Oeinion
Restani, Judge:
In Minivans from Japan, USITC Pub. 2529, Inv. No. 731-TA-522 (July 1992) (“Final Det. ”), a majority of the commissioners [698]*698of the International Trade Commission (“ITC”) found that the United States minivan industry was not materially injured or threatened with material injury due to less than value (“LTFV”) sales of Japanese minivans. Plaintiffs, General Motors Corporation (“GM”), Ford Motor Company, and Chrysler Corporation, challenge the determination.1
I. Background
A. The Minivans:
Minivans are on-road motor vehicles with certain distinct characteristics; the primary feature distinguishing minivans from other passenger vehicles is substantial interior space. The plaintiffs, GM, Ford and Chrysler, are the only U.S. producers of minivans. Chrysler also produces minivans in Ontario, Canada; these minivans are imported into the United States. The defendant-intervenors, Toyota Motor Corporation, Toyota Motor Sales, U.S.A., Inc., Mazda Motor Corporation and Mazda Motor Sales, Inc. (“Toyota” and “Mazda”), accounted for most of the minivans imported from Japan during the period of investigation.2
Chrysler introduced its minivans in 1983. The first Chrysler minivans, the Dodge Caravan and Plymouth Voyager, were manufactured in Canada. In 1987, Chrysler began to offer “long wheelbase” versions of its minivans. These vehicles, called the Dodge Grand Caravan and the Plymouth Grand Voyager, were produced in the United States, and had added body length between the front and rear axles. In 1989, Chrysler introduced the Town and Country minivan, which is manufactured in the United States.3
In late 1984, GM introduced the Chevrolet Astro and GMC Safari. The Astro/Safari minivans were built on truck frames, and as a result were less car-like in handling than the Chrysler minivans. In 1989, GM introduced “extended” Astro/Safari minivans, which had added body length behind the rear axle. In 1989, GM produced a second minivan series, consisting of the Chevrolet Lumina APV Pontiac Trans Sport, and Oldsmobile Silhouette (“APV triplets”). The APV triplets were intended to be more passenger oriented than the Astro/Safari vans.
Ford introduced its first minivan, the Ford Aerostar, in 1985. The Aerostar, like the GM Astro/Safari, was built on a truck frame. In 1988, Ford produced an “extended” Aerostar.
The only Japanese minivans currently produced for the U.S. market are the Mazda MPV (multi-purpose vehicle) and the Toyota Previa. Neither minivan is available in an extended or long wheelbase version. The Mazda MPV was introduced in late 1988 as a 1989 model. The Toyota Previa was introduced in February 1990, as a replacement for the first minivan that Toyota introduced into the U.S. market in 1983.
[699]*699B. The Antidumping Investigation:
On May 31, 1991, GM, Ford and Chrysler filed an antidumping petition, alleging that an industry in the United States was injured or threatened with material injury by reason of LTFV sales of Japanese minivans. On July 15, 1991, ITC notified the Department of Commerce that there was a reasonable indication of material injury to the U.S. industry producing minivans. Minivans from Japan, USITC Pub. 2402, Inv. No. 731-TA-522 (July 1991) (preliminary). On January 2,1992, the Department of Commerce issued a preliminary determination, finding that Japanese minivans were sold in the United States at LTFV New Minivans from Japan, 57 Fed. Reg. 43 (Dep’t Comm. 1992) (preliminary determination of sales at LTFV). On May 26, the Department of Commerce published a final affirmative determination. New Minivans from Japan, 57 Fed. Reg. 21,937 (May 26,1992) (final determination of sales at less than fair value).
ITC then instituted its final investigation. On June 24, 1992, ITC reached a negative determination, and on July 3, 1992, issued its opinion. The notice of final negative determination was published in the Federal Register on July 15, 1992. Minivans from Japan, 57 Fed. Reg. 31,388 (July 15, 1992).
