume shifts is not relevant in this context. See Nippon II, 223 F.Supp.2d at 1358 n. 12 (“[t]he court ... sustained the Commission’s finding of low but significant subject import volume as an isolated finding and does not revisit the issue, except as it affects the ultimate causation conclusion.”). Therefore, Purchaser A’s pricing data shows a lack of correlation between subject imports and domestic prices.

b. Purchaser C

In Nippon II, the court also held that the Commission had not adequately ex*1203plained its finding of a correlation between Japanese imports and domestic prices, considering that Purchaser C was able to secure domestic price decreases between 1997 and 1998 without making any subject import purchases until 1999. 223 F.Supp.2d at 1358-59. The Commission now explains that the link between Japanese imports and Purchaser C’s ability to secure domestic price decreases was Purchaser F. Second Remand Determ, at 46-47. Because Purchaser F bought Japanese imports in 1998 and was in a purchasing alliance with Purchaser C, the Commission reasons that Purchaser F probably shared Japanese pricing information with Purchaser C.39 Based on this information, the Commission finds a correlation between Japanese imports and the price at which Purchaser C bought U.S. TCCSS. Plaintiffs argue that this “possible indirect effect” does not establish a correlation between Japanese imports and U.S. prices.40 PI. ’s Resp. at 13.

In the absence of concrete record evidence, the Commission’s “belief’ that Purchaser F “likely” informed Purchaser C of Japanese pricing does not constitute substantial evidence. Moreover, Purchaser F did not buy Japanese imports until 1998, and reported that it was not until “late ’98” when it decided to take advantage of pricé variances it observed between foreign and domestic mills. Purchaser F Questionnaire Resp. at II — 2. Therefore, contrary to the Commission’s conclusion, the record fails to demonstrate a connection between Japanese imports and the 1997 and 1998 prices Purchaser C paid for domestic TCCSS.

c. Purchaser D

In Nippon II, the court also held that the Commission failed to address why Purchaser D bought more expensive domestic TCCSS between 1999 and 2000, despite lower Japanese prices. 223 F.Supp.2d at 1359. The Commission responds that this price increase was related to the filing of the antidumping petition. Second Remand Determ, at 49. Plaintiffs do not challenge this explanation.

Under 19 U.S.C. § 1677(7)(I), the Commission may consider whether any change in price effects since the filing of the petition is related to the pendency of the investigation, and if so, it may reduce the weight accorded to that data. In this case, the Commission points to evidence indicating that the filing of the antidump-ing petition in October 1999 had an effect on domestic pricing in 2000.41 Thus, the Commission’s decision to place less weight on data for 2000 is reasonable.

*1204From 1998 to 1999, the price at which Purchaser D bought domestic merchandise remained stable. The Commission contends that because of the domestic industry’s cost-price squeeze, this stable pricing indicates that Japanese underselling prevented domestic price increases that would have otherwise occurred. Second Remand Detenu, at 48. Plaintiffs argue that nothing in the record shows that U.S. suppliers would have raised Purchaser D’s prices in 1999. Pl.’s Resp. at 13. Although Purchaser D paid the same price for domestic TCCSS in 1998 and 1999, it did not buy Japanese imports until 1999. Given the insignificant levels of underselling with regard to this purchaser, it is only the general cost-price squeeze analysis that could support a finding of price suppression. The pricing to Purchaser D is not supportive.

d. Purchaser B

Finally, the court in Nippon II found that the Commission failed to address Purchaser B’s data. 223 F.Supp.2d at 1359. The Commission now examines the pricing data and finds a correlation between Japanese underselling and domestic price depression. Plaintiffs argue that there was no correlation because Purchaser B bought increasing volumes of domestic TCCSS over the period of investigation. The issue here, however, is price effects. From 1997 to 2000, Purchaser B obtained wider U.S. and Japanese discount rates.42 Although this evidence indicates a correlation between Japanese imports and domestic price depression over the period of investigation, Purchaser B did not consider underselling margins in 1999 to be significant and there was not a significant trend of underselling over the period of investigation. Consequently, this evidence is problematic. The most that can be said of the pricing data is that it pulls in two directions.

Thus, the Commission’s conclusion as to price depression or suppression is only supported by the generalized sub-conclusion of a cost-price squeeze. Neither lost sales analysis nor particularized pricing data positively correlates with Japanese import pricing. If the pricing data is viewed in the context of particularized underselling data it supports a negative determination. If viewed in the context of the generalized cost-price squeeze conclusion it is slightly supportive of price suppression. Overall, it is not helpful to the question of price depression or suppression.

D. Conditions of Competition

The material injury statute directs the Commission to evaluate all relevant economic factors, including price effects, “within the context of the ... conditions of competition that are distinctive to the affected industry.” 19 U.S.C. § 1677(7)(C); Nucor Corp., 318 F.Supp.2d at 1215. In its Final Determination, the Commission found that U.S. prices were affected by the industry’s high degree of price sensitivity and by the use of Japanese prices in domestic negotiations. Final Determ, at 8, 16. The court in Nippon II held that the Commission failed to assess these findings in the context of the “peculiar conditions of competition” of the TCCSS industry. 223 F.Supp.2d at 1350.

1. Price Sensitivity

An industry is price sensitive if small differences in price will induce purchasers *1205to shift from one supplier to another. Second Remand Determ, at 57-58. Price sensitivity is relevant to a material injury analysis because the more price sensitive the market, the greater the ability of lower-priced imports to impact the domestic industry’s sales and prices. Id. at 58. In this case, the Commission previously found that the TCCSS market was characterized by a high degree of price sensitivity. Final Determ, at 8. The court in Nippon II held that this finding was unsupported by substantial evidence, noting that the record indicated that non-price factors outrank the importance of price in purchasing decisions. 223 F.Supp.2d at 1361. In its Second Remand Determination, although the Commission provides a somewhat plausible explanation for its decision to give less weight to non-price factors highly ranked in purchaser questionnaires, the record shows that the TCCSS market is only low to moderately price sensitive.

a. The ITC’s Decision to Give Less Weight to Purchaser Responses to Questions IV-11 and III — 18 is Reasonable

In purchaser questionnaires, non-price factors, including quality and on-time delivery were consistently ranked as more important than price. As a result, the court instructed the Commission to consider these non-price factors in assessing whether the TCCSS market is price sensi-five. The Commission now explains that it accords less weight to the high ranking of non-price factors in purchaser questionnaires because (1) the wording of Question IV-11 is misleading, and (2) Questions TV-11 and III — 18 do not take into account the qualification process in the TCCSS industry.

First, Question TV-11 asks purchasers to rate various factors as “very important,” “somewhat important,” or “not important” to their purchasing decisions. Overall, “lowest price” was ranked seventh of approximately ten factors.43 Staff Report at 11-12, Table II-4. The Commission gives less weight to this high ranking of non-price factors, however, because it finds that the wording of question IV-11 .is misleading. Specifically, Question IV-11 refers to “lowest price” rather than simply “price.” The Commission contends that while price may be a key factor to purchasers’ sourcing decisions, obtaining the “lowest price” may not be crucial if other factors are not satisfactory. Second Remand Determ, at 61. Plaintiffs do not challenge this theory. Because price is the only factor in Question TV-11 to use a superlative modifier such as “lowest,” the wording may in fact be misleading.44

Second, the Commission explains that the high ranking of non-price factors in response to Questions TV-11 and III — 1845 may not be accurate because the questions *1206do not take into account the qualification process in the TCCSS industry. Plaintiffs argue that the Commission is merely repeating its previous arguments and not addressing the court’s concerns. The Commission, however, is directly responding to the court’s concern that it consider whether Questions IV-11 and III — 18 refer to the post-qualification stage. Nippon II, 223 F.Supp.2d at 1361 (noting that neither the Commission nor the Defendant-Inter-venors attempted to rebut the fact that “Questions III — 18 and IV-11 of the purchaser questionnaires clearly ask purchasers to rank the importance of ‘lowest price’ and other considerations in choosing among qualified suppliers only”). Moreover, the court agrees with the Commission that because Questions IV-11 and III — 18 do not reference a certain stage of the qualification process, their answers may not be entirely accurate with regard to purchaser considerations. Nonetheless, the responses cannot be discounted entirety-

The TCCSS industry is characterized by long-term relationships established through a qualification process. Purchasers begin by buying limited quantities from suppliers, and only after a potential supplier has proven that it can deliver the desired quality and quantity in a steady and rehable manner does it become qualified as a supplier. Staff Report at II — 10. Thus, when a qualified supplier enters negotiations with a purchaser, the general quality and reliability of that supplier’s product has been established, leaving price and volume the essential factors to be negotiated. See Hr’g Tr. at 66-67 (“If you’re not through the qualification process, which is determined by all these other non-price factors, you don’t get a chance to offer your price.”). Questions IV-11 and III — 18, however, do not ask purchasers to rank the importance of these factors at a certain stage of the qualification process, and do not reflect the fact that although quality and on-time delivery may be important considerations at the pre-qualification stage, price may be a more important factor after qualification. Thus, it is reasonable for the Commission to accord somewhat less weight to the responses to Questions IV-11 and III — 18, where purchasers ranked non-price factors as more important than price in their purchasing decisions.

Even though the Commission has provided a credible explanation for its decision to give less weight to these responses, it may not ignore the fact that quality and on-time delivery are nonetheless critically important considerations to TCCSS purchasers. See Staff Report at II — 10 (“quality was most frequently reported as the critical consideration in their purchasing decisions”); Koplan Dissent at 2 (“reliable delivery is extremely important to purchasers ...”); Final Determ, at 8 (“The record indicates that non-price factors such as product quality, product consistency, and on time delivery are very important in choosing suppliers.”).

b. The Record Shows that the TCCSS Market is Low to Moderately Price Sensitive

In its two prior determinations, the Commission found that the TCCSS market was “highly” price sensitive. Final De-term. at 8; First Remand Determ, at 17. The Commission now contends that the market is merely “price sensitive.” Second Remand Determ, at 63. Because the four categories of evidence relied upon by the Commission support a finding of low to moderate price sensitivity, the Commission’s analysis should reflect that less expensive subject imports had a low to moderate ability to effect domestic sales and prices.

*1207First, the Commission contends that the record shows that most purchasers were likely to shift suppliers based on small changes in price. Id. at 58. Evidence that purchasers shift suppliers in response to small changes in price indicates that a market is price sensitive. Plaintiffs correctly point out, however, that the evidence cited by the Commission .leads to precisely the opposite conclusion. Pl.’s Resp. at 17. For example, a memo cited by the Commission actually shows that Purchaser A was unwilling to shift volume in light of small price changes.46 In addition, evidence that purchasers negotiated discount rates to the hundredth percent does not indicate price sensitivity. See Nippon I, 182 F.Supp.2d at 1345 (noting that “it is [not] apparent why the degree of price specificity in negotiations would be necessarily indicative of price sensitivity”). Finally, purchaser estimates of determinative price differentials indicated that most purchasers were not likely to shift suppliers because of price differentials ranging from two to six percent. See discussion supra Section I.B.2.b.ii. Therefore, this evidence does not show that the domestic TCCSS market is price sensitive.

Second, the degree of substitutability between products is relevant to whether the market is price sensitive because “the more fungible the product, the more likely that purchasers will make decisions based on price differences ... [but] when products are highly differentiated,, price is less likely to determine product selection.” General Motors Corp. v. United States, 17 CIT 697, 706, 827 F.Supp. 774, 784 (1993). Although the Commission claims that domestic and Japanese TCCSS are highly substitutable, the record indicates that they actually have at best a moderate degree of substitutability. Although producers reported that Japanese and U.S. merchandise are “relatively close substitutes” with the exception of some Japanese specifications, respondents reported that the two products were less fungible.47 As a result, the Staff Report estimated the elasticity of substitution, or how easily purchasers switch suppliers when prices change, to be low to moderate.48 Id. at II-17-11-18.

