WILKEY, Circuit Judge:
The case comes before this court on a petition for review of Order 73-4-64 of the Civil Aeronautics Board of 13 April 1973, approving certain International Air Transport Association rate agreements for the North Atlantic, dismissing petitioners’ complaint against the fare structure, and declining to institute a requested investigation of North Atlantic rates at this time. The two individual petitioners represent themselves to be users of transatlantic airline services; and petitioner Aviation Consumer Action Project (with which the individual petitioners are connected) is a nonprofit consumer organization which has participated in CAB hearings.1 Highly conscious of the appropriate limits to the scope of our review, we have given the fullest deference to the administrative judgment of the Board. Despite that approach, we can detect no substantial evidence to support the Board’s major and most critical findings, and we are not satisified that the Board exercised its discretion in setting policy in accordance with the approach intended by Congress. Accordingly, the Board’s order must be vacated.
I. Actions of the Civil Aeronautics Board and Carriers
This case brings before us again the now familiar scenario of lengthy, tedious, and complex negotiation among the eighteen carriers over the North Atlantic to establish a unanimously agreed upon rate structure under the aegis of the International Air Transport As[1020]*1020sociation.2 This latest chapter begins with the last general transatlantic fare agreement, reached in early 1972 to last for one year only, approved by the CAB to be effective 1 April 1972.3 In approving this agreement, the CAB strongly indicated that the compromise fare structure left much to be desired, specifically that it was not adequately cost-related nor did it provide sufficient revenue, and administered the first (in this chapter) of its admonitions to all IATA carriers to do better next time.4
In mid-1’972 the eighteen IATA North Atlantic carriers dutifully entered upon negotiations for a new fare structure for the 1973 travel season. After several months of stalemate, “it became apparent that a consensus on revised fares for the 1973 season could not be reached through the IATA machinery, and intergovernmental consultations undertaken by the United States thereafter . similarly did not result in an accommodation for the upcoming summer season generally acceptable to all governments.”5 There is no intimation here or elsewhere in the record that United States Government representatives made any effort with any other individual government or foreign airline to achieve any bilateral rate agreement on any single route, although the international intergovernment airline framework is based on bilateral underlying agreements between the United States and each individual foreign country.6
As a result of the complete collapse of the IATA and the intergovernmental multilateral negotiations, major North Atlantic carriers filed individual tariffs with the CAB and foreign governments, to be effective upon expiration of the previous limited one-year IATA agreements on 1 April 1973. By Order 73-1-76 of 26 January 1973 the CAB approved the new fare structure as filed by U.S. carriers, noting that “it contained certain features which we consider significant improvements” and “most importantly, however, the U.S. carrier proposal was expected to provide some improvement in yield and would have moved the structure in the direction of more closely relating fares to the cost of providing the respective services.”7
The tariffs filed by nine foreign carriers reflected the sharp fundamental disagreement on scheduled airline rate policy which had stalemated the multilateral negotiations. The Board exercised its newly granted statutory power to suspend these foreign tariffs,8 find[1021]*1021ing that they bore little or no relationship to cost and would have a deleterious effect on carrier yield. Apparently, to every action there is an equal and opposite reaction, in international economics and politics as well as physics; hence the tariffs filed by Pan American, TWA, and National were protested and rejected by several European governments pursuant to their respective bilateral agreements with the United States. The Board states that “in subsequent consultations with those governments, we sought to no avail to gain their acceptance of the Pan American and TWA filings and, indeed, it proved impossible to reach an agreement on any mutually acceptable alternative.”9
The Board then concluded that insistence on the proposed Pan American and TWA fare structure could only result in “an intergovernmental confrontation which could lead to a cessation of air services.” Preferring not to risk this by either further insistence on the proposed U.S. carrier rates or compromise rates, the Board then approved the IATA carriers’ agreement to extend the status quo through 1973 “to avoid an open rate situation,”10 although the Board was “unable to conclude that extension of present fares for a further period of effectiveness constitutes a rational and economic basis for a provision of transatlantic service.”11
The Board justified this acceptance of a fare structure, which for two successive years it had condemned, by reasoning “that failure of the carriers to reach agreement within the IATA machinery on a more economic pattern of fares and the imminence of the peak travel season combine to make any solution other than the status quo wholly unrealistic.”12 [1022]*1022And, just as it had done in accepting the unsatisfactory fare structure for 1972, it accompanied its acceptance with a second exhortation to the carriers and other governments to “face up to the fact that this interim solution does nothing to simplify the fare structure or to produce the revenue improvement all consider necessary.”13
II. Applicable Statutory Standard
The statutory standard applicable to the CAE’s action herein is found in § 412(b) of the Federal Aviation Act of 1958,14 which requires “the Board shall by order disapprove any such contract or agreement . . . that it finds to be adverse to the public interest . . . and shall by order approve any such contract or agreement that it does not find to be adverse to the public interest.” Taking this as the appropriate standard, petitioners strenuously urge that the findings cited to support the Board’s basic conclusion that a continuation of the 1972 IATA fare structure is in “the public interest” are not supported by substantial evidence and that, to the extent the public interest determination represents an instance of expert agency exercise of discretion, discretion exercised without support is an arbitrary abuse of the discretion which Congress intended to be exercised.