C. The ITC Determination:
The plurality defined the like product as minivans, and the domestic industry as minivan producers. Final Det., at 5, 8-12; 19 U.S.C. § 1677(4), (10) (1988).4
It then considered the state of the domestic industry. 19 U.S.C. § 1677(7)(C)(iii). It rejected the plaintiffs’ request to consider several factors in this assessment, including the impact of lost Canadian sales on Chrysler’s U.S. production, the impact of lost minivan sales on sales of other less fuel efficient vehicles, and the impact of “brand loyalty” considerations on future lost sales of other vehicles. Final Det., at 7-10. It found that the industry was not mature, and that in a growing market a new minivan model expands market size without displacing existing sales. Id. at 17-18. It noted certain economic conditions in the domestic industry and additional factors related to product design and marketing that affected the condition of the domestic industry.5 Id. at 18-20. Based on all these factors, ITC found adverse industry trends in the domestic industry, with mixed trends in 1989 and 1990 and declining trends in 1990 and 1991. Id. at 20-21.
The plurality then considered whether these adverse trends were caused by LTFV imports, taking into account the volume and price effect of imports and their impact on the domestic industry. See 19 U.S.C. § 1677(7)(B). In terms of volume, it found that the domestic industry ac[700]*700counted for the majority of shipments during the period of investigation; Canadian imports accounted for a minimum of twenty percent; and subject imports comprised less than fifteen percent of the market. Final Det., at 22. In terms of price effects, it found domestic prices were not suppressed to a significant degree by subject imports. Id. at 30. It found that due to a number of factors domestic and Japanese minivans were of limited substitutability, so price was unlikely to determine vehicle choice. Id. at 29. The plurality also found no evidence of underselling. Id. at 31. It reached this conclusion after expressing reluctance to rely on certain price data, which it viewed as flawed, and, in any event, not probative due to limited substitutability. Id. at 30-31. It refused to place great weight on the interim data for 1992. Id. at 34. In conclusion, the plurality reached a negative determination, finding volume, price effect, and impact on the domestic industry insignificant. Id. at 36.
Commissioner Rohr also reached a negative determination, but issued a separate opinion. He agreed with the plurality’s findings concerning like product, the domestic industry, and industry trends. Id. at 44, 48. He found that the increased volume of subject imports had little impact on domestic sales, due to several factors: the increase was chiefly due to introduction of the Toyota Previa, which created its own demand; and only small percentages of buyers of either a domestic or imported minivan considered purchase of the other. Id. at 50. He also found no evidence of price depression or suppression due to LTFV imports, and like the plurality, discounted the 1992 data Id. at 51, 53.
Finally, neither the plurality nor Commissioner Rohr found that the evidence indicated a threat of material injury.6 Id. at 37, 55-57.
II. Standard of Review
ITC’s determination shall be upheld unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B).
III. Discussion
A. Applicable Law:
ITC’s task is to determine whether an industry in the United States is materially injured, or threatened with material injury by reason of imports or sales at LTFV See 19 U.S.C. § 1673d(b)(1). To reach a determination, ITC considers import volume, effect on domestic prices, impact on domestic producers, and other relevant economic factors. 19 U.S.C. § 1677(7)(B).
B. Statutory Construction:
The plaintiffs argue that ITC erred in failing to take into account the following factors: the adverse effect of lost Canadian sales on Chrysler’s U.S. operations; the negative impact of lost minivan sales on U.S. pro[701]*701ducers’ ability to sell less fuel efficient vehicles; and the loss of future sales of other vehicles due to brand loyalty considerations.
1. Lost Canadian Sales:
The plaintiffs argue that because planning, parts production, and administrative costs are shared between U.S. and Canadian operations, lost Canadian sales lead to increased per-unit costs for vehicles manufactured in the United States. They also argue that lost Canadian sales mean more intense competition between U.S. and Canadian minivans. The plaintiffs claim that ITC should have considered these factors. The plaintiffs cite statutory provisions that give ITC discretion to consider all relevant economic factors.7 Neither the antidumping statute nor the legislative history supports the plaintiffs’ position.
The statute provides that ITC is to consider the impact of imports “on domestic producers of like products, but only in the context of production operations within the United States.” 19 U.S.C. § 167T(7)(B)(i)(III) (emphasis added). The legislative history states that foreign or offshore operations of a U.S. producer “should not be considered part of the domestic industry for injury purposes” and “are not to be considered in measuring the impact of imports on the domestic industry.” S. Rep. No. 71, 100th Cong., 1st Sess. 115 (1987); H. Rep. No. 40, Part 1, 100th Cong., 1st Sess., 128 (1987). As the majority and the dissenting commissioners found, the specific instruction to disregard foreign production operations takes precedence over more general provisions granting discretion to consider all relevant factors. ITC’s interpretation is consistent with the statutory language and legislative history, and will be upheld.