Third, the Commission claims that the market is price sensitive because TCCSS accounts for a large percent of the total cost of production of many purchaser operations, and thus lowering the. cost of TCCSS is the most effective way to increase profit margins. Depending on the final product, however, TCCSS can be a very large or relatively small part of the final product cost.49 Because purchaser questionnaires indicate that TCCSS accounts for approximately half of their total cost of production, the evidence as a whole supports the conclusion that the market is, at most, moderately price sensitive.

*1208Fourth, the Commission argues that the price sensitivity of the market is reflected by the fact that purchasers consistently rated price as one of the three most important factors in their purchasing decisions. As discussed above, however, purchasers generally rated price as third in importance, after quality and on-time delivery. Therefore, this evidence does not support a finding of high price sensitivity.

In sum, the evidence relied upon by the Commission indicates that the TCCSS market was low to moderately price sensitive — that lower priced subject imports had a low to moderate ability to impact the domestic industry’s sales and prices. Contrary to the Commission’s assertion, this evidence does not support an affirmative material injury determination. Rather, it supports the opposite conclusion. See United States Steel Group v. United States, 18 CIT 1190, 1192, 1213, 873 F.Supp. 673, 680, 695 (1994) (holding that the Commission’s negative material injury determination reasonably supported by finding that market was not highly price sensitive, and imports were not greatly substitutable).

2. Negotiation Practices

In its Final Determination, the Commission found that contrary to purchaser testimony, importers often conduct negotiations simultaneously and use aggressive Japanese pricing in domestic negotiations to leverage lower prices. Final Determ. at 16. The court held that even if Japanese and domestic negotiations take place contemporaneously, the Commission must still address other evidence indicating that these negotiations run on separate tracks according to different procedures and criteria. Nippon II, 223 F.Supp.2d at 1352, 1361-62. Nonetheless, on its first remand, the Commission again failed to address evidence indicating that (a) price negotiations are bifurcated, (b) a division of major and minor tonnage keeps negotiations compartmentalized, (c) supply agreements prevent foreign price competition in domestic negotiations, (d) different delivery times segregate negotiations, and (e) Weir-ton did not consider Japanese competition when calculating its prices. Id. Although the Commission argues otherwise, this evidence continues to indicate that Japanese and domestic negotiations are compartmentalized.

a. Price Negotiations are Bifurcated

The court held that the Commission failed to address the extent to which bifurcation of price negotiations were representative of purchasers’ practices. Id. at 1362. In response, the Commission finds that there is a “significant body of evidence on the record” to show that prices of the Japanese product were in fact used to obtain lower prices from domestic producers.50 Second Remand Determ, at 66. Plaintiffs disagree. PI. ’s Resp. at 19. The evidence relied upon by the Commission does not substantially support its position.

First, the Commission continues to rely on internal negotiating memoranda from Purchasers E and F, which the court previously held were irrelevant. Nippon I, 182 F.Supp.2d at 1348, n. 37 (noting that only Purchaser F’s memo indicates that lower priced Japanese imports were taken into consideration during negotiations with domestic suppliers, but the price discussion appeared incidental). Second, the testimonies of Silgan and U.S. Can actually cut against the Commission’s finding that purchasers used subject pricing in domestic *1209negotiations.51 Third, a U.S. supplier’s questionnaire response indicates that it believed that the Japanese price was lower because Japanese volume increased, not because purchasers used aggressive pricing in negotiations.52 Fourth, Weirton’s claims are undermined by the fact that it was unable to document its allegations of Japanese price competition. See discussion infra Section I.D.2.e. Finally, although the Commission claims that six U.S. suppliers reported lost revenue, only three actually responded affirmatively when asked whether they reduced prices or rolled back announced price increases to avoid losing sales to Japanese competitors. See Producer Questionnaires at IVC. Of those three, only one reported specific instances of lost revenue.53 See Staff Report at V-22. Therefore, contrary to the Commission’s finding, there is not significant record evidence indicating that purchasers used subject imports to obtain lower domestic prices. Rather, price negotiations appear to be bifurcated.

b. Purchasers Buy Major Domestic Tonnage and Minor Japanese Tonnage

The court also held that the Commission failed to address the extent to which a division of major and minor tonnage between domestic and subject suppliers kept negotiations on separate tracks. Second Remand Determ, at 68-69. The Commission now finds that the fact that purchasers buy more domestic product than subject imports does not mean that domestic negotiations are sealed off from subject import pricing. Id. at 69. The Commission contends that the record shows that purchasers developed negotiation strategies encompassing all of their suppliers, domestic and foreign, before beginning negotiations with domestic mills. Plaintiffs argue that the evidence cited by the Commission-does not show that subject prices were used in domestic negotiations. Pl.’s Resp. at 19-20. Although the evidence cited by the Commission confirms that purchasers bought major tonnage from domestic suppliers and minor tonnage from subject suppliers, it does not show that purchasers used foreign pricing to leverage lower domestic prices.

For example, although an internal memo indicates that Purchaser F planned to allocate more volume from Japanese sources and less from domestic suppliers, it does not show that such volume allocation af*1210fected price negotiations.54 Similarly, although another memo indicates Purchaser F’s plan to buy more volume from foreign suppliers and less from domestic sources would save the purchaser money, it does not show that subject pricing was used to leverage lower domestic prices.55 Finally, the fact that subject import volume increased over the period of investigation does not prove that subject prices were used by purchasers to obtain lower prices. The issue here is price effects.