We should note at the outset that our objections to the Board’s analysis do not stem from a narrow misconception that a temporary continuance of “uneconomic” rates could never fit within a valid broader vision of the “public interest”; we recognize that “the public interest” in international fares embraces more than the economic underpinning decisive in domestic fares, and we recognize that the Board made'an attempt to justify its position under this broader view. It is crystal clear to us that the Board’s central concern boiled down to a fear of an open rate situation, or as CAB counsel preferred to put it, the fear of the consequences of an open rate situation. The Board concluded that “[t]he fact remains, however, that failure of the carriers to reach agreement within the IATA machinery on a more economic pattern of fares and the imminence of the peak travel season combined to make any solution other than the status quo wholly unrealistic. . . . [I]t is imperative that firm prices now be approved so that firm travel plans can be made. This in our view is the paramount public interest- at this point in [1023]*1023time. . . . ”15 That vision of imminent peril, and the Board’s decision to favor the retention of IATA multilateral agreements over any other possible method of arriving at international fares, are the subjects of our review and disapproval.
III. Substantial Evidence
To the extent that the Board’s “public interest” determination involved factual predicates, the substantive statute requires that they be supported by substantial evidence.16 “Substantial evidence” is the requirement, no matter what the issue, and no matter how esoteric are the facts to be considered in support of the CAB decision on the issue. The Board cannot wrap its decision in some mystique of foreign policy or purported expertise in international negotiation to achieve a nonreviewable status for the facts underlying its most important and sensitive decisions. This, and the particular issue in the ease at bar, we anticipated in National Air Carrier Ass’n v. CAB (NACA I), where we cautioned the Board that “the specter of an open rate situation does not absolve the Board of its responsibility . . . ”17 Here the board conjures up this same specter, yet the evidence of record as to why it should quail at the thought of open rates remains essentially ectoplasmic.
As the Board recognized, there was absolutely nothing to support extension of the 1972 rates as a matter of economics. The CAB accepted, in two successive years, for periods of first twelve months and then for nine months, a rate structure which it has condemned in the strongest terms as being uneconomic, inequitable among the classes of fares, and eventually destructive of the financial capacity and stability of the United States airlines involved: Even if the 1972 rates had been viewed as satisfactory for the period, the economic assumption behind an acceptance of the 1972 fare structure for the year 1973 would necessarily have been that the year 1973 economically and business-wise will be and is a rerun of the year 1972. This may be true, as far as airline business is concerned, but we could not be confident of this assumption, without some justification or evidence to that effect on which the CAB could act. Yet the 1972 fare structure, plus six percent, was accepted for 1973 without any attempt to justify anything except the six percent increase in rates.