2. Lost “CAFE" and Brand Loyalty Benefits:
Under the corporate average fuel economy (“CAFE”) standards of the 1975 Energy Policy and Conservation Act, as amended, 15 U.S.C. §§ 2001-2012, the government sets average fuel economy standards for passenger cars and trucks. For CAFE purposes, the minivans at issue are classified as trucks. As minivans have a fuel economy that is higher than the average requirement for trucks, the plaintiffs claim that for every minivan sale, they are able to sell another truck that is less fuel efficient. Hence, lost minivan sales result in lost CAFE benefits. The plaintiffs also argue that brand loyalty evidence demonstrates that vehicle purchasers are more likely to buy a second vehicle from the same manufacturer than from another manufacturer. They argue that lost minivan sales will result in future lost sales of other vehicles. The plaintiffs claim that ITC should have considered these factors.
The statute clearly provides that “the effect of * * * dumped imports shall be assessed in relation to the United States production of a like [702]*702product. ” 19 U.S.C. § 1677(4)(D) (emphasis added). In this case, the like product consists of minivans; lost sales of other vehicles are not to be considered. ITC’s determination on this point was proper.
C. Competition Between U.S. Minivans and LTFV Imports:
The plurality found that price competition in the minivan industry was limited due to several factors. It reasoned that because price competition was limited, price was less likely to be a determining factor in purchasing decisions. This fact, in conjunction with the size of the dumping margin and limited import market share, led the plurality to a negative determination. For similar reasons, Commissioner Rohr also reached a negative determination.
The plaintiffs claim that the majority’s findings were in error. They argue that the only trustworthy evidence in the record indicates significant competition between U.S. and Japanese minivans. In addition, they claim that the majority considered certain unreliable data and disregarded other data that demonstrates demand for LTFV imports was price sensitive. The court will first address the plaintiffs’ arguments concerning evidence that should and should not have been considered, and will then consider whether the determination is based on substantial evidence.
1. Interim Data:
Interim data showed sticker price increases for 1992 Toyota and Mazda minivans, a relative decline in retail sales for these minivans during the first five months of 1992, and an increase in U.S. retail sales for the same period. See Conf. Doc. 461, INV-P-104. The plaintiffs assert that the data prove the importance of price to market share.
The majority refused to place great weight on the interim information because it did not distinguish between Chrysler’s U.S. and Canadian minivans, and it was neither complete nor comparable to data collected during the period of investigation. In addition, the commissioners expressed reluctance to draw conclusions about a full year based on interim data, particularly in the minivan industry where quarterly sales could be influenced by many factors. The plurality also found that, even if it were to consider the data, there was insufficient evidence in the record to conclude that an increase in U.S. minivan sales was related to price increases in LTFV imports.
ITC did not act unreasonably. The questionnaires, which sought data to the end of the 1991 calendar year, were sent out in early 1992, with a return date of March 27, 1992. The parties were given an adequate opportunity to comment on the questionnaires in early 1992; although the plaintiffs filed comments at that time, they did not seek collection of data beyond 1991. It was not until their prehearing brief, filed May 14, 1992, just one week prior to the hearing date and five weeks prior to the Commission’s vote, that they first raised arguments concerning the 1992 data. Given the lateness of the plaintiffs’ allegations, ITC’s deci[703]*703sion not to conduct a supplemental investigation was reasonable.8 See Florex v. United States, 13 CIT 28, 37, 705 F. Supp. 582, 591 (1989) (not unreasonable for ITC to reject data submitted after due date for questionnaire responses, when data would have required supplemental mailing).
