This is a judicial review of a determination 1 in which the U.S. International Trade Commission (ITC) found that an industry in the United States would be threatened with material injury if an antidumping duty order issued against television sets from Japan 2 were to be modified or revoked. The ITC had to first find that circumstances had changed sufficiently since the issuance of the antidumping duty order to justify the type of administrative review provided in section 751(b) of the Tariff Act of 1930 (19 U.S.C. § 1675(b)).3 The result of the ITC’s review was then challenged by the proponents of revocation4 in an action brought under 19 U.S.C. § 1516a(a)(2)(B)(iii), which action requires the Court to hold unlawful *856any determination found to be unsupported by substantial evidence, or otherwise not in accordance with the law. 19 U.S.C. § 1516a(b)(l)(B).
The Court has examined the administrative record, studied the briefs of the parties and listened to oral argument. As a result, the Court concludes that in one essential respect the determination was not based on substantial evidence. There was no substantial evidence to support the conclusion that the level of importations from Japan would be injurious if the antidumping order were to be revoked.
I
The Court begins with a holding that, if the ITC is the only agency from which such a review is sought, it has to presume, as it did, that such imports as will be made will be at less than fair value. This may produce some anomalies5 but it is the only result which is consistent with the fact that the International Trade Administration of the Department of Commerce (ITA) has the power to conduct' a “changed circumstances” review of its original determination of sales at less than fair value.6 What the ITA thought about the likelihood of future sales at less than fair value, was unknown to the ITC and that unknown factor must operate as a presumption in the ITC’s review.
The ITA’s review powers would be pointless if the question of whether there will be sales at less than fair value could be reviewed and resolved by the ITC alone. In order to give meaning to the separation of function which runs throughout the law, the ITC must presume that future sales will be at less than fair value if the ITA has not reviewed the matter.7
In sum then, the choices of a party seeking a “changed circumstances” review under section 751(b) are two. It can seek review by the ITA on the question of sales at less than fair value and, if the ITA still finds the presence or likelihood of sales at *857less than fair value, it can move on to the ITC and seek a conclusion that, nevertheless, material injury will not result. As an alternative, a party can forego review by the ITA and go directly to the ITC. The latter strategy must result in the unalterable presumption in. the ITC review that such sales as there will be, will be at less than fair value.
II
The presumption that future sales will be at less than fair value represents the limit of the presumptions which the ITC can make in this -review. As will be explained later, in all other respects the review is a neutral investigation of the likelihood of the recurrence of injury. The ITC has stated its task to be an investigation of whether an industry in the United States would be materially injured or threatened with material injury by reason of imports of the merchandise covered by the antidumping order if the order were to be modified or revoked.8
The ITC determination of this question consisted of two subsidiary findings, first, that the U.S. industry was still in a delicate state of health and second, that the levels of imports from Japan, after the anti-dumping duty order was revoked, would cause injury. The finding as to the health of the U.S. industry was supported by substantial evidence in the form of evidence that it was highly competitive, price sensitive and showing low profitability. Although there were other, possibly more significant, signs of recovery in the industry and, although competition could be seen as an indicator of health,9 these are matters of weighing and interpreting economic evidence on which the Court will defer to agency expertise.
On the other hand, the subsidiary finding that import levels would be injurious in the absence of the antidumping duty order was not supported by substantial evidence even in the loosest sense. This finding took the ^orm of a prediction that import levels from Japan would increase significantly. The finding about future import levels gave three supporting reasons, first, the present intentions of the Japanese, second, their production capacity, and third, the incentives or motivations which would govern their future behavior.
' [3] The judgment of present intentions is a proper, and possibly controlling element of a review by the ITC. City Lumber Co. v. United States, 59 CCPA 89, 96, C.A.D. 1945, 457 F.2d 991 (1972). Here, however, it was not based on any positive evidence tending to show an intention to increase the levels of importation. The judgment that import levels would intentionally increase was expressed by statements of the ITC that there was no basis to believe otherwise;10 that the ITC was not convinced otherwise by the *858statements of counsel;11 that there was no direct testimony by counsel or corporate executives;12 that there was an absence of credible submissions.13 All this demonstrates that, despite its protestations that it was not placing a burden of proof on the proponents of revocation,14 the ITC relied on an asserted failure of the Japanese interests to introduce evidence, combined with a decision that the evidence which was submitted lacked credibility.
In this case, on the subject of intention, this approach indicates that a presumption was made that the intention to increase imports (which was temporarily restrained by the existence of an antidumping duty order) remained undiminished.
Such a presumption is not permitted by the law, which envisages a review conducted by the ITC, and gives no inkling that in marked contrast to all the other phases of an antidumping investigation, here for the first time a burden is placed on a party. If a party can be said to have a burden of proof in these matters it is the burden of proving to the agency that circumstances have changed sufficiently to justify a review. In the review itself, however, the only burden on a party is the burden of cooperating with the investigation of the administrative agency. The credibility of testimony is important only when a party has a burden of proof, or when there is conflicting testimony as to facts. If, after all is said and done, the ITC has nothing more than its dissatisfaction with a party’s evidence on which to base a determination, it is not relying on substantial evidence.
Ill
To a large extent, the difficulties in this administrative determination had their origin in a misconception of the nature and purpose of the administrative review.
The statutory language gives no details about the conduct of these administrative reviews.15 The legislative history is unrevealing.16 The ITC has stated that its review should be a determination of whether the industry would be threatened with material injury from imports of the product covered by the order if the antidumping duty order were to be modified or revoked. In the review, the ITC saw itself in the position of having to honor the basic protective intent of the antidumping law; that is to say, it believed it had to treat the continuing validity of.the existing antidumping *859order as a premise of the review.17 This led it to go beyond its neutrally stated objective and to engage in a presumption that injury would recur and to place a burden on the proponents of revocation to prove otherwise. However, although the purpose of antidumping duties is unquestionably to protect American industry, the provision for review of final injury determinations cannot be given the same protective motive.
In the opinion of the Court, in order for the review provision to operate consistently within the structure of this law and in order for it to function in harmony with one of the international agreements which the law was intended to implement, the review must establish the continuing need for the injury determination. If it cannot do so, the original determination has become vestigial.
This conclusion is necessitated by the statutory prerequisite to administrative review, namely, that circumstances have changed since the issuance of the antidumping order. A preliminary determination that circumstances have changed means, in this case, that there is some evidence to support the view that the original injury determination is no longer valid. This is the fundamental justification for a review of the injury determination. If the ITC had found that there was not a sufficient change of circumstances to justify a review, that decision would have been judicially reviewable under 19 U.S.C. § 1516a(a)(l)(A)(ii).18 On the other hand, the finding of changed circumstances sets in motion the review process. This can only mean that a new investigative finding is needed to maintain the validity of the injury determination in the new circumstances. If the threat of injury cannot be found, or if a finding that there is a threat is not supported by substantial evidence, the unavoidable inference from the evidence of changed circumstances is that the threat of injury has dissipated.
Another important factor is the necessity and desirability whenever possible, of harmonizing this law with the international agreements it was intended to implement. Section 2(a) of the Trade Agreements Act of 1979 (19 U.S.C. § 2503(a)) expressly approved the Agreement on Implementation of Article' VI of the General Agreement on Tariffs and Trade (GATT) which relates to antidumping measures, and is referred to as the International Antidumping Code. Article 9(a) of the Code relates to the review of antidumping duty orders and provides that “an anti-dumping duty shall remain in force only as long as, and to the extent necessary to counteract dumping which is causing injury.” The injury referred to includes the threat of material injury.19 It would be *860unreasonable to take or maintain the potent measures of these laws for anything more diffuse than the threat of material injury. It must follow that when the continued necessity for the antidumping duty is placed in question by a change in circumstances, the review required by section 751(b) must either find reason for continuation of the duty or lead to revocation.
The formalization of the review process, the passage of time,20 the change of circumstances, the international commitment not to leave antidumping orders in place longer than necessary, the elaboration of the related revocation power of the ITA21 all suggest that in this review the continuation of an antidumping order depends on a finding that injury is still a threat. As noted, it is also possible to reason from the lack of an explicit burden of proof on the proponents of revocation and from harmony with the earlier stages of these investigations in which there is no burden of proof on a party. From all this, it can be concluded that the ITC has the duty to investigate whether there is a threat of injury and if it cannot find one, it must recommend revocation of the order.
The Court is satisfied that this does not make a foregone conclusion of the review process and does not present the ITC with an impossible task. In keeping with normal principles of administrative law, the Courts sustain a determination which is supported by substantial evidence even if it does not represent the view of the evidence which the Court might have taken. Consolo v. Federal Maritime Commission, 383 U.S. 607, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966).
In this matter, however, it became apparent that the determination, although disarming in its candor, went beyond the limits of administrative fact-finding in several important respects. There is nothing to support the ITC’s subsidiary finding that the Japanese presently intend to increase imports if the antidumping duty order is revoked or modified. The ITC could not derive intent from a presumption that the antidumping order was presently restraining intent, or from a failure by the proponents of revocation to offer evidence of a lack of intent, or solely from a disbelief in the credibility of the evidence offered.
IV
The ITC’s concentration on intent makes it difficult to isolate the grounding of the ITC determination in objective economic factors. To a certain extent even the later conclusions that develop those factors are clouded by the element of assumed intention and the vestiges of a presumed increase in import quantities. Nevertheless, the evidence regarding the objective factors of production capacity and economic incentives or motivation has been analyzed by the Court and has been found wanting.
Evidence of production capacity alone does not support the ITC’s conclusion that imports would increase. First, it is not stated by the ITC as a self-sufficient basis for finding that import levels would increase and it is questionable whether raw capacity could provide such a basis, without being reconciled with the tremendous increase of the production capacity of related Japanese-owned companies in the United States.
In this regard, the development of a coherent finding was probably impeded by the undisputed fact that the Japanese had established significant production facilities in the U.S. with which they were supplying approximately twenty-five percent of U.S. *861demand at the time of the ITC review determination. A determination that the Japanese would import at increased levels would have had to confront the question of the apparent irrationality of such behavior vis a vis their American production facilities. The ITC did not see sustained or prolonged increases in importations. It saw only short-term supplemental penetrations, which, as will be explained, even if correctly foreseen, would not cause material injury.
Second, even in the combination of capacity with other economic factors the ITC did not have substantial evidence for concluding that import levels would be injurious. This emerges from a close examination of the examples of injury predicted, which will be referred to as the scenarios. The first scenario,22 combines increased demand in the U.S. market (for which the evidence is definitely substantial) with a virtual assumption that, given the production flexibility of the Japanese, importations of their products will increase.
The first scenario is not really a finding that import levels will be injurious, but a step beyond that to a prediction, which is probably a truism, that prices would be suppressed if the Japanese were to increase import levels at less than fair value. In effect, it assumes that import levels will increase and finds that prices will then be suppressed.
The second scenario,23 is somewhat more specific. It sees the Japanese using imports to supplement their American production when there are short-term, cyclical increases in demand, in order to secure the increased market share until the U.S. production can be expanded. Although there may have been substantial evidence for finding that there would be short-term increases in imports,24 the outcome of this scenario would not be injury but a form of assistance to the U.S. industry. If the inclusion in the U.S. industry of Japanese-owned production facilities is accepted, (and this is a premise of the ITC review), then any short-term import strategy designed to temporarily occupy a market segment for the benefit of related American subsidiaries, would be a benefit for the industry.25 If it does *862harm, it does so to competitors within the U.S. industry, not to the industry as a whole.
The third scenario has two variants, both of which are conjectural in origin and development. The first variant26 begins with diffuse predictions of surplus production of Japanese picture tubes leading to the use of those tubes to increase Japanese television production and then to increased exportations to the attractive U.S. market. The second variant27 begins with the unsupported projection of restraints on the importation of Japanese televisions or tubes in various important foreign markets and moves directly to the shift of the resulting excess production to the U.S. market.
The Court agrees that this review can be predictive in nature, looking beyond the revocation of the antidumping duty order.28 In this sense, it may have to be somewhat more predictive than the determination in the original investigation that there is a threat of injury. This may be necessary because in this review the antidumping order, operating as a strong corrective or deterrent, can be presumed to distort the meaningfulness of observable data regarding present conduct in the U.S. market.
However, it does not follow that these circumstances justify the use of something less than substantial evidence as the starting point for agency reasoning.29 The nature of the factual predicate in a review may vary. It may have to consist of facts derived from outside the U.S. market or outside the producers’ home market. It may have to involve facts regarding important component parts of the products in question. It may have to include other measurable economic factors upon which it is reasonable to base inferences. But, in all instances, the factual predicate, i.e., the base on which the inferences are made must have a present, verifiable substance to it. This is lacking in the ITC findings. There was no real evidence of restraints in any other markets on the sale of Japanese television sets or tubes. As a result, the diversion of those sets to the U.S. market was only a speculation. The future occurrence of excess tube production in Japan was, in *863itself, more a prediction than a finding of fact. It then required a further prediction, that excess tube production would force excess television production and would then require increased exportation to the U.S. This was a chain of predictions set on an insubstantial base and was entirely conjectural. It is of crucial importance to the fair administration of this law that these determinations, even in their most predictive mode, must detect real and imminent events and must have a basis in substantial evidence.
V
The Court has found no support in the case law for the reworking of the substantial evidence standard for which the ITC is, in effect, asking. It has not been persuaded that the ITC review determination involved matters of such an inherently indeterminate and inchoate nature that it is impossible to base decisions on conventional evidence.