On the other hand, since the question of sufficiency of the evidence necessarily turns on sufficiency in relation to the issue and the findings on which the agency action was based, we should most appropriately look to what evidence supports the Board’s view of the consequences of an open rate situation. When we do so, we find hardly anything in the record bearing on that question at all. The record we have in this case consists of three Board orders of March 1972, January 1973, and April 1973; justifications of National, Pan Am, and TWA; comments in April 1973 of the National Air Carrier Assocation, which represents supplemental carriers; an extract from minutes of the London Currency Conference of March 1973; and the complaint of Aviation Consumer Action Project before the Board in the instant ease. The airline justifications deal with cost and traffic data, similar to what we would expect in a domestic rate proceeding. The Minutes of the Currency Conference reveal the divergent views of the various national airlines, and tend to support the undisputed conclusion that no multilateral agreement on a new fare structure was then possible in March 1973. The orders of the Board, particularly those of March 1972 and January 1973, are principally [1024]*1024devoted to analysis of cost and traffic data. Only in the order of 13 April 1973 does the Board touch upon the decisive issue on which the CAB action hinges — the consequences as of April 1973 of an open rate situation.
A. Harm from Open Rates
There is no evidence in the record to support the Board’s view that an open rate situation would lead to significant passenger inconvenience during the peak season. While the absence of an IATA agreement would most probably have led to differing rates, each carrier would have remained able to file firm rates, as indeed thay did earlier in the year. It must be conceded that the possibility of Board suspension of some tariffs, and the threat of retaliatory cessation of service, would have given rise to some uncertainty. But there is no visible basis for the Board’s dire view of the likelihood of such confrontation, not to mention any persuasive indication regarding how prolonged or unsettling some confrontation might be.
While the Board’s order darkly predicts “an intergovernmental confrontation which could lead to the cessation of air services,” there is no evidence in the record whatsoever to suggest the possibility of a complete breakdown of North Atlantic air traffic. We are not cited to anything in the history of airline negotiations to support the idea that the use of the suspension power recently conferred upon the CAB by Congress would necessarily produce an unresolvable confrontation with a foreign country. While we are told that the proposed fares by Pan Am and TWA “were protested by numerous key European governments,”18 we are not told how many countries originally protested, to what portions of the fare each took objection, how far different the U.S. proposed fares were from those proposed by other individual national airlines, whether in truth the proposed U.S. fares were actually agreed to by a majority of European countries. Indeed, the inference to be derived from the Board’s reference to the hostility of “key European governments” is that air services on a mutually satisfactory basis with many countries were entirely possible.
It is obvious, as Board counsel stated at oral argument, that the consequences of an open rate situation are mutual for the United States and the foreign governments.19 It is equally obvious that the foreign governments he.re involved have an enormous stake in the continuance of air traffic across the North Atlantic. In the year 1972, 582,411 United States citizens flew United Kingdom aircraft across the North Atlantic; U.S. passengers on the airlines of Germany, France, the Netherlands, Italy, and Switzerland in the North Atlantic ranged from 364,803 to 214,520 passengers each.20 Two thirds of the passengers on all eighteen carriers on the North Atlantic route are U. S. citizens. In these foreign countries the stake of the tourist industry — hotels, restaurants, bus and tour services, etc. —in addition to the U.S. passengers carried by their national airlines, is enormous.21
[1025]*1025The consequences of a cessation of air service between the United States and any one of these foreign countries must be every bit as unfortunate for the foreign government as for the United States to contemplate. Perhaps further explication by the Board would make this clear, but on this state of the record we are unable to see why in bilateral negotiations a compressed time span should always put more pressure on the United States side to accede to the demands of the other carrier for an IATA agreement. In multilateral negotiations the United States is one nation among several nations; in the North Atlantic route area one among sixteen, with three carriers among eighteen; in bilateral negotiations the United States is in a one versus one position, with whatever advantages this might bring.22 In a bilateral negotiation the cessation of all air traffic might be just as far beyond the contemplation of a foreign government as on this record the CAB has revealed that it is beyond its own contemplation.
Nor are we given any reason to believe that an unresolvable confrontation with one or more countries would necessarily be disastrous to the United States in foreign policy terms. We bear in mind that the CAB, after all, does not have the final word on international routes or fare structures; the President has power to set aside the CAB suspension of any foreign airline’s tariffs or landing rights in the United States, if in his judgment other foreign policy reasons call for this.23
While there is concededly more stability when a long-term IATA agreement has been accepted by foreign governments and approved by the Board, it should be noted that perhaps the very term “open rates” overstates the fluidity of the situation which would result if the Board did not approve the IATA agreement. Even if the Board acted to suspend subsequently filed individual [1026]*1026carrier tariffs, under the power granted to it by 49 U.S.C. § 1482(j) (1), the statute Specifically provides that “the affected air carrier or foreign air carrier shall maintain in effect and use the rate, fare or charge . . . which was in effect immediately prior to the filing of the new tariff.” In other words, even under the dire circumstance of suspension, the situation pending further bilateral negotiation would be precisely the same as that perpetuated by the IATA agreement approved here (less six percent, to be sure).
And, in that light it is of considerable interest that, according to the uncontested assertions of petitioners at oral argument, the recent 60-day price freeze gave rise to a sort of open rate situation in the North Atlantic — without any of the dire results predicted by the Board. In pleading to the Cost of Living Council for an exemption from the application of the price freeze Executive Order, the Board went through much the same litany of horrors, citing “the delicacy of the matter,” as they have recited here.24 While the Council’s denial of an exemption apparently operates merely to invalidate the Board’s approval of an across-the-board six percent increase, it does operate to negate an IATA agreement — and instead of immediate world airline chaos and bedlam, the Board’s predictions here, in contrast we have a rather deafening silence.
B. Benefits from, Status Quo
The affirmative and hopeful side of the Board’s findings reflects a conviction that just this one more delay will give the time needed for negotiation of a satisfactory rate structure. However, we are cited to nothing in the basic motivations or past conduct of the carriers to support this view. For example, the ten-page extract from the March 1973 London Currency Conference deals with positions of various airlines on the percentage surcharge which should be imposed on the 1972 fare structure, not with positions taken on redesigning a more equitable new fare structure; apparently the latter effort had long since been abandoned.
Nor are we cited to any change in the Board’s conduct or position which would support the conclusion that the situation will improve next time around. In response to questions at oral argument, the Board’s counsel was unable to inform us as to what new or different activity is contemplated by the Board during the remainder of 1973 which would serve to support a belief that the ultimate result for the 1974 fares will be different from that of 1972 and 1973.25
[1027]*1027The Board’s major articulated positive support for choosing the status quo was the benefit to passengers of stable rates. Yet, even if we assume that the CAB’s decision helped, to assure temporarily a stable and predictable fare structure, there is no showing that that contributed to earlier achievement of a fare structure which the Board could find to be in the public interest as an economic matter. Indeed there is an argument that the CAB’s action provided time for leisurely negotiation — whereas a realistic prospect of confrontation might have forced quick IATA agreement on a more acceptable fare structure.
IV. Arbitrary Exercise of Discretion
To counter the suggestion that the record is devoid of evidence to support its factual predicates, the Board sought to characterize this as a case in which the decision subject to review was the sort of policy-making judgment-call which is not to be tested by “substantial evidence” criteria in the first place.26 To the extent that a decision to avoid an open rate situation represented a combination of “administrative prediction” and strategy in the face of essentially unprovable levels of risk, we must agree that this decision did constitute an exercise of agency discretion — the sort of determination for which this court should have the utmost deference in view of administrative expertise. However, as can be seen from the extensive quotations herein, the Board’s discussion is a recital of conclusions, not facts, and of conclusions with little supporting rationale visible. As a reviewing court we are thus placed in the position of either accepting, on the basis of sheer faith in the Board’s “expertise,” the unsupported conclusions, or of vacating the order and remanding for reexamination and creation of a record relevant to what the Board has clearly designated the decisive issue. The very nature of the Board’s rationale for its action reveals that it conceived of approval of the multilateral IATA agreement as the only available alternative. We find that arti^ ficial narrowing of options to be arbitrary and capricious and, in effect, an abuse of the discretion specifically delegated to the Board by Congress.27
[1028]*1028Even prior to the specific legislative grant of power to suspend international fares, this court reminded the CAB that “the fact that IATA has some leeway to make the threat of an open rate situation imminent rather than remote . does not relieve the Board of its responsibility to make an adequate appraisal of the public interest under section 412 of the Act.”28 The court went on to say “the specter of an open rate situation does not absolve the Board of its responsibility to employ procedures which will allow it to discharge the task adequately .... To abrogate the right to disapprove an IATA agreement on the basis of its disruptive effects would reduce Government review to a perfunctory exercise negating any need for review.” 29
The Board’s total rejection of all options except approval of an IATA agreement, because it apparently felt almost any agreement was preferable to an open rate situation, is rendered even more improper by the subsequent passage of legislation specifically designed to lessen the terror of the open rate situation in order to strengthen the Board’s hand in reviewing IATA agreements.
All previous history of airline rate negotiations demonstrates the CAB and the U.S. carriers seeking to achieve a multilateral IATA agreement. Prior to 1972, the alternative to a multilateral rate agreement was the individual filing of tariffs by U.S. and foreign carriers, and the CAB had no authority to suspend the tariffs of foreign carriers. The economics of the air carrier situation then meant that foreign carriers would file what the CAB and the U.S. carriers rightly considered to be uneconomically low rates. However, in 1972, aware of the weakness of the United States position, at the request of the CAB and'the U.S. airlines the Congress conferred on the CAB the power to suspend foreign tariffs, a power which foreign governments had long had with reference to United States international carriers.
In so doing, the Congress expressed its belief that it was strengthening the CAB bargaining position, and that it expected
During the period prior to and following suspension of a rate, the United States would engage in discussions with the other government or governments involved as provided in the bilateral agreements. The discussions would seek resolution of the disrupted fare situation, and the Board would be expected to take the results of the discussion into account in reaching any decisions on whether to reject a fare.
The power to suspend and reject rates will arm the Board with sufficient authority to exercise broad influence on international ratemaking and will keep in effect subparagraph (f) of Article II of the Annex to the U.S.-Bermuda-type air agreements so that unreasonable rates may be held in abeyance by the U.S. Government pending investigation or a negotiated agreement.30
It is obvious from this that the Congress contemplated bilateral as well as multilateral negotiations as an available alternative for the Board. It is also obvious that the Congress considered that it had given the Board a strong hand and that use of the suspension power would be considered a viable alternative. Quoting from a previous Senate report in regard to the deplorable backdown of [1029]*1029the United States in the 1963 air fare dispute, the Senate said: ^It is more than slightly ludicrous that this Nation, which is predominant in the field of aviation and international air travel, should be unable to resist the threats and dictates of foreign governments with respect to international air fares. With the additional powers which would be added by favorable action on S. 1540 (as enacted in 1972, S. 2423), the Board will be able to effectively utilize and wield this economic power in negotiations with foreign countries.” 31
From what appears on this record, U. S. Government officials and the CAB have never sought to negotiate a bilateral agreement with any other given country covering flights from the U.S.A. to that country alone. While the Board did conduct bilateral discussions, these were all aimed at procuring unanimous multilateral agreement, rather than individualized understandings pertaining to separate routes. There is a vast difference between the strategy appropriate to achieving bilateral as opposed to unanimous agreement — and the Board’s commitment to unanimity above all else is demonstrated by the fact that it continued to allow each individual country to exercise effective veto power over any change in the overall rate structure.32 From the record of successive Board orders, it appears that the Board adhered to unanimous IATA multilateral agreement as being the only conceivable alternative and viewed the open rate situation (no matter what consequences flowed from it) as unthinkable, even though Congress had specifically created a new bargaining tool for the CAB to deal with this anticipated situation. In" other words, the CAB did not actually or rationally choose among the possible alternatives, as contemplated by Congress.u
In significant contrast, in its later oyder in Docket No. 25519, which is currently under review in this court in No. 73-1777, the Board did think the unthinkable, at least as to the possibility of bilateral agreements. The Board said that the reasons that “the Board has required unanimity of voting within IATA since its inception” are in part “based on circumstances some of which have since changed, not the least being the Board’s recently acquired suspension power over international tariff filings.” It is of course this new grant of power which changes the whole rationale behind seeking unanimity. Previously, if the CAB did not achieve unanimous IATA agreement, any foreign carrier could undercut U.S. carriers with lower rates. Now, the suspension power protects against such a result and vitiates the rationale of requiring unanimity.
In weighing the duty of the CAB to consider alternatives other than unanimity by IATA agreement, it should not be forgotten that this whole IATA concept as the most desirable and so far inevitable outcome of negotiations is definitely contrary to the philosophy of the antitrust laws, contrary to our usual view of the public having the benefit of either competitive rates or rates set by a regulatory body in the public interest.
We stress we are not second-guessing the CAB on its choice of alternative courses of action. Rather, our concern has been that the CAB has not made the first guess — has not made a considered evaluation of the presently available alternatives in light of the new 1972 suspension power. We are not attempting" to weigh for the CAB the comparable consequences of a possible cessation of air services and the respective strengths of the parties in bilateral negotiations; we are pointing to the absence of any indication in this record that the CAB or other United States representatives ever seriously considered the alternatives which Congress had specifically conferred upon the CAB in the 1972 statute for the specific purpose of [1030]*1030strengthening the hand of the CAB in these negotiations.
In his typically pungent style, Judge Tamm’s dissent phrases the Board’s only viable choice as “obtaining for American carriers . . . part of something rather than all of nothing.” If the CAB in truth viewed these as the only two choices, if it thought that “Our boy can’t fight”, then its error becomes crystal clear. For certainly there is no substantial evidence in this record supporting a conclusion that “all of nothing” was ever a realistic alternative. Indeed, all logic from the available record points to an opposite conclusion. Nor do we know whether “Our boy can’t fight”; he was never sent into the ring in the manner Congress intended.
Yet an administrative agency’s discretion is the power to make a reasoned choice of alternatives within a class of permissible actions. Whether the CAB be viewed as having made a choice, but without any reasons having been given for that choice, or viewed as having in actuality made no choice at all, merely homed to the IATA agreement from historical habit, it appears to us the net result is the same; the order must be vacated as either an unsupported exercise of discretion or an arbitrary failure to exercise discretion at all.
V. Conclusion
Just as the Board predicted that chaos would result from any decision other than acquiescence in the IATA agreement, CAB counsel hinted at oral argument that the collapse of airline traffic as we know it would follow this court’s failure to adopt unquestioningly the agency’s line. We do not agree. As previously mentioned, adverse effects on foreign relations can be controlled by means of the President’s right to review the CAB’s decisions concerning foreign air transportation. Given the vested interests involved, and the direction of net benefits flowing from continued air service, we think a total breakdown in North Atlantic air service extremely unlikely — and any continuation of service will not only fulfill emergency needs but also act to force recalcitrant parties to quicker agreement. In addition, assuming no widespread cessation of service, even a full confrontation and tariff suspension scenario will result in the continuation of the status quo as to rates for an interim period.
We have no doubt that at least some time will be needed for the carriers and the Board to adjust to this new state of affairs. This is usually the case when a regulatory agency’s order is vacated. Some leeway is allowed by the fact that, as a matter of course, our mandate will not issue until twenty-one days after this opinion issues. We also note that the extract from the Minutes of the London Currency Conference 33 indicates that the next IATA rate conference is scheduled for early October and that the North Atlantic Policy Group has been meeting “as necessary” in preparation, with a meeting definitely planned for September.
We have no doubt that the Board, the President, and this court can act effectively to prevent intolerable interruptions in foreign air transportation in the North Atlantic. We do not feel that speculative “horribles” should deter a firm rejection of action by the Board unsupported by substantial evidence relevant to the critical issue, action which appears, not as a principled choice of alternatives, but as failure to exercise discretion.
Whether the Board’s order extending the basic North Atlantic fare structure through 1973 is viewed as ratemaking, involving findings of fact, or an exercise of policy-making discretion, we find the decision unable to withstand applicable standards of review.34 Since vacation of
[1031]*1031this central feature of Order 73-4-64 will no doubt cause the Board to rethink its fundamental premises, we see no purpose to be served by separately reviewing the petitioners’ additional arguments concerning the six percent devaluation increase and the Board’s decision not to launch an investigation. Accordingly, the entire order is vacated and the case is remanded for proceedings consistent with this opinion.
So ordered.