Of course, ITC bears the burden of collecting all data necessary to its investigation, and the fact that the plaintiffs did not make a specific and seasonable request is not dispositive. See USX Corp. v. United States, 11 CIT 82, 96, 655 F. Supp. 487, 499 (1987). Here, however, the majority found the information unreliable; on this basis, it had discretion as the trier of fact to discount it. Wieland Werke, AG v. United States, 13 CIT 561, 576, 718 F. Supp. 50, 61-62 (1989); see also Iwatsu Elec. Co. v. United States, 15 CIT 44, 56,7 58 F. Supp. 1506, 1517 (1991). Its decision was not unreasonable. The data consisted of retail sales information only; the absence of evidence concerning other factors that might have influenced minivan sales rendered the data incomplete. In other respects, the data were not comparable to those covering the period of investigation, in part because retail prices were used whereas the questionnaire responses contained wholesale price information. In addition, the data did not distinguish between Chrysler’s Canadian and U.S. minivans, and Canadian shipments had a demonstrated effect on the data throughout the investigation. Finally, the majority permissibly discounted the information because it covered less than an annual period, and sales were subject to short term fluctuations. See, e.g., British Steel Corp. v. United States, 8 CIT 86, 93-94, 593 F. Supp. 405, 410-11 (1984) (rejecting argument that ITC is obligated to consider quarterly analysis of most recent data).
The plaintiffs argue the merits of the 1992 data at some length. These arguments need not be addressed as the court upholds the decision to discount the data. In any event, the court notes that the plurality correctly concluded no evidence linked the claimed increases in domestic sales to import pricing, and that the gains could easily have been due to other factors.
2. Toyota Data:
The majority examined “cross-shopping” and “second-choice” data; these data indicate the various minivans that purchasers inspected before making their selections and the minivans they would have purchased had their first choices not been available. The plurality considered all these data, but gave greater weight to Toyota’s Consolidated Dynamic Study (“CDS”), which consisted of second-choice information. Several reasons were given for placing greater reliance on the CDS information: it did not confuse Chrysler’s U.S. and Canadian minivans; it included complete data, to which ITC typically gives greater weight; and [704]*704the plaintiffs indicated they did not object to it. Although Commissioner Rohr did not cite the Toyota data specifically, he states that he considered all the second-choice data. The plaintiffs challenge the Toyota data on several grounds.
First, the plaintiffs argue that the data were submitted in Toyota’s posthearing brief, and they had no opportunity to respond. ITC counters that the plaintiffs had adequate response time, and are estopped from challenging the data. The court agrees with ITC.
In both the preliminary and final investigations, ITC requested surveys or studies regarding consumer preferences for minivans. ITC requested complete copies of the survey information. Following the May 21,1992 hearing, one commissioner requested cross-shopping data that distinguished between U.S. and Canadian minivans. The plaintiffs and Mazda were unable to comply as their surveys did not permit such a distinction. Toyota, on the other hand, attempted to eliminate the Canadian information from its tables. The resulting CDS data, which was filed with Toyota’s posthearing brief on May 29, 1992, showed that only modest percentages of import buyers listed a domestic vehicle as their second choice.9
On June 5,1992, the plaintiffs filed a letter responding to certain unrelated allegations made by Mazda and Toyota on May 28,1992.10 Conf. Doc. 40. At that time, the plaintiffs did not raise any arguments concerning the CDS data. On June 19, 1992, three weeks after posthearing briefs were filed, the plaintiffs attempted to file a letter, ostensibly regarding the CDS data; the comments were returned as untimely.11
Certainly, there is some asymmetry in that the plaintiffs were not given a specific opportunity to respond to information requested at the hearing and submitted in a posthearing brief. See Suramerica de Aleaciones Laminadas v. United States, 14 CIT 366, 370 (1990). There is also some danger that information accepted by ITC and not subject to rebuttal may be erroneous. Id. Nevertheless, as the plaintiffs have pointed out, material injury investigations are not adversarial in a formal sense, and it is ultimately ITC’s responsibility to evaluate the data it gathers. Conf. Doc. 40, at 2. Moreover, it is ITC’s prerogative to set and enforce time limits for submission of data. See, e.g., Avesta AB v. United States, 12 CIT 493, 510-11, 689 F. Supp. 1173, 1188 (1988). The notice of investigation provided that written submissions were to be filed by May 29, 1992; no provision was made for rebuttals. Minivans from Japan, 57 Fed. Reg. 2785, 2786 (ITC 1992) (institution and scheduling of final antidumping investigation); see 19 C.F.R. § 207.24 (1992). While [705]*705ITC did accept a submission from the plaintiffs one week after the deadline, it did not act unreasonably in refusing a submission filed three weeks after the deadline and just three working days before the Commission’s vote, especially since the plaintiffs’ challenge is not so extensive as to justify the delay. Accordingly, the plaintiffs’ first argument that they had no opportunity to respond to the submission is without merit.
The plaintiffs’ second argument is that the data are misleading and unreliable. Specifically, the plaintiffs claim that the data depend on the impossible premise that consumers can distinguish with precision between Chrysler models manufactured in Canada (Dodge Caravan/ Plymouth Voyager) and those manufactured in the United States (Dodge Grand Caravan/Plymouth Grand Voyager). They argue that it is unreasonable to expect consumers to recall and report vehicle designations accurately. The plaintiffs also argue that the figures reported in the CDS data are inherently incredible, because they indicate that the Dodge Grand Caravan and Plymouth Grand Voyager, which account for a significant market share, do not compete in any material way with either domestic or imported minivans.
To a degree, the plaintiffs’ arguments have merit. Certainly, all consumers may not have been able to recall the model names of Chrysler vehicles. Nevertheless, the majority did not err in relying on the data. The choice was between two sets of imperfect data, one set that included Canadian imports, the other that excluded them, perhaps imperfectly. Moreover, the data do not seem absurd on their face. The plaintiffs’ arguments are relevant only to the Chrysler vehicles. Yet, the figures show that purchasers of Japanese minivans express a similar low level of second choice preference for other domestic minivans, including GM vans, which also hold significant market share. The plaintiffs now argue that ITC would have reached a more reasonable conclusion by eliminating Canadian vehicles based on relative sales. The plaintiffs did not suggest this methodology to ITC, and cannot now complain that ITC did not adopt it.
The plaintiffs make two additional arguments. They claim that the plurality’s remaining justifications for relying on the Toyota data — that only Toyota submitted complete data, and that the plaintiffs conceded ITC could place greater weight on the data — are incorrect. As to the first point, it is the plaintiffs who are incorrect. Toyota did submit complete raw survey data, Conf. Doc. 45 (M-P), whereas the plaintiffs, by their own admission, submitted only summaries. Conf. Doc. 40, 4-5.12 As to the second point, it appears that the plaintiffs’ statement that they would not object if ITC relied more heavily on the Toyota data (which essentially contained the same cross-shopping figures as the GM data), [706]*706was a reference to the Toyota data prior to exclusion of Canadian imports. So, in fact, the plaintiffs made no concessions concerning the CDS data. Even without this justification, however, the reasons for relying on the CDS data were sufficient. Finally, the majority did not disregard other data in the record, but emphasized that it considered all the survey data. There was no error. The court will now consider whether the determination was based on substantial evidence.
3. Price Effects:
In evaluating the effect of imports on domestic prices, ITC considers whether “there has been significant price underselling by the imported merchandise,” and whether “the effect of 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).
Here, the plurality found no evidence of price depression because prices were rising. It also found no evidence of significant underselling, primarily because it believed the data to be flawed and unreliable. In terms of price suppression, it noted several relevant factors, such as the degree of substitutability between minivans, the availability of substitute vehicles (fairly traded imports, resale minivans, and non-minivan substitutes), the size of the dumping margin, and the market share held by imports. Final Det., at 24.
It found that the presence of substitutes for domestic and Japanese minivans limits price suppression. It found that the size of the dumping margin also limits the impact of dumping on domestic sales, and that given the size of the margin, demand would not have increased significantly even if the subject imports had been fairly traded.13 It noted that Japanese imports had only a limited market share, and thus any decrease in import market share would not have a significant impact on domestic sales. Id. at 29.
On the question of substitutability, the plurality, relying on basic economic principles, noted that the more fungible the product, the more likely that purchasers will make decisions based on price differences. On the other hand, when products are highly differentiated, price is less likely to determine product selection. It found that minivans are highly differentiated, and numerous factors limit their substitutability, including vehicle features, perceptions of quality and reliability, price differences, and brand loyalty considerations. Id. at 24-27. Commissioner Rohr also found that competition in the minivan industry was more a matter of features than price. Id. at 51. These findings will be discussed in turn.
[707]*707(a) Vehicle Features:
The plurality found that minivans are differentiated by many features, including interior space, seating configuration, cargo capacity, handling, engine size, front- or rear-wheel drive, two- or four-wheel drive, length or wheelbase, available options, and safety features.14
As a general matter, the plaintiffs argue that the feature differences identified by the plurality do not limit competition between U.S. and Japanese minivans to any significant degree. They concede that these differences exist and influence purchasing decisions, but argue that competition takes place across differences. They claim that for each of the two Japanese models, one or more U.S. made minivans is a closer match in terms of features and characteristics than the other Japanese model. The plaintiffs misconstrue the plurality’s reasoning in a fundamental way. The plurality did not draw systematic distinctions between U.S. and Japanese minivans. It found that there are differences between all minivans in terms of features and characteristics, which limit their interchangeability. As a result, minivans compete based on nonprice factors; to a much lesser extent, they compete based on price or deal offered. This particular finding, which was central to the determination, is also supported by survey evidence.15 The evidence indicates that most minivan purchasers considered as many as five other factors before price; for purchasers of Japanese minivans, price was an even less important consideration.16 Conf. Doc. 43A, at 6-7.
In addition to their broad objection, the plaintiffs challenge specific findings concerning wheelbase and engine size. On the subject of wheelbase, the plurality stated:
All shipments of Japanese minivans were of standard length while U.S. minivans were more evenly split between standard length and extended length wheel bases.
Final Det., at 25. The plaintiffs claim that the plurality misinterpreted the data. They point out that only Chrysler makes a minivan with two different wheelbases, but GM and Ford sell their vehicles in regular and extended lengths. In addition, the plaintiffs note that Japanese minivans are not of “ standard” length; the Previa is a relatively long minivan with significant cargo capacity, whereas the MPV is a much smaller vehicle in terms of length and cargo capacity. According to the plaintiffs, the Previa is comparable to Chrysler’s U.S.-built long wheelbase vehi[708]*708cles or the extended length Ford Aerostar, whereas the MPV is comparable to the regular length domestic minivans or the Canadian-built Chrysler minivans. Hence, argue the plaintiffs, even if the plurality intended “wheelbase” as a surrogate for “size,” neither length nor cargo capacity distinguishes U.S. from Japanese minivans.
This argument is without merit. During the investigation, ITC used two minivan categories: “Type wheelbase: standard length,” and “Type wheelbase: extended length.” See, e.g., Conf. Doc. 461, at B-72-B-73. All long versions — both extended length and long wheelbase — were placed under the second category. Hence, ITC’s designation of “extended length wheel base” refers to both long wheelbase and extended length minivans. Moreover, the fact that no systematic distinctions exist between U.S. and Japanese minivans based on length and cargo capacity is not relevant. The mere availability of an extended length or long wheelbase vehicle, combined with other differences, is a valid basis for product differentiation, which limits substitutability.17
As for engine size, the plurality noted:
Many Japanese minivans are sold with only 4-cylinder engines [referring to the Previa],18 while all U.S. minivans sold in 1991 had the larger engine size. The smaller engine size has been cited as a drawback to some of the Japanese vehicles.
Final Det., at 25. For this proposition, the plurality relied on an ITC memorandum, which in turn cited an article in the trade press. In a road test comparison of six minivans, the article described as a “drawback” Previa’s “2.4 liter, 16-valve inline four cylinder engine,” and noted that its “138 horsepower was noticeably overmatched in this field of V-6s.” Rik Paul, The Great Mini-Van Comparison, Motor Trend, at 94 (May 1992), Pub. Doc. 337MM (56). The article then goes on to compliment the Previa’s performance, in spite of its “valiant but overmatched” 2.4 liter engine. The plaintiffs claim that the critical issue is not cylinder count, but overall engine performance including horsepower and torque ratings. When these factors are considered, according to the plaintiffs, the trade press has compared the Previa favorably to six-cylinder vehicles. Here, the plaintiffs attempt to prove too much. The plurality stated only that the Previa’s four-cylinder engine has been cited as a drawback; substantial evidence supports this proposition. That this report, or others, praise the Previa after pointing out that it has a four-cylinder engine does not undermine the point that the trade press perceives a difference based on cylinder count, which is relayed to consumers. After [709]*709all, product differentiation may be based on actual or perceived differences.19
Other than wheelbase and engine size, the plaintiffs do not challenge the plurality’s specific findings on vehicle features. The findings are sound. Although all minivans share certain very basic characteristics, there are many differences that fragment the market in various ways.20
(b) Perceptions of Quality and Reliability:
The plurality also found that perceptions of quality and reliability are important nonprice considerations influencing vehicle selection. Final Det., at-26. Indeed, the staff report states that the quality of products and dealership service are critical elements of the auto industry. Conf. Doc. 461, at A-26. Survey evidence also indicates that quality and reliability are among the most important considerations for buyers. Conf. Doc. 43A, at 6.
The plurality found that in general Japanese minivans are viewed as higher quality than domestic minivans. Final Det., at 26. Substantial evidence in the form of warranty data and quality survey data supports this finding.21 Conf. Doc. 461, at A-26-A-31. The plaintiffs do not put up a serious fight on this front, except to point to some less persuasive evidence in the record which they claim the plurality ignored.
(c) Price Differential:
The plurality found a significant price differential between domestic and imported minivans, which also limits substitutability. Final Det., at 27. It found that Japanese minivans were generally sold at higher prices, and to a greater extent in higher price ranges than the domestic vehicles.22 Id. It found that this factor limits substitutability to the degree that purchasers are constrained by income considerations, but noted that price was a less important criteria for buyers of Japanese minivans than for domestic purchasers. Id. at 27-28. Substantial evidence, including the plaintiffs’ own concessions, supports the plurality’s findings. Conf. Doc. 461, at A-151; Conf. Doc. 23, at 29; Conf. Doc. 43A, at 6-7. The evidence shows that Japanese minivans compete in the higher price ranges, and that price is less important to purchasers of these vehi[710]*710cles.23 Again, this evidence indicates that the market is fragmented into smaller markets. Finally, the plaintiffs’ argument that reliance on average prices distorts the data on the price differential is without merit. ITC is not prohibited from relying on averages. See generally, Copperweld Corp. v. United States, 12 CIT 148, 162, 682 F. Supp. 552, 566 (1988) (neither statute nor legislative history requires ITC to adopt any particular analysis when market consists of different segments).
(d) Brand Loyalty:
The plurality also found that brand loyalty is an important nonprice factor that affects consumer preferences and limits substitutability. Final Det., at 26. That is, the first purchase of any type of vehicle tends to dictate future vehicle selection. The cross-shopping and second-choice data measure brand loyalty to an extent. The plurality found these data indicate that only a minority of minivan buyers “shop” both a U.S. and Japanese minivan. Id. at 27. Commissioner Rohr also found limited cross-shopping. Id. at 50.
The significance of the cross-shopping and second-choice data is disputed, and each side claims victory based upon it. The majority considered all the data. The cross-shopping data support the finding that relatively few purchasers consider purchase of both domestic and import brands. Focusing on what are agreed to be the more relevant second-choice data, that is, the percentage of Japanese-import purchasers listing a domestic vehicle as their second choice, and relying on the Toyota CDS data, which ITC reasonably found more reliable, the evidence also indicates that a relatively small percentage of the market would even consider purchase of a domestic vehicle if the import were not available.
That is not the end of the inquiry, however, because the plaintiffs maintain that the data demonstrate significant competition between domestic and Japanese minivans. The data do indeed indicate that competition occurs between U.S. and Japanese minivans, although the court would not characterize it as significant. The majority did not find otherwise. The data also do not indicate the basis of any competition that does exist. That is, if competition takes place primarily based on nonprice factors, the majority’s findings concerning price suppression are correct. Moreover, as the defendant and the defendant-intervenors point out, the second-choice data are not to be equated with lost sale information. The data indicate only the possibility that a purchaser would have settled for another vehicle, had his or her first choice not been available.