Therefore, although this evidence confirms that purchasers bought major volumes of domestic product and minor volumes of subject imports and indicates that subject import volume increased over the period of investigation, it does not show that purchasers used foreign pricing to leverage lower domestic prices.

c. Supply Agreements Limit Price Competition to Domestic Bids

Five purchasers lease facilities from two U.S. suppliers.56 As part of these leasing arrangements, purchaser/lessees and supplier/lessors enter into supply agreements, which set forth volume and pricing terms for extended periods of time. Staff Report at III — 3, n.5. Although these supply agreements vary, most require purchasers to buy the majority of their TCCSS from the supplier and require suppliers to compete only with other domestic prices.57 In Nippon II, the court held that the Commission did not address Plaintiffs arguments relating to the prevalence and impact of these supply agreements. 223 F.Supp.2d at 1363. In response, the Commission now finds that these supply agreements are not prevalent in the industry and do not preclude purchasers from using foreign prices to negotiate lower domestic prices. Second Remand Determ. at 72. Plaintiffs correctly assert that the Commission’s findings are unsupported by substantial evidence.

First, the Commission finds that supply agreements are not prevalent in the industry because the volume of TCCSS sold pursuant to such agreements over the period of investigation was a small percentage of the total apparent U.S. consumption.58 The apparent U.S. consumption, *1211however, includes total domestic consumption from all sources. Moreover, the Commission’s calculations reveal that the vast majority of TCCSS bought by purchaser/lessees from supplierflessors were pursuant to supply agreements.59 The fact that most of the TCCSS bought by four large purchasers from two of the seven domestic suppliers was insulated from Japanese price competition, weighs against the Commission’s conclusion that purchasers used Japanese pricing in domestic negotiations.

In addition, the Commission’s finding that supply agreements do not preclude purchasers from using foreign prices to negotiate lower domestic prices is unsupported by substantial evidence. The Commission contends that the language of supply agreements, which imposes an obligation on the supplier, does not preclude a purchaser from using subject prices in negotiations with suppliers. Second Remand Determ, at 72-73. Plaintiffs argue that this argument is inconsequential because the record does not show that purchasers used subject imports to leverage lower prices. Pl.’s Resp. at 21. The Commission is unable to cite an instance when a purchaser, party to a supply agreement, used subject pricing to negotiate lower domestic prices.60 Even if such a purchaser attempted to use subject pricing to negotiate lower domestic prices, however, a supplier would be insulated, at least partly, from meeting those prices. Indeed, the record shows that domestic suppliers simply refused to compete with foreign prices. See Hr’g Tr. at 224 (“The domestic mills do not recognize foreign mill prices as competitive situations that they can — they choose to meet or being asked to meet. They flatly, absolutely do not recognize it.”).

d. Domestic Producers Enjoyed a Lead Time Advantage

The court also held that the Commission failed to evaluate purchaser perceptions with respect to the domestic industry’s lead time advantage as a potential explanation for keeping negotiations compartmentalized. Nippon II, 223 F.Supp.2d at 1363. In response, the Commission acknowledges that the proximity of domestic mills to their purchasers generally gives them a lead time advantage, but it finds that this advantage is mitigated by several factors. Second Remand Determ, at 75. Specifically, the Commission cites evidence that purchasers minimize (but concededly do not eliminate) longer delivery times by negotiating for core specifications in advance of their production needs, and requiring Japanese suppliers to carry the cost of larger consignment inventories at storage facilities in the United States. Id. at 75-77. Plaintiffs argue that this evidence cuts against the Commission’s conclusion. PI. ’s Resp. at 21. The fact that purchasers negotiate for advance specifications and for consignment inventories with Japanese suppliers but not with domestic suppliers, *1212undercuts the Commission’s finding that U.S. and Japanese negotiations take place on equal footing.61

e. Weirton was Unable to Provide Documentation of Alleged Japanese Price Competition

At the public hearing, the Commission requested that Weirton submit documentation to support its claim that subject import pricing damaged its negotiating leverage. In response, Weirton submitted “Competitive Price Allowance” sheets for several purchasers during the period of investigation and conceded that “the competitors listed ... are always other domestic firms.” See Pet’r Posthr’g Br., Ex. 20, at 1, (A.R.2-104), ITC App., Tab 1, at 4. Based on this information, the court found that Weirton “apparently derives its pricing allowance range solely according to pricing data of domestic producers submitted by its sales department.” Nippon I 182 F.Supp.2d at 1348. The Commission disagreed and accorded little weight to Weirton’s inability to evidence its Japanese price competition. First Remand Determ. at 22-23. The court in Nippon II held that the Commission failed to justify its decision to accord little weight to this lack of evidence. 223 F.Supp.2d at 1364. Nonetheless, the Commission again finds that Weriton’s lack of documentation should be accorded little weight. Second Remand Determ, at 78-80. Plaintiffs counter that since Weirton claimed to be highly sensitive to the alleged competitive pressures of subject imports, this lack of documentation is significant. PL’s Resp. at 22.

This court has held that “what a party says, or does not say, concerning its economic condition can be important evidence of injury or lack thereof.” Allegheny Ludlum Corp. v. United States, 24 CIT 858, 116 F.Supp.2d 1276, 1299 (2000) (citing Suramerica De Aleaciones Laminadas, CA v. United States, 44 F.3d 978, 984 (Fed.Cir.1994)). The Defendant-Interve-nor here, Weirton, alleges that it adjusted pricing because of Japanese competition. Nonetheless, it is unable to back up this allegation with documentation. The Commission reasons that this lack of documentation is insignificant because purchasers do not specify the identity of suppliers with which they are negotiating, but generally mention offshore suppliers to obtain lower domestic prices. For support, the Commission notes that Weirton only discovered “after the fact” that it was competing with Japanese prices.62 Under this explanation, however, it is not clear why Weirton had documentation regarding specific domestic competitors, did not document “offshore” competition, and did not document Japanese competition “after the fact.”

Accordingly, the Commission’s explanation for according little weight to this lack of documentation is unsupported by substantial evidence, and the fact that Weir-ton — a party to this action and principal supporter of the petition — is unable to pro*1213vide evidence supporting its allegations, is important evidence of lack of injury.

II. Causation

After assessing whether the volume, price effects, and impact of the subject imports on the domestic industry are significant, the statutory “by reason of’ language implicitly requires the Commission to “determine whether [these] factors as a whole indicate that the subject imports themselves made a material contribution to the overall injury.” Taiwan Semiconductor v. United States, 23 CIT 410, 414, 59 F.Supp.2d 1324, 1329 (1999) (quoting Gerald Metals, Inc. v. United States, 132 F.3d 716 (Fed.Cir.1997)); see also 19 U.S.C. § 1673d(b)(l). This “by reason of’ language “mandates a showing of causal — not merely temporal — connection between the [subject imports] and the material injury.” Gerald Metals, Inc., 132 F.3d at 720. To establish this causal connection, a minimal or tangential contribution to the material harm is insufficient. Id. at 722.

As part of this analysis, “the Commission must examine other factors to ensure that it is not attributing injury from other sources to the subject imports.” Statement of Administrative Action, H.R. Doc. No. 316, 103rd Cong., 2nd Sess. (1994), reprinted in Uruguay Round Agreements Act, Legislative History, Vol. VI, at 851-52. Although the Commission need not find that alternative causes entirely negate the likelihood of subject imports having any adverse impact on domestic pricing, it must determine whether “their combined effect may dilute the effect of the LTFV imports, preventing [them] from being a material factor.” Taiwan Semiconductor v. United States, 93 F.Supp.2d 1283, 1291 (CIT 2000) (quoting Gerald Metals,. Inc., 22 CIT at 1014, 27 F. Supp 2d at 1355 n. 8). As explained by this court,

Frequently, several events — each of which is a necessary antecedent and has án appreciable effect — contribute to overall injury to an industry. In some cases,- another event may have such a predominant effect in producing the harm as to-make the effect of the LTFV imports insignificant and, therefore, to prevent the LTFV imports from being a material factor. (This is not to say, however, that there may not be more than one material factor to injury.) In addition, even if no contributing factors independently have a predominant effect, their combined effect may dilute the effect of the LTFV imports, preventing the LTFV imports from being a material factor. The statute requires that the Commission determine whether the LTFV imports themselves made a material contribution to the injury suffered by the domestic industry.

Gerald Metals, 22 CIT at 1015, 27 F.Supp.2d at 1355.

In this case, the court held that the Commission had not sufficiently ensured that it was accurately attributing the harmful effects to lower-priced subject imports because “[t]he record reflects that the increased subject import volume must be attributed largely to purchaser priorities that are unrelated to. price.” Nippon II, 223 F.Supp.2d at 1371. In particular, the court held that the Commission had not reasonably explained why injury was not caused by (A) U.S. quality and delivery problems, or (B) non-subject imports.

A. The Record Consistently shows that Purchasers Bought Subject Imports Largely because of U.S. Quality and Reliability Problems

Non-price factors such as product quality and reliable delivery are important considerations to TCCSS purchasers. Many purchasers testified that they experienced *1214U.S. quality and on-time delivery problems over the period of investigation. Final Determ, at 26. Nonetheless, in its First Remand Determination, the Commission was not persuaded by this evidence, which it found to be “inconsistent and contradictory” with other record evidence. First Remand Determ, at 29-41. In contrast, the court held that the evidence was not inconsistent. Nippon II, 223 F.Supp.2d at 1367-69. In its Second Remand Determination, the Commission continues to find that evidence that purchasers turned to Japanese sourcing solely because of domestic quality and delivery time problems is inconsistent with other evidence showing that purchasers bought subject imports because of price. Second Remand Determ. at 87-117. Plaintiffs argue that the record consistently shows that purchasers turned to Japanese imports for these non-price reasons. PI. ’s Resp. at 24.

An agency has the discretion to weigh and judge the credibility of conflicting evidence. Nippon III, 345 F.3d at 1381: Chung Ling Co. v. United States, 16 CIT 636, 648, 805 F.Supp. 45, 55 (1992). While the court may not substitute its judgment for that of the agency, the agency’s findings must be reasonable and supported by substantial evidence. Bratsk Aluminum Smelter, Slip Op. 04-75 at 10. In this case, because the evidence cited by the Commission reinforces purchaser testimony by showing that purchasers bought subject imports because of quality and delivery considerations, the Commission’s finding that they are inconsistent is unreasonable and unsupported by substantial evidence. Contrary to the Commission’s implication, the fact that some of the same evidence indicates that purchasers also considered price when making purchasing decisions does not establish a conflict. Indeed, most purchasers reported that several factors drive their purchasing decisions. The question is whether subject imports made a material contribution to the injury, or whether the combined effect of other factors were of such a magnitude as to prevent subject imports from being a material factor in the injury. See Taiwan Semiconductor, 23 CIT at 416, 59 F.Supp.2d at 1331.

Although some evidence indicates that price was a purchasing factor, the record on the whole consistently shows that purchasers bought subject imports largely because of U.S. quality and reliability problems. In light of this consistent evidence, the Commission cannot reasonably ensure that it accurately attributed harm to lower priced subject imports rather than to these other problems. Moreover, the combined effect of these problems may have prevented lower-priced Japanese imports from being a material factor in the injury. Indeed, the record shows neither that price was a material factor, nor that subject imports had a significant effect on domestic prices. See discussion supra Section I; see also Gerald Metals, Inc., 22 CIT at 1014, 27 F.Supp.2d at 1356 (noting that factors as a whole must indicate that LTFV imports themselves made a material contribution to the injury). Consequently, the record fails to show that harm was by reason of subject imports.

1. BWAY

At the public hearing, a BWAY representative testified that due to “a series of delivery and quality disappointments with certain U.S. mills,” BWAY made “a strategic decision to diversify its sourcing including additional sourcing [of TCCSS] from abroad.” Hr’g Tr. at 190. Nonetheless, the Commission finds that other record evidence contradicts BWAY’s testimony and shows that BWAY purchased subject imports based on price. Second Remand Determ, at 95. The court agrees with Plaintiffs, however, that the record does *1215not support the Commission’s conclusion. PI. ’s Resp. at 25.

First, BWAY’s questionnaire response is consistent with its testimonial evidence. Although the Commission notes that BWAY’s questionnaire response states that it decided to take advantage of price differences offered by foreign suppliers over the period of investigation, the next sentence states that BWAY had concerns with domestic supply disruptions and quality problems.63 Similarly, BWAY stated elsewhere in its questionnaire that although it dropped two U.S. suppliers due to price and commercial terms, it cut two others because of poor performance related to quality and delivery.64 Therefore, BWAY’s questionnaire and testimony do not contradict the quality and delivery representations.

Second, BWAY’s comparison of domestic and Japanese product corresponds with its hearing testimony, that it considered quality an important purchasing consideration. In its comparison, although BWAY rated U.S. delivery time as superior to the Japanese,65 it rated the quality of Japanese TCCSS as superior to the U.S.

Third, although the Commission suggests otherwise, BWAY’s failure to report any quality or delivery problems with some U.S. suppliers is not inconsistent with its hearing testimony. In the hearing, the BWAY representative stated that BWAY had problems “with certain U.S. mills.” Hr’g Tr. at 190. In addition, the Commission even concedes that this testimony is backed up by BWAY’s questionnaire, in which it reported that it shifted volumes from a certain U.S. supplier because of quality and delivery concerns.66

Therefore, the evidence cited by the Commission consistently indicates that BWAY purchased subject imports because of U.S. quality and reliable delivery concerns.

2. Crown

Crown indicated that it increased purchases of Japanese imports in 1999 because of quality and performance problems, and a shortage of West Coast supply.67 Although the Commission rejects the credibility of Crown’s assertion on the grounds that it is inconsistent with other record evidence, the Commission again fails to provide evidence that contradicts Crown’s explanation for its shift to Japanese imports.

First, the evidence cited by the Commission does not contradict Crown’s explanation that it shifted to Japanese imports due to a lack of alternate East Coast suppliers. Although the Commission notes that a certain East Coast supplier had ample available capacity that could have *1216been used to supply the volumes Crown sourced from Japan in 1999,68 it fails to point out that this supplier reported on-time delivery problems associated with railway complications from June 1999 to 2000. Staff Report at III — 2 n.2. Therefore, even assuming that this supplier had available capacity, the record indicates that it could not meet Crown’s West Coast on-time delivery requirements in 1999.69 Indeed, Crown noted that “delivery performance” was a factor in its decision to buy Japanese imports that year.

Second, Crown’s comparison of Japanese and U.S. merchandise bolsters its explanation that it shifted to Japanese imports because of quality considerations. In the comparison, although Crown rated U.S. availability and delivery time as superior to that of the Japanese suppliers, it rated Japanese quality and consistency superior to that of the U.S. Crown Questionnaire Resp. at IV-10. In addition, the fact that Crown rated Japanese and U.S. prices as comparable suggests that it purchased from these sources based on factors other than price. Id.

Third, evidence that Crown informed a U.S. producer that Japanese prices were lower than U.S. prices neither contradicts its explanation for why it purchased subject imports, nor constitutes substantial evidence that Crown bought subject imports because of price.70 It is undisputed that Japanese prices were lower than U.S. prices in 1999. The fact that “economic harm, to domestic industry occurred when LTFV imports [we]re also on the market is not enough to show, that the imports caused a material injury.” Gerald Metals, Inc., 132 F.3d at 719. .

In sum, the Commission fails to provide substantial evidence to discredit Crown’s explanation'that it shifted to subject imports because of inadequate U.S. quality and reliable delivery problems.

3. Silgan

Silgan testified that it purchased Japanese imports for .specialized applications that were either not available from the domestic industry or of a “quality level” not obtainable from U.S. producers. Hr’g Tr. at 200-02. Silgan also attributed its increase in Japanese imports to its acquisition of Campbell’s Soup, which used small quantities of TCCSS produced by Nippon because of superior quality and unique specifications not available from U.S. producers. Id. at 202. In addition, Silgan testified that it terminated Weirton as a supplier for failing to meet Silgan’s quality and service requirements. Id. at 200-01. Finally, Silgan stated that if it were to purchase according to price,' it would purchase from Brazil, Korea, and Taiwan. Although the Commission finds that this tes*1217timony is undermined by other record statements indicating that Silgan based its purchasing decisions on price, it fails to support this finding with substantial evidence.

First, the fact that Silgan increased its purchases of Japanese imports at the same time Japanese producers were underbidding domestic producers is not inconsistent with Silgan’s testimony. In fact, it corresponds with Silgan testimony that it increased its purchases of subject imports during the period of investigation due to its acquisition of Campbell’s Soup.71 Moreover, as discussed above, a temporal connection between LTFV imports and material harm is insufficient to establish causation. Gerald Metals, Inc., 132 F.3d at 719 (citing United States Steel Group v. United States, 96 F.3d 1352, 1358 (Fed.Cir.1996)) (“To claim that the temporal link between these events proves that they are causally related is ... fallacy....”).

Second, the Commission concedes that Silgan’s questionnaire response and hearing testimony verify that Silgan discontinued buying from Weirton because of quality problems. Second Remand Determ, at 114-15. In a weak attempt to discredit this consistent evidence, the Commission reasons that “Silgan made no other comments on this issue in its questionnaire response, did not describe these failures in any detail, and provided no documents to the Commission showing what quality or delivery problems Weirton was having pri- or to 1998.” Id. at 115. This lack of particular types of evidence, however, does not undermine otherwise consistent evidence.72

Third, other statements by Silgan support its testimony that it purchased Japanese imports because U.S. suppliers were unable to meet certain specifications. A Silgan representative testified, “we purchased ... wide tin-free steel from Japan ... because our equipment is designed to run ... wide coils- No U.S. mill can provide ... wide tin-free steel. If we buy U.S. produced narrower coils, we lost [a percentage] of our output. This not only affects our costs, but it constrains our capacity.” Hr’g Tr. at 201. The Commission claims that this statement contradicts its testimony by proving that “Silgan was able to, and did, use the domestic product in place of the wider product supplied by the Japanese.” Second Remand Determ. at 117. On the contrary, this statement is consistent with Silgan’s testimony that U.S. suppliers could not meet Silgan’s “specialized applications.”73 Hr’g Tr. at 202.

Accordingly, the record consistently shows that Silgan purchased subject im*1218ports largely because of non-price concerns.

4. U.S. Can

U.S. Can testified that it increased its volume of Japanese imports because of U.S. quality and on-time delivery problems, and because of the desire to source TCCSS on a hiore global basis to serve its increasingly international operations. Hr’g Tr. at 196-97. Although the Commission concedes that this testimony is consistent with U.S. Can’s testimony at the preliminary staff conference, it finds that it is undermined by other record evidence. The court agrees with Plaintiffs, however, that the statements cited by the Commission are neither inconsistent nor contradictory.

First, purchasing data for 1999 substantiates U.S. Can’s hearing testimony that it increased its purchases of Japanese imports while decreasing its purchases of U.S. TCCSS. The Commission asserts that this purchasing history conflicts with another U.S. Can statement: “[U.S. Can] did not favor the Japanese mills and take tons away from Weirton Steel.... Japanese tonnage [was] reduced [during certain months in 1999] for reasons of inventory control.” Id. at 199. Evidence indicating that U.S. Can increased its total 1999 Japanese tonnage is not necessarily inconsistent with testimony that U.S. Can decreased Japanese tonnage. during certain months of 1999. Moreover, as Plaintiffs point out this evidence is not relevant to causation. It is undisputed that U.S. Can increased its purchases of Japanese imports in 1999; the issue is whether or not it did so by reason of Japanese pricing.74

Second, two internal memos regarding U.S. Can’s negotiations with a particular U.S. supplier reinforce U.S. Can’s hearing testimony that it increased purchases from Japan due to U.S. reliability problems. Although noting that foreign suppliers were offering “aggressive pricing,” both memos emphasized that U.S. Can was experiencing on-time performance problems with U.S. suppliers and reduced volumes accordingly.75 Therefore, these internal memos reinforce U.S. Can’s hearing testimony.

Third, U.S. Can’s rating of U.S. and Japanese TCCSS also supports its hearing testimony. Although U.S. Can rated U.S. delivery time and availability as superior to that of the Japanese, it rated Japanese quality, consistency, and reliability superi- or to the that of the U.S. suppliers. U.S. Can Questionnaire Resp. at TV-10. This supports U.S. Can’s testimony that it increased its purchases of Japanese imports due to U.S. quality and reliability problems.76

*1219Therefore, the evidence cited by the Commission is fully consistent with U.S. Can’s hearing testimony that it increased its purchases of subject imports for reasons other than price.

B. The Record indicates that Non-Subject Imports were a Significant Factor in the U.S. Market

During the period of investigation, non-subject imports accounted for a greater proportion of total U.S. market share than Japanese imports.77 Final Determ, at 10. Nonetheless, the Commission rejected the contention that non-subject imports accounted for declines in domestic pricing, finding that “[ajlthough non-subject imports were a significant factor in the domestic market during the period of investigation, subject imports grew more rapidly and were generally priced more aggressively.” Id. at 22. The court in Nippon I held that this finding was unsupported by substantial evidence because the Commission failed to consider whether non-subject imports were predominant in the regions where the majority of domestic shipments were concentrated, and failed to compile its price comparison data in a manner to facilitate the court’s review. 182 F.Supp.2d at 1354-55. On its first remand, the Commission did not comply with the court’s instructions. Nippon II, 223 F.Supp.2d at 1369-71. In its Second Remand Determination, the Commission again finds that non-subject imports were not the predominant cause of injury to the domestic industry during the period of investigation. Second Remand Determ, at 127. In making this determination, however, the Commission fails to (1) evaluate whether non-subject import volume was predominant in regions where the majority of domestic shipments were concentrated, and (2) reasonably interpret non-subject import pricing data.

1. The ITC Fails to Evaluate whether Non-subject Import Volume was Predominant in Regions where the Majority of Domestic Shipments were made

The majority of U.S. producers are located in the East and Midwest and supply purchasers in those regions. Staff Report at II — 1. Japanese producers, while competing heavily in the West, supply purchasers throughout the United States. Final De-term. at 10. Non-subject producers compete only in the Eastern and Midwestern United States. Id. In light of this regional competition, the court instructed the Commission to evaluate whether non-subject imports were predominant in regions where the majority,of domestic shipments were concentrated. The court explained that “[e]ven if subject and non-subject market share levels are ‘comparable’ on the whole, non-subject imports are not necessarily precluded from constituting the predominant source of injury where, as in this case, they are concentrated in regions to which most domestic shipments were made.” Nippon I, 182 F.Supp.2d at 1355.

In its Second Remand Determination, the Commission finds a correlation between the increased volumes of subject imports to the West Coast, and the apparent financial declines of the domestic West Coast producers. Second Remand De-term. at 122. Plaintiffs argue that because the majority of domestic TCCSS shipments and non-subject import volumes are concentrated on the East Coast, non-subject imports constituted the predominant source of injury during the period. Pl.’s Resp. at 34-35. The Commission once again refuses to comply with the court’s *1220instructions.78 By refusing to assess non-subject import volumes in the East and Midwest, the Commission has no basis for ensuring that it did not attribute the harmful effects from non-subject imports to the subject imports.

2. The ITC Fails to Reasonably Interpret Non-subject Import Pricing Data

In Nippon II, the court criticized the Commission’s analysis of the underselling patterns of subject imports compared to non-subject imports. 223 F.Supp.2d at 1370-71. The court held that by collapsing the data into two-year increments, the Commission obscured the fact that there is no clear pattern when analyzed from year to year. Id. On its second remand, the Commission prepared new charts and represented the data. Based on the revised data, the Commission finds that underselling by non-subject imports and subject imports over the period of investigation was “mixed.!’ Second Remand Determ, at 125. This mixed data, the Commission contends, establishes that subject imports had more than a minimal or tangential amount of injury on the domestic industry. Id. at 126. Plaintiffs argue, on the other hand, that the data is not mixed, but rather when analyzed on an individual purchaser basis, shows that non-subject imports almost always undersold subject imports. PI. ’s Resp. at 35.

The Commission presents two charts. The first chart compares the weighted average prices that six large purchasers paid for Japanese and non-subject merchandise over the period of investigation. Second Remand Determ, at 124. It reveals that over the period of investigation, three purchasers paid less for non-subject TCCSS, two paid less for Japanese TCCSS, and one paid less for Japanese TCCSS in 1997 and 1998, but more in 1999.79

*1221The second chart compares non-subject bids to Japanese bids. Although there were more total non-subject bids “above” than “below” all Japanese bids over the entire period of investigation, from 1997 to 1999, the number of non-subject bids “below all Japanese bids” steadily increased. In addition, in 1999, the year in which the Commission finds the most instances of subject underselling, there were more non-subject bids “below all Japanese bids” than “within” or “above.” Second Remand De-tem. at 125, Table Second Remand 5.

Although subject imports need not be the sole or principal cause of injury, Nippon III, at 1381, “a positive correlation concerning non-subject importfs] ... in conjunction with other factors, may be sufficient to cut the causal connection between subject imports and any harm suffered by the domestic industry.” Altx, Inc. v. United States, 167 F.Supp.2d 1353, 1361-63 (CIT 2001). In this case, the record not only indicates that non-subject imports potentially had a significant effect on U.S. prices, but it also affirmatively demonstrates that quality and delivery problems affected the domestic industry. These aspects of the record indicate a gap in the causal connection between lower priced imports and material harm to the domestic industry. See Gerald Metals, Inc., 132 F.3d at 721. The record simply does not support a finding that subject imports caused injury to the domestic industry.

CONCLUSION

The Commission’s affirmative injury determination is unsupported by substantial evidence. As indicated previously, the record does show some increase in the volume of subject imports, but despite some isolated fragments of positive evidence, the record does not show that subject imports had a significant effect on domestic prices, or that purchasers bought significant volumes of subject imports by reason of lower prices. On the contrary, because the record shows that the effect of subject imports on domestic prices was insignificant and that harm suffered by the domestic industry was not caused by lower-priced Japanese TCCSS, it compels a negative material injury determination. See INS v. Elias-Zacarias, 502 U.S. 478, 481, 112 S.Ct. 812, 117 L.Ed.2d 38 (1992) (“To reverse the [agency’s] finding we must find that the evidence not only supports that conclusion, but compels it ... ”) (emphasis in original).

Regarding price effects, not only was Japanese underselling and domestic price depression or suppression insignificant over the period of investigation, but certain conditions of competition also minimized any effect subject imports could have had on domestic prices. Specifically, lower-priced imports had only a low to moderate ability to impact the domestic industry’s sales and prices; Japanese and U.S. price negotiations were compartmentalized; the majority of the industry’s product was supplied by domestic producers; several large purchasers bought U.S. TCCSS pursuant to supply agreements that limit price competition domestically; superior domestic lead times seem to translate into price premiums and segregate Japanese and U.S. price negotiations; and the principal supporter of the petition had no documentary evidence of Japanese price competition. Therefore, the record indicates that the effect of subject imports on domestic prices was not significant.

With respect to causation, the record does not show that lower-priced Japanese imports were a material factor in the domestic industry’s harm. On the other hand, it does show that purchasers bought increased volumes of Japanese imports because of concerns with domestic producers’ *1222product quality and reliable delivery, and that non-subject imports were an important competitive factor in the domestic market during the period of investigation. Thus, the record shows that the harm suffered by the domestic industry was not by reason of subject imports.

In sum, -the record fully supports a negative determination and will not support an affirmative one. The court has considered whether to leave to the Commission’s discretion, as it ordinarily would, the issue of reopening the record for further investigation, particularly because non-subject imports were not fully studied, but such information would not change the result. It likely would be more support for a negative determination. The agency has had three opportunities to investigate this matter and its attempts to obtain new supportive information on price effects have not been successful. Further, it is not fair to Plaintiffs to delay this matter when lack of adequate investigation is not the primary problem. While flawed, the investigation gathered most of the relevant material. It simply does not support an affirmative determination.

Accordingly, the court concludes that because the Commission is unable to obtain new evidence to significantly supplement the record, due to the passage of time and other reasons, further investigation or reconsideration in this matter is futile. The Commission’s Second Remand Determination is remanded with instructions to issue a negative material injury determination.80 See 19 U.S.C. § 1516a (“If the [court’s] final disposition of an action brought under this section is not in harmony with the published determination of ... the Commission, the matter shall be remanded to ... the Commission, as appropriate, for disposition consistent with the final disposition of the court.”).

The court has previously declined to remand this matter for a determination of threat of material injury, largely on the basis that Weirton neither raised the issue of threat before the court, nor presented a viable threat case in its post-hearing brief before the Commission. Nippon Steel Corp. v. United States, No. 00-09-00479, 2002 WL 31163095 (CIT Sept. 26, 2002). Upon further review of Nippon III, the court concludes that it is better practice for the agency in the first instance to determine whether a threat of injury dispute remains.

The remand determination is to issue within sixty days hereof. Objections *1223thereto may be filed within eleven days thereafter.