The decision in Industrial Union Department v. American Petrol. Inst., 448 U.S. 607, 100 S.Ct. 2844, 65 L.Ed.2d 1010 (1979) is by far the most instructive of the many cases cited by the parties. Its lesson is all the more important because the agency in that case was operating in a relatively unknown area to an even greater extent than the ITC. In that case, the Supreme Court affirmed the holding of the Court of Appeals that the Secretary of Labor lacked substantial evidence to lower from ten parts per million (10 ppm) to 1 ppm the maximum worker exposure to the cancer-causing chemical, benzene. The new standard was based on assumptions that exposure at levels below 10 ppm caused cancer and that the risk of cancer would decrease as exposure levels decreased. The Supreme Court found that the burden was on the Secretary to show on the basis of substantial evidence that existing standards presented a significant risk of material health impairment and further held that this was not a disabling burden. It remained within the agency’s responsibility to determine what it considers significant,30 it could have some leeway in making that determination on the frontiers of scientific knowledge, and could even take the more conservative position. The Supreme Court further stated,
Thus, so long as they are supported by a body of reputable scientific thought, the Agency is free to use conservative assumptions in interpreting the data with respect to carcinogens, risking error on the side of overprotection rather than underprotection. 448 U.S. at 656, 100 S.Ct. at 2871.
The aforementioned case shows both the weakness of the ITC approach and, at the same time, offers encouragement to proper determinations.
The same situation is seen in a more competitive context in Boise Cascade Corp. v. Federal Trade Commission, 637 F.2d 573 (9th Cir.1980). In that case, the Court of Appeals found that the Federal Trade Commission (FTC) lacked substantial evidence to conclude that a system of delivered pricing (using West Coast freight rates) used by certain Southern plywood mills, had an anti-competitive effect in violation of section 5 of the Federal Trade Commission Act, 15 U.S.C. § 45. The Court found that the FTC’s conclusion was “little more than a theory of the likely effect of the challenged pricing practices,” (637 F.2d at 578) and was “largely a deduction from the Commission’s reasoning about the tendencies of the challenged practice.” (637 F.2d at 580). In addition, the Court held that the FTC’s approach had the effect of requiring the Southern mills to prove that the challenged practices had no anti-competitive effect.
In the interests of enforcement they may not take leave of normal practice, may not place burdens where they should not be placed, and may not use impermissible assumptions and suppositions rather than substantial evidence.
The only possible distinction in this case is that here the administrative agency *864is doing its work on its own previous determination. However, as has been pointed out, that earlier determination can be given no special standing in setting the burdens and responsibilities of the ITC. In all respects, this review has the character of a renewed investigation, in changed circumstances, into the existence of a threat of material injury.
A case such as Industrial Union Department, AFL-CIO v. Hodgson, 499 F.2d 467 (D.C.Cir.1974) does not help the ITC. In that case the decision made by the Secretary of Labor under the Occupational Safety and Health Act of 1970, 29 U.S.C. § 651 et seq. (OSHA) setting low limits for worker exposure to asbestos was upheld as rational and within his authority. The Court of Appeals frankly stated that it could not use the statutory substantial evidence standard for those of the Secretary’s findings which were on the frontiers of scientific knowledge, and which were, in reality, policy or quasi-legislative decisions which he was authorized to make. The decision-making process of the ITC does not have the same difficulties and it does not require the same freedom of action. In any event, the ITC does not have the same mandate for quasi-legislative action in its fact-finding investigations as the Secretary of Labor has in promulgating general rules and standards. The ITC cannot engage in the same “better to be safe than sorry” reasoning. Questions of whether there is a threat of injury which precludes revocation have been resolved without speculative reasoning or legislative creativity and there is no reason to think that the ITC cannot find substantial evidence in its area of expertise.
In sum, the Court finds that there was a lack of substantial evidence to support the ITC’s determination regarding a threat of material injury from imports of television receiving sets from Japan. A remand for further clarification or investigation is not called for in this situation. This is not a case in which the ITC failed to sufficiently articulate its rationale to allow judicial review as in SCM Corp. v. United States, 84 Cust.Ct. 227, C.R.D. 80-2, 487 F.Supp. 96 (1980); nor is it a case in which a corpus of correct data has to be investigated for the first time as in Sprague Electric Co. v. United States, 84 Cust.Ct. 260, C.R.D. 80-6 (1980) modifying on reh. 84 Cust.Ct. 243, C.R.D. 80-3, 488 F.Supp. 910 (1980).
Here, the unavoidable conclusion from the lack of substantial evidence of a threat of injury is that, in the light of changed circumstances, there is no threat of injury. Accordingly, it is hereby
ORDERED that the determination of the ITC is reversed and it is further
ORDERED that the ITC shall reach a new determination consistent with this opinion.