(e) Underselling:
On the issue of underselling, which requires more specific information than assessments of price differentials for different vehicle categories, the plurality refused to place great weight on the pricing data because selection of comparable vehicles was difficult. In addition, it re[711]*711jected the validity of certain proposed price comparisons, such as between “regular length” Japanese minivans and “extended-length” U.S. vehicles. The plaintiffs concede that price comparisons are difficult; however, they restate their previous objection to the characterization of Japanese minivans as “regular length.” This issue has already been discussed. Suffice it to say that differences between minivans, including wheelbase or length, made specific price comparisons unreliable. The plurality was correct in regarding such comparisons with skepticism.
4. Volume Effects:
The plaintiffs acknowledge that the majority’s substitutability finding was also the basis of its conclusions concerning the effect of import volume. They argue, however, that the analysis is flawed on independent grounds.
On the issue of volume, ITC is to consider whether volume or increase in volume is significant, either in absolute terms or relative to U.S. production or consumption. 19 U.S.C. § 1677(7)(C)(i). The plurality found that it was not significant. It noted that increased volume was due to introduction of the Previa, which, in large part, created its own demand.24 It noted that throughout the period of investigation the market share of LTFV imports in terms of quantity and value was less than fifteen percent. It also stated that its finding concerning the significance of volume increases was made in view of nonprice factors, such as the level of sub-stitutability. Commissioner Rohr’s findings concerning volume were based on a similar analysis.
The plaintiffs argue that the increased volume of LTFV imports had a significant impact because the minivan industry is capital intensive, and small changes in sales volume lead to steep profit losses. In addition, they argue that sales of LTFV imports were at the high end of the market, which is more profitable to U.S. producers. There is no authority for the proposition that these factors, above all others, should dictate a result. ITC has discretion to consider import volume in light of the “conditions of trade, competition, and development regarding the industry concerned.” S. Rep. No. 249, 96th Cong., 1st Sess. 88. As this court has recognized, the significance of import volume may depend on whether the product is fungible and price sensitive, or whether the market is highly differentiated. See Negev Phosphates, Ltd. v. U.S. Dep’t of Commerce, 12 CIT 1074, 1084, 699 F. Supp. 938, 947 (1988) (import volume alone cannot gauge effect of imports because if product is price sensitive, underselling of even small volumes may cause price suppression); H.R. Rep. 317, 96th Cong., 1st Sess. 46 (1979) (small volume of imports may or may not have significant impact, depending on industry). Here, the [712]*712maj ority’s finding on import volume was dependent on its finding of limited substitutability, and the plaintiffs have not identified any independent error. Its reasoning was legally sound.
D. Other Arguments:
The plaintiffs argue that the plurality failed to state the standard of causation and weighed various causes. They claim there is evidence an incorrect standard was applied. This argument is without merit. The plurality’s analysis clearly tracks the statutory factors, and there is no indication that the plurality weighed causes.
The plaintiffs also argue that the plurality attributed to other factors injury that was actually caused by the LTFV imports, and overlooked the fact that U.S. producer fleet sales, which are less profitable, were required because of competition from LTFV imports. These arguments are legally insufficient. ITC found no nexus between LTFV imports and injury to the domestic industry. It did not err in noting other possible causes of injury, S. Rep. No. 249, 96th Cong., 1st Sess., at 75; nor was there sufficient evidence to indicate that fleet sales were necessitated by the subject imports.
IV Conclusion
The plaintiffs have failed to demonstrate legal error in ITC’s determination. In addition, the negative determination was supported by substantial evidence, in that the volume and market share of the subject imports were low, price depression was absent, and no price suppression or underselling was found, based, in part, on low substitutability. In addition, limited substitutability also led to a finding that the volume and market share which did exist were not significant. The vehicle largely responsible for increased import market share, the Previa, created a good part of its own demand. Purchasers of Japanese minivans did not usually consider domestic minivans, and prospective purchasers not satisfied with Japanese minivans had other imports from which to choose, as well as nonsubject merchandise, such as station wagons. Thus, the majority’s findings, that competition between Japanese and domestic minivans was limited and impact on the domestic industry was insignificant, are supported. The factors to which the plaintiffs point as demonstrating injury do not amount to evidence of a significant contribution by the subject merchandise to material injury, and are insufficient to undermine the substantial evidence on which the ITC majority relied. The determination is sustained.