dissenting:
Because in my view, the court, like the Maritime Subsidy Board before it, misinterprets the purposes and mandate of a key section of the Merchant Marine Act of 1936,1 I must respectfully dissent.
I
Additionally to construction-differential subsidies,2 the Act offers operating-differential subsidies — the concern of this appeal — for qualifying shipowners and operators.3 The Board 4 is authorized to grant an operating-differential subsidy when “re*114quired to meet foreign-flag competition and to promote the foreign commerce of the United States . . . .”5 Congress provides these subsidies because operating costs — primary wages — for foreign-flag ships are lower and because many foreign nations subsidize their own commercial vessels, thus rendering a wholly unsubsidized United States merchant marine economically untenable.6 To protect other United States-flag carriers and the federal fisc,7 however, Section 605(c) of the Act8 requires that when another United States line is already serving the trade route for which an operating-differential subsidy is sought, the Board must conduct a hearing encompassing designated issues and make findings thereon.9
*115American President Lines, Ltd. (APL)10 seeks Board approval of additional service11 on one of its subroutes.12 The extension of service involves a larger number of sailings of vessels already subsidized, and the Board analyzed APL’s application under Clause 1 of Section 605(c),13 which states that
[n]o contract shall be made under this title with respect to a vessel to be operated in an essential service served by citizens of the United States which would be in addition to the existing service, or services, unless the Secretary of Commerce shall determine after proper hearing of all parties that the service already provided by vessels of the United States registry is inadequate, and that in the accomplishment of the purposes and policies of this Act additional vessels should be operated thereon . . . ,14
Both APL and Sea-Land Service, Inc., the party challenging the subsidy contract, agree that the trade route in question is an “essential service,” and that there is “existing service” on it. The disputed question of fact before the Board was whether that existing service is inadequate and the question of law — the one we must now answer15 —is how adequacy is to be determined.
*116Fortunately, we are not left without guidance. As the Board itself has often indicated, adequacy of existing service on a particular trade route generally' is to be ascertained in light of the objectives of the Act.16 And surely a court is duty bound to that course even though the language of Section 605(c) can be read to indicate that the two are separate considerations.17 Relatively little is more firmly embedded in our law than the canon that legislative purpose is the touchstone of statutory meaning.18 “Adequate” cannot sensibly be taken to connote anything but whatever is “adequate to achieve the purposes and policies of the Act.”19
*118II
The major objectives of the Merchant Marine Act of 1936 are readily ascertainable. As APL itself has aptly observed,
the predominant concern of the enacting Congress was the fear that foreign-flag service of the American trades would be terminated abruptly when war or other foreign emergency forced the withdrawal of the vessels. An American-flag merchant fleet was viewed as indispensable because it would remain in the service of our commerce when the foreign vessels were withdrawn.20
♦ :J« * * sje $
The other great purpose of the Act is to provide the ships necessary to the national defense.21
That these are the grand designs of the Act is set forward with relative clarity in Section 101:
It is necessary for the national defense and development of its foreign and domestic commerce that the United States shall have a merchant marine (a) sufficient to carry its domestic water-borne commerce and a substantial portion of the water-borne export and import foreign commerce of the United States and to provide shipping service essential for maintaining the flow of such domestic and foreign waterborne commerce at all times, (b) capable of serving as a naval and military auxiliary in time of war or national emergency . . . ,22
These concerns — for uninterrupted export of our own products and import of foreign goods essential to our economy, and for a merchant fleet available and sufficient to supplement our military effort during an era of crisis — were articulated often during congressional consideration of what became the Merchant Marine Act of 1936.23
*119The Board, expressly relying upon the legislative aims set down in Section 101,24 has established fifty percent as a basic guideline fór determining adequacy.25 This figure, however, can have no meaning unless one knows what must be fifty percent of what. Nonetheless, the Board has never bothered to explain what relationship the figure describes. This is but a reflection of the problem central to this case: How, under the provisions of the Act properly construed, adequacy is to be ascertained; in other words, adequacy involves the relationship of what to what? Both the Board and the court say that the relationship is between the amount of cargo in our commerce and the capacity of United States-flag ships routinely devoted to that cargo, and that since the total United States-flag capacity ordinarily filled with United States-bound cargo is less than fifty percent of all United States-bound cargo on the trade route in question, the existing service is inadequate. Without impungning in any way the Board’s prescription of fifty percent as the mathematical point of demarcation between adequacy and inadequacy, I submit that the wrong factors have been compared.
Both of the predominant purposes of the Act relate directly to the facts of shipping capacity and volume of waterborne cargo during an emergency — war or a foreign country’s cutoff of its shipping services for any reason whatever. Thus, at least with reference to these two critical objectives, adequacy is to be measured not by whether a substantial amount of United States cargo is ordinarily carried by United States-flag vessels, but rather by whether it could be so carried if foreign-flag service to the United States were to be appreciably interrupted. The Act itself is filled with references to that hypothetical context: “sufficient to carry,”26 “capable of serving,”27 just to mention two. It will also be recalled that the statute is so worded as to protect “essential” routes, not all those habitually traveled.28 This is a further indication of *120the overriding congressional concern with assuring against a loss of shipping vital to the domestic economy and the national defense.
Most determinations under the fifty percent standard are relatively straightforward because routine carriage of foreign wayport cargo in United States bottoms seldom occurs. In the normal instance, if United States-flag vessels are now carrying, and by projection will continue to carry, fifty percent of our waterborne foreign commerce, it can safely be assumed that they will be able and available to do so in an emergency.
The situation under scrutiny in this case, however, is complicated by the location of the subroute involved. Because the British Columbia ports of Vancouver and Victoria are so close to our Pacific Northwest ports, United States-flag vessels serving our ports commonly carry Canadian cargo.29 Thus, present-day figures on United States-flag carriage of United States-bound cargo, as well as projections of what a future uneventful year will look like, are not descriptive of the merchant marine’s capability in a crisis. Regardless of whether in routine 1975 business United States-flag vessels transported substantial Japan-to-Canada cargo, in an emergency all of the tonnage of those ships would be available to maintain the commerce of the United States and to provide support services for the military. As the court admits, “a ship is a ship” regardless of its current usage.30
The routine carriage of Canadian cargo thus appears irrelevant to the goal of subsidization, which, as noted earlier, is to assure that in the absence of foreign-flag service United States-flag carriage could meet our country’s basic economic and defense needs.31 In line with that objective, the pertinent comparison is between those needs — here, the projected traffic inbound to United States ports32 — and our merchant fleet’s ability to serve those needs.33 I fail to discern why that capability should be established on the assumption that, even in the event that foreign-flag service to the United States is rendered unavailable to the Nation’s shippers, our merchant marine will continue, and will be allowed to continue, on a business-as-usual basis to devote a substantial percentage of its resources to serving the economy of Canada.34 Some *121cogent rationale might lurk behind the Board’s ipse dixit that inclusion of Canadian cargo in its calculation of adequacy is the “more reasonable approach.”35 Yet if so, the Board gave no hint either to its existence or its shape, and I cannot accept my colleagues’ attempt to give it form; that is the agency’s job, not the court’s.36
APL, the Board and the majority argue that the tonnage of United States-flag ships customarily devoted to carrying cargo in another country’s commerce should be excluded from the computation of adequacy. Accordingly, only the capacity ordinarily utilized for United States-bound cargo would be compared to the total United States-bound cargo.37 This interpretation reads Section 605(c) as a statement that “[n]o contract shall be made . . . unless the Secretary of Commerce shall determine . . . [that United States-flag ships are not customarily carrying substantial portions of all United States-bound cargo.]”
As a matter of statutory construction, this would seem reasonable if, as the majority appears implicitly to assume, the statutory purpose were simply to have much of the cargo in the stream of our foreign commerce routinely carried in United States-flag vessels. It can be argued that one purpose of the Merchant Marine Act of 1936 was merely to meet day-to-day foreign competition in commercial shipping,38 but the argument mistakes means for ends.
The legislative history of the Act reveals that more United States-flag ships in operation was not an end in itself39 but was a *122means of assuring that the vessels, and as well trained and experienced crews and shipbuilders,40 would be available in an emergency.41 Congress believed that having the ships on the open seas was preferable to having them in mothballs.42 And it was obvious that to keep them sailing it would be necessary for our Government to provide operating subsidies to combat the lower operating costs of competing foreign-flag vessels and their subsidization by their own governments.43
III
The preceding analysis seems rather unexceptionable, and it may well be that if APL had applied for increased subsidization for the increased service which it proposes to institute, my colleagues would agree that the application should have been denied. Perhaps at the heart of what bothers them is the fact that APL apparently is not asking for more money.44 It asserts that it seeks simply to initiate a more efficient *123method of shipping — containerization—that would allow its currently subsidized vessels to make more voyages per year.45 It would seem eminently sensible on pragmatic grounds to allow APL to give the Board and the Nation more capacity for the public dollars it is already receiving, and the court strives to garner this valuable benefit.46 Nonetheless, the Board has decided that APL’s request is subject to the requirement of a Section 605(c) hearing,47 and APL has not challenged that holding here.48 I am unwilling to distort the standards employed in Section 605(c) determinations to reach a wished-for practical result in a case to which the section perhaps should not be applied at all.49
My colleagues’ main argument is that the statute should not be interpreted to give unsubsidized lines an undue competitive advantage and that, if adequacy is found in this case, only the unsubsidized lines could expand their services to meet increased demand.50 To the extent that this is true, it remains true no matter how adequacy is defined if the facts of the given case meet that definition. Thus, by twisting the defi*124nition of adequacy with a view to finding inadequacy in this situation, the majority effectively reads the adequacy requirement out of Section 605(c) for this class of cases.
I cannot accept the majority’s premise that unless inadequacy is somehow found, and consequently justification for approval of APL’s application is established, the subsidized line is unfairly denied the chance to expand to enjoy the opportunities created by new demand. APL would not be prohibited from competing for that increased demand; it simply could not be subsidized in doing so if there is already adequate service. It may compete, and indeed it has competed,51 on an unsubsidized basis, and the Act confines it to that role if existing service meets the minimum level of adequacy. Apparently what the court means is that because APL is not asking for any more dollars it in reality is merely another unsubsidized line with respect to the increased demand and should not be limited by its base-demand subsidization in seeking to meet the enlarged demand. This argument also gains its pragmatic force, if any, from the situation created by the Board’s unchallenged conclusion that increases in service but not in subsidy are subject to Section 605(c).52 I cannot accept ¿his distortion of the Act on such a tenuous basis.
IV
My colleagues praise the Board’s “concern for the real world,”53 and APL complains that the District Court’s ruling would “produce an artificial and distorted result.” 54 Although I agree completely that regulatory agencies should not “depart from this worldly context,”55 both the majority and APL seem misguided as to which “real world” facts are relevant to the purposes of the Merchant Marine Act.
The court’s formula does describe the volume of goods in our commerce carried from day to day in United States vessels, but the normal commercial world is not the world with which Congress was primarily concerned in enacting this legislation. Congress in 1936 looked toward a future world, but one — as the events of 1939 to 1945 showed — whose futurity did not make it any less real. The world with which Congress was preoccupied was a very real world of potential conflict — in particular, a world in which foreign-flag service might, and undoubtedly someday would, be interrupted. And, in the view of Congress, our *125merchant marine is sufficiently fortified to protect against the dangers of that future reality if it can carry a substantial portion of our foreign commerce and can act as a naval and military auxiliary. The fact that something less than a substantial portion of our foreign commerce is routinely carried in United States-flag vessels is interesting but, on this record, is irrelevant to the statutory objectives.
It may be that a grant of APL’s request is justified on some ground other than the Board’s erroneous interpretation of what should guide a determination of adequacy, and I would give the Board an opportunity to remedy its deficiencies. For example, if it should find that, notwithstanding the inclusion in its equation of capacity routinely devoted to Canadian cargo, United States-flag service is inadequate, regardless of the numerical percentage arrived at, and that increased subsidized service will comport best with the purposes of the Act, we all would agree that it may grant the application.56 But I cannot subscribe to the notion that the Board may set the amount of subsidized service by reference to factors bearing no apparent relationship to the purposes of the Act.
. Act of June 29, 1936, ch. 858, 49 Stat. 1985, as amended, 46 U.S.C. §§ 1101-1294 (1970 & Supp. V 1975).
. Merchant Marine Act of 1936, tit. V, 46 U.S.C. §§ 1151-1161 (1970 & Supp. V 1975).
. Id. tit. VI, 46 U.S.C. §§ 1171-1183a.
. The Secretary of Commerce has delegated her authority under the Act to the Maritime Subsidy Board. See 46 C.F.R. § 252.13 (1976). The Secretary denied a petition to review the Board’s decision in this case.
. Merchant Marine Act of 1936, § 601(a)(1), 46 U.S.C. § 1171(a)(1) (1970).
. E. g., S.Rep. No. 1721, 74th Cong., 2d Sess. 7, 10 (1936); S.Rep. No. 748, 74th Cong., 2d Sess. 3 (1935); H.R.Rep. No. 1277, 74th Cong., 2d Sess. 3, 12-17 (1935); Hearings on H.R. 7521 Before the House Comm, on Merchant Marine and Fisheries, 74th Cong., 1st Sess. 61-62 (1935) (testimony of Alfred E. Haag, Chief, Division of Shipping Research, United States Shipping Board Bureau) (hereinafter cited as “House Hearings ”).
. See Sea-Land Serv., Inc. v. Connor, 135 U.S.App.D.C. 306, 313, 418 F.2d 1142, 1149 (1969) (“since these grants are a part of the government wealth, it is incumbent upon the Secretary to follow sound administrative procedures to determine the necessity of expenditures and thereby serve the master, public interest”).
. 46 U.S.C. § 1175(c) (1970).
. Sections 601 and 605(c) appear on their face to establish three basic standards for subsidization. When the line seeking the subsidy desires to institute new service on a route already served by another United States line, it must show that the existing service is inadequate to meet the purposes of the Act. Merchant Marine Act of 1936, § 605(c), 46 U.S.C. § 1175(c) (1970). This differs from the showing requisite in the absence of service on the route by another United States line. In that case, the Commission in its discretion may subsidize United States-flag capacity that exceeds the bare minimum necessary for achievement of the purposes of the Act. See id. § 601(a), 46 U.S.C. § 1171(a). This distinction is soundly based upon the desire to preserve and promote existing lines if possible, and on the fear that subsidies might eventually drive out unsubsidized lines. H.R.Rep. No. 1277, supra note 6, at 23 (purpose of § 605 [then-§ 535] is “to protect operators on existing routes”); cf. Pacific Far East Line, Inc. v. Federal Maritime Bd., 107 U.S.App.D.C. 155, 156, 275 F.2d 184, 185, cert. denied, 363 U.S. 827, 80 S.Ct. 1597, 4 L.Ed.2d 1523 (1960) (Congress intended to prevent subsidized lines from diverting subsidy funds to unsubsidized domestic operations “to the disadvantage of an unsubsidized operator”). See generally Whitehurst, The Merchant Marine Act of 1936: An Operational Subsidy in Retrospect, 8 J.L. & Econ. 223, 240-242 (1965) (despite safeguards, subsidization drove out many unsubsidized lines over time). As the Board’s predecessor said in one of the earliest interpretations of the Act:
We have stated as a matter of policy that we prefer that private United States-flag operations be conducted in the foreign trade without Government aid, but we will enter into contracts for the payment of operating-differential subsidies, in accordance with the provisions of the law, whenever this is found necessary to maintain adequate United States-flag service on essential foreign trade routes.
Bloomfield S.S. Co., 3 F.M.B. 299, 306 (Maritime Comm’n 1946); accord, American South African Line, Inc., 3 F.M.B. 314, 321 (Maritime Comm’n 1947) (“it should be understood, however, that the Commission would not pay a subsidy to an American-flag operator for operating an essential foreign service that could and would be adequately maintained on a long-range basis by an American-flag operator without subsidy”). If current service is below the minimum level of adequacy, the Board cannot afford to confine operations on the route to existing lines, but otherwise their interests and the desire to avoid unnecessary expenditures should normally triumph. Final Opinion and Order of the Maritime Subsidy Board, Joint Appendix (J.App.) 87 (hereinafter cited as “Board Opinion”); Brief for Federal Appellants at 26-28; see Hearings on S. 2582 Before the Senate Comm. on Commerce, 74th Cong., 1st Sess. 512, 547 (1935) (hereinafter cited as “Senate Hearings on S. 2582 ”) (finding of inadequacy referred to as “certificate of necessity”); House Hearings, supra note 6, at 691, 737 (testimony of Ira A. Campbell, counsel for the American Steamship Owners Association and draftsman of the' original version of § 605(c)) (“it is not sound for the Government to subsidize two carriers to go into competition with each other” unless “it is necessary so that Americans will be able to carry 50 percent or 51 percent of her commerce”).
A middle category involves situations in which the service sought to be subsidized was being provided by the applicant on an unsubsidized basis prior to the subsidy application, *115and another United States line was also providing service on the same route. Merchant Marine Act of 1936, § 605(c), 45 U.S.C. § 1175(c) (1970); Brief for Federal Appellants at 7 n.2. There, if foreign operating costs are low enough to justify a subsidy, the Board has discretion to grant one even though the aggregate service is adequate, unless the other existing line shows that the subsidy would be “unduly prejudicial.” Pacific Transp. Lines, Inc., 4 F.M.B. 7, 19-20 (Bd.1952). The assumption seems to be that when both lines are already providing service the demand must be such that the other line is less likely to be injured than it might be if new or additional service were being sought. Thus, the desire to assure that the applicant for the subsidy is not forced to abandon its service is given greater weight.
To summarize, if the current service is inadequate to meet the Act’s goals, subsidization is allowed regardless of the presence of United States-flag competitors. If it is inadequate but there are no such competitors, the Board may still grant a subsidy through a § 601 determination. If existing service is adequate and both lines provided prior service, the line not seeking a subsidy must show enough to warrant a § 605, cl. 2, determination that it would be unduly prejudiced. Finally, if existing service meets the minimum threshold of adequacy, and the line seeking the subsidy did not serve the route prior to its application, a § 605, cl. 1, denial of the subsidy will generally be in order. See generally The Ocean Freight Industry, Report of the Antitrust Subcomm. of the House Judiciary Comm., 87th Cong., 2d Sess. 354 (1962) (hereinafter cited as “The Ocean Freight Industry").
This synopsis of the statute seems to conflict, at least on its face, with a number of recent Board decisions indicating that even if the existing service is adequate the Board may approve subsidization to attain even greater United States-flag capacity. E. g., United States Lines, 5 S.R.R. 969, 977 (Bd.1965). As may appear from my discussion in note 19 infra, however, these cases might simply mean that a higher standard of adequacy is appropriate if excess overall demand makes increased United States-flag subsidized service “practically available” with little or no injury to unsubsidized lines. In effect then, the presence of excess demand might result in a situation more closely analogous to that contemplated by clause 2 of § 605(c) than by clause 1.
. American Mail Lines, the original applicant, was merged into APL just before the Board’s decision. See Brief for Appellant APL at 4.
. The Board rejected APL’s contention that, because APL had instituted unsubsidized service after applying for the subsidy, its application should be viewed as one seeking subsidization of existing service and accordingly judged under the second clause of § 605(c). Board Opinion, J.App. 88. APL did not challenge that conclusion in the District Court.
. The general route in question is Trade Route 29, which includes shipping between United States Pacific ports and ports in the Far East. Initial Decision of Eugene E. Hunt, Administrative Law Judge, J.App. 4 n.l (hereinafter cited as “Initial Decision”). This dispute particularly involves ships calling at the Pacific Northwest ports of Seattle, Tacoma and Portland.
. See text infra at notes 46-49 for a discussion of the applicability of § 605(c) in this context.
. 46 U.S.C. § 1175(c) (1970).
. Although we yield “great deference to the interpretation given [a] statute by the officers or agency charged with its administration,” Udall v. Tallman, 380 U.S. 1, 16, 85 S.Ct. 792, 801, 13 L.Ed.2d 616, 625 (1965), that deference is more emphatically summoned when the question is one requiring expertise in the subject area, such as a dispute over whether the agency’s policy satisfactorily fulfills the statutory aims. The question here, however, is *116what those purposes are, and “the courts are the specialists, whether analysis of legislative history is called for, or whether the main process is one of finding the meaning of the words.” K. Davis, Administrative Law Treatise § 30.09, at 243 (1958); accord, Wilderness Soc’y v. Morton, 156 U.S.App.D.C. 121, 143-144, 479 F.2d 842, 864-865 (en banc), cert, denied, 411 U.S. 917, 93 S.Ct. 1550, 36 L.Ed.2d 309 (1973) (agency construction contrary to congressional intent must be overturned). The Board’s interpretation, moreover, was not adopted contemporaneously with enactment of § 605(c), is not a longstanding construction, and has not been implicitly relied upon by Congress, see NLRB v. Bell Aerospace Co., 416 U.S. 267, 275, 94 S.Ct. 1757, 1762, 40 L.Ed.2d 134, 143 (1974), and therefore it does not bear those characteristics calling for maximum judicial deference. Indeed, this appears to be the first time this issue has arisen, and the Board’s implicit decision to deemphasize the purposes of the Act when interpreting what is “inadequate” contradicts earlier decisions by itself and its predecessor. See cases cited in notes 16 and 24 infra. Lastly, it is hard to extend deference to a conclusion lacking reasoned explication. See note 19 infra and text infra accompanying note 36. It is impossible to be sure that the Board construed the Act in light of the congressional objectives or whether it merely decided, without reference to the Act, that it would be more “reasonable” to decide the way it did. See note 35 infra.
. Bloomfield S.S. Co., 4 F.M.B. 305, 317, 324 (Bd.1953) (standard of adequacy must be established in light of the purposes of the Act); accord, Majority Opinion (Maj. Op.), text at note 18; American President Lines, Ltd., 4 F.M.B. 681, 695 (Bd.1955). As the author of the original language of § 605(c) explained his proposal:
If the service is inadequate, then the new contract should be made. If additional vessels should be operated to accomplish the purposes of this whole Act, the permission should be given.
Senate Hearings on S. 2582, supra note 9, at 512 (testimony of Ira A. Campbell, counsel for the American Steamship Owners’ Association).
. As the court points out, the determination of adequacy must relate to the purposes of the Act. Maj. Op., text at notes 18, 50. The Board’s disagreement with that conclusion is totally semantic. It may be that the Board has employed the 50% guideline, see note 25 infra, so often that it has forgotten why it is employed; that it has made “adequacy” determinations so often that it has forgotten why it is ' doing so. This is indicated by the Board’s opinion in this case, wherein it explained the difference it perceives between “adequacy” and “purpose-fulfillment” with an “extreme example.” Board Opinion, J.App. 123. The example was “the time period preceding the entrance into war by the United States when our allies are engaged in hostilities and our ships are supplying them with goods while ‘competing’ against the few foreign-flag ships plying the embattled seas.” Id. The Board felt that in such circumstances a subsidy could be granted even though the United States-flag shipping was found to be adequate. Id. In what way the shipping could possibly be deemed adequate was left wholly unelucidated. The only way it can be explained is that the Board slavishly adheres to the 50% figure despite the facts of the particular situation and the goals of the Act.
On the other hand, “adequacy” and “the accomplishment of the purposes and policy of this Act” must be independently ascertained in one situation. That is where the existing service is found to be inadequate. The Board cannot grant a subsidy solely on the basis of inadequacy of shipping in a hypothetical crisis, but must still see that the other conditions and considerations of the Act, such as those in § 601(a), are met. The Board believes that some of these other “purposes and policy” matters, such as financial feasibility and subsidy costs, are not within the scope of a § 605(c) hearing, Board Opinion, J.App. 124, but we are not called upon today to review that conclusion.
. Philbrook v. Glodgett, 421 U.S. 707, 713, 95 S.Ct. 1893, 1898, 44 L.Ed.2d 525, 532-533 (1975), quoting United States v. Heirs of Boisdore, 49 U.S. (8 How.) 113, 122, 12 L.Ed. 1009, 1013 (1849).
. Both the Board and my colleagues tend to confuse adequacy to meet the Act’s purposes with the adequacy of the total supply of shipping to meet the demand of all cargo shippers. See especially Maj. Op. at note 69. If there is statutory inadequacy, the Act makes overall demand inadequacy irrelevant: Even if the to*117tal all-flag supply of shipping is adequate for the existing demand, the Board, in the event of statutory inadequacy, must still subsidize the operation of United States-flag vessels, which then presumedly would successfully compete with foreign-flag ships for the existing demand.
If there is the minimum level of statutory adequacy and no existing service by a United States-flag line, the Board may of course inform its discretion to subsidize operations above the minimum level by looking to whether overall demand exceeds overall supply. In such a situation, the Board could reasonably conclude that inadequacy of supply enhances the chances for success of increased subsidized service.
If there is a competing United States-flag line, however, the language of § 605(c) prohibits further subsidization once a minimum level of adequacy is established. This the court completely disregards. Its rationale rests almost totally on the premise that an increase in subsidized United States-flag service is “commercially feasible.” Maj. Op. at note 69. That is an important factor under § 601, but under § 605(c), even if a subsidized line could succeed economically on a given route, the subsidy must be denied if existing service is “adequate” and if the subsidized line’s success would come at the expense of an unsubsidized line. And even as to § 601 itself, Congress by calling for United States-flag service for a "substantial portion” of our commerce obviously intended that subsidies not be granted for service sufficient to carry all our trade although that service would be “commercially feasible.” Congress knew that there is a point of disutility at which the cost of subsidization is not justified by the marginal gains in protection against emergencies. And although it delegated to the Board the task of determining that point in each individual case, it did not direct the Board to shower tax dollars on every ship operator who can come up with a commercially feasible proposal, especially if that proposal would result in harmful competition with operators hearty and independent enough to do without a subsidy. The court today decides that “it would be highly desirable to have all foreign commerce carried in American ships,” Maj. Op. at note 69, and that that goal is so desirable that the Federal Government should pay for it, whatever the ultimate cost. I cannot agree that Congress has embarked upon any such thing.
Nonetheless, the fact that overall demand exceeds overall supply, one aspect of commercial feasibility, might indicate that further subsidization would not substantially injure the unsubsidized line. Because one important aim of § 605(c) is to fend off unnecessary injury to unsubsidized lines, it might be prudent for the Board, defining adequacy in light of the statute’s mix of purposes, to raise the adequacy standard where there is excess overall demand. Cf. Matson Nav. Co. v. Conñor, 258 F.Supp. 144, 157 (N.D.Cal.1966), aff’d per curiam, 394 F.2d 514 (9th Cir.), cert, denied, 393 U.S. 998, 89 S.Ct. 482, 21 L.Ed.2d 463 (1968) (“residual benefit to a company’s domestic carriage resulting from subsidization of voyages stopping at both domestic and foreign ports did not result in “unfair competition” with domestic carrier because “the California-Hawaii trade will grow enough to accommodate both”). This seems to be what is behind the Board’s decisions that it can approve requests resulting in greater service than is called for by the general guidelines for adequacy of United States-flag capacity even though there is existing service. See cases discussed in note 9 supra. If indeed that is the Board’s guiding rationale, I for one would benefit from more careful choice of its language. “Overtonnage” and “adequacy” have little or no meaning when set adrift from a particular context of capacity versus demand or capacity versus statutory purposes.
My colleagues appear to rely on the Board’s projection of excess demand to support their approval of its decision. At 107 of 185 U.S. App.D.C., at 772 of 566 F.2d. Yet the Board did not itself rest the decision on that basis and it is firmly settled that “[t]he premises upon which the validity of an administrative order is to be decided are only those upon which the agency predicate[s] its action.” Tygrett v. Washington, 177 U.S.App.D.C. 355, 362, 543 F.2d 840, 847 (1974), and cases cited in note 49 thereof. “[A] simple but fundamental rule of administrative law [is] that a reviewing court, in dealing with a determination or judgment which an administrative agency alone is authorized to make, must judge the propriety of such action solely by the grounds invoked by the agency.” SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 1577, 91 L.Ed. 1995, 1999 (1947); accord, Westminster Broadcasting Corp. v. FCC, 148 U.S.App.D.C. 332, 336, 459 F.2d 1356, 1360 (1972). And a reviewing court is not free to advance “an alternative, unstated ground to support an agency’s decision if that ground is one that ‘the agency alone is authorized to make.’ ” Gulf States Utils. Co. v. FPC, 411 U.S. 747, 764, 93 S.Ct. 1870, 1880, 36 L.Ed.2d 635, 647 (1973), quoting SEC v. Chenery Corp., 318 U.S. 80, 88, 63 S.Ct. 454, 459, 87 L.Ed. 626, 633 (1943).
Moreover, even assuming that the statute does not prohibit the approach taken by the majority, I am unwilling to speculate that the Board would both adopt the majority’s rationale and find it applicable here. For instance, an unsubsidized United States-flag carrier, such as Sea-Land, might be able to supply the excess demand, see at 108 of 185 U.S.App.D.C.. at 773 of 566 F.2d, and the policy of § 605(c) may require that it be given the chance to do so at the expense of the subsidy applicant. See cases discussed in note 9 supra.
*118The Board assumed that Sea-Land might suffer some detrimental effect but that the impact was acceptable because, as an unsubsidized line, Sea-Land could move its ships to a different route with greater ease than could APL. Board Opinion, J.App. 129. This course of reasoning pays scant heed to the Act’s plain demand that, whenever possible, foreign-flag carriers, and not unsubsidized United States lines, suffer competitive injury at the hands of subsidized vessels. Furthermore, Sea-Land should not be made to suffer from APL’s inability to respond quickly to competitive conditions caused by the Board’s cumbersome regulation of subsidized lines.
. Brief for Appellant APL at 24.
. Id. at 25.
. 46 U.S.C. § 1101 (1970). The 1936 Act adopted the goals of two earlier statutes, H.R. Rep. No. 1277, supra note 6, at 3-4, which called for a United States merchant marine “sufficient to carry the greater portion of its commerce and serve as a naval or military auxiliary in time of war or national emergency.” Merchant Marine Act of 1920, § 1, 46 U.S.C. § 861 (1970); see Merchant Marine Act of 1928, ch. 675, Pub.L. No. 70-463, § 1, 45 Stat. 689 (1928) (adopting purposes of the 1920 Act).
. E. g., 80 Cong.Rec. 10571 (1936) (Representative Moran) (discussing national defense and threat of cutoff of foreign shipping); id. at 9904-9906 (Senator McAdoo) (same); id. at 9899 (Senator Guffey) (same); 79 Cong.Rec. at 10290 (1935) (Representative Crowe) (same); id. at 10258 (Senator Copeland, Chairman of the Senate Committee) (same); id. at 10207 (Representative Ramspeck) (national defense); id. at 10107 (Representative Wigglesworth) (same); id. at 10095 (Representative Welch) (same); id. at 10089 (Representative Connery) (same); id. at 10088 (Representative Bland, Chairman of the House Committee) (same); id. at 10075-10076 (Senator Fletcher) (same); S.Rep. No. 1721, supra note 6, at 5 (“[i]f a match should be set to this tinderbox of national jealousies and conflicting ambitions, a large part of the foreign merchant fleet upon which the United States depends for carrying two-thirds of its foreign commerce would be withdrawn from this service, and the products of our American farms would again lie rotting in the fields or on wharves as they did in 1914 to 1917, and our warehouses would bulge with the unsold products of our factories. There would not be sufficient ships to carry them”); H.R. Rep. No. 1277, supra note 6, at 4-5 (discussion of the two purposes); S.Rep. No. 713, 74th Cong., 1st Sess. 7 (1935) (same); House Hearings, supra note 6, at 833-834 (Alfred E. Haag, Chief of the Division of Shipping Research, United States Shipping Board Bureau) (“we should have proper representation upon the seas so that we need not rely upon foreign shipping in the case of a national emergency or be deprived of effective means to compete in world markets”); id. at 47-48, 62-65, 121-125 (discussion of effects of wars on merchant marine); id. at 1239 (text of radio address by Secretary of Commerce Roper) (adequate merchant marine is necessary to protect commerce and national defense in event of war or other *119emergency); id. at 1236 (text of radio address by Representative Walsh) (same); H.Doc. No. 118, 74th Cong., 1st Sess. 1 (1935) (message from President Roosevelt) (discussing purposes of proposed legislation); id. at 16-17 (General Report of the Postmaster General to the President) (discussing war-time concerns); id. at 21, 25-27 (Report of the Interdepartmental Committee on Shipping Policy) (discussion of need for strong merchant marine to support national defense and development of commerce).
The purposes behind federal subsidization of a privately owned merchant marine were most clearly and logically explained by then-Senator Hugo L. Black, who noted that the objectives of such subsidies are:
To make certain that American farmers, manufacturers, and all American producers of goods can transport their products to foreign markets, regularly, speedily, and at reasonable cost regardless of economic or war disturbances in any part of the world.
To make available constantly an efficient and sufficient fleet of potential naval auxiliaries, manned by American citizens, whose ability, training, courage, and loyalty will assure the successful operation of the fleet in time of peace as well as war.
In accomplishing the foregoing to give steady employment to American working men in shipyards and industries supplying shipbuilding material and the plants wherein they work, thus keeping ready for instant action the means of rapid ship construction in time of emergency.
S.Rep. No. 898, 74th Cong., 1st Sess. 2 (1935). Whether the two primary statutory goals carry the same weight now as they did in 1935 and 1936 might be open to question. G. Gilmore & C. Black, Law of Admiralty 977 (2d ed. 1975) (“new patterns of power, by making long distance land wars unwinnable, may have affected drastically the indispensable premise of all ship construction and shipping subsidy legislation— the premise that a national carrying fleet is essential to the defense and to national power”).
. Lykes Bros. S.S. Co., 4 F.M.B. 455, 464 (Bd. 1954).
. Bloomfield S.S. Co., supra note 16, 4 F.M.B. at 317 (50% guideline based on legislative history of Merchant Marine Act); accord, Board Opinion, J.App. 116-117.
. Merchant Marine Act of 1936, § 101, 46 U.S.C. § 1101 (1970).
. Id.
. Id. §§ 211, 601(a), 46 U.S.C. §§ 1121(a), 1171(a). In determining that a route was essential, the Board has looked to whether the cargo transported over the route was so important to our economy that it would be dangerous for the United States to be “dependent upon a *120transportation system over which it had virtually no control.” American President Lines, 3 F.M.B. 457, 462 (Bd. & Maritime Admin.1951).
. Maj. Op., text accompanying note 22.
. Maj. Op. at note 65 and accompanying text.
. See text accompanying notes 20-23 supra.
. Inbound cargo alone was considered because the amounts of cargo already containerized and susceptible to containerization were greater inbound than outbound. Board Opinion, J.App. 116-117.
. Thus, the total projected capacity of United States-flag vessels, 432,000 long tons, should have been compared to the projected United States-bound cargo, 755,000 long tons, a ratio of 57%. Because this exceeds the Board’s 50% guideline, no one seriously contends that the Board’s error, if any, was harmless. No argument appears to have been made that the United States suffered appreciably from a lack of shipping capacity during the years of the Vietnam War.
. Of course, in an emergency, the country to which the wayport cargo heads might well be our ally, and we might want to continue carrying its cargo so as not to disrupt its economy. But the Board did not make such a determination in this case, and it is not our place to make it, even if we had the information to do so. There is not, for example, any data in the record as to what amount of the United States-bound cargo, if any, is routinely carried in Canadian-flag vessels or in the vessels of our other traditional allies. If the assumption is that we must be prepared in a crisis to carry our allies’ goods as well as our own, it seems equally logical to assume that they will continue to carry our goods as well as their own. Furthermore, despite our commitment to the country in question, such a foreign policy decision might be one that neither the Board nor this court is capable of making. Finally, Congress might have taken away from the Board the power to undertake such a decision. A proposal to amend the bill that became the Merchant Marine Act to incorporate as its purpose an assurance that the United States-flag fleet would be adequate to carry a substantial portion of the commerce of the United States and of the rest of the world was not accepted by Congress in 1936. Hearings on S. 3500 Before the Senate Comm, on Commerce, 74th *121Cong., 2d Sess. 298-299 (1936) (testimony of Victor B. Bendix, ship and freight broker).
. Board Opinion, J.App. 113. The Board’s rationale is very unclear. Its entire discussion of the issue was:
Even if we assume, for present purposes, that Section 605(c) does not permit to be considered in the pool of cargo any cargo “external” to the U. S. foreign commerce, the Judge’s treatment of Canadian cargoes is in error. In a determination of ship container capacity to containerized cargo, Canadian cargo, which the parties did carry in 1970 and which there is no evidence to indicate they will not carry in 1975, must be recognized. The Judge accepted [APL’s] and other ship operators’ container capacity without making any reduction for this cargo (it does not appear that the parties other than [APL] made any such reduction), but he excluded it from the 1975 pool of containerized cargo. A more reasonable approach is to provide for a reduction in ship container capacity with no reduction in the pool of cargo of containerized cargo.
Board Opinion, J.App. 112-113. It is agreed by all that the Board did not mean “with no reduction in the pool of containerized cargo,” and that what it instead had in mind was “along with a reduction in the pool of containerized cargo.” Brief for Federal Appellants at 18 n.26.
. See note 19 supra.
. Total cargo inbound to the United States was projected for 1975 at 755,000 long tons, and the projected 1975 capacity customarily devoted to United States-bound cargo, without approval of APL’s application, of United States-flag vessels was 281,000 long tons. This resulted in a ratio of approximately 38%, a figure below the Board’s guideline of 50%.
. This is apparently part of the court’s rationale. See at 122-123 of 185 U.S.App.D.C., at 787-788 of 566 F.2d.
. Some “goals” discussed might relate directly to increasing routine operations by United States-flag ships, but these were of minor, if any, importance. House Hearings, supra note 6, at 694 (Representative Welch) (“if you eliminate national defense from this bill, you might as well throw it out the window, because Congress would not be concerned about it at all”). They were more in the nature of beneficial impacts of the statute and not the purposes for which it was enacted. Compare S.Rep. No. 1080, 91st Cong., 2d Sess. 9-11, [1970] U.S. Code Cong. & Ad.News 4188, 4190 (purposes of 1936 Act are still relevant to 1970 amendments: “[t]o permit our security and economy to become totally dependent upon foreign vessels, operated by foreign crews, subject to the wishes of foreign governments^] would be to run an unacceptable risk”) with id. at 21-22, [1970] U.S.Code Cong. & Ad.News at 4195-4196 (1970 amendments would beneficially affect unemployment, freight rates, and the balance of payments deficit). None of these benefits, moreover, was discussed by the Board, and none of them appears on its face to be relevant to the determination of adequacy in this case. For instance, the desire to provide service where there is not already adequate service by anyone else, H.R.Rep. No. 1277, supra note 6, at 9, certainly does not apply to Trade Route 29.
*122Only two ancillary benefits were of much importance. Many Representatives and Senators mentioned the beneficent effects on the depression-era unemployment picture of increasing the demand for shipbuilders and seamen. See H.R.Rep. No. 1277, supra note 6, at 11; S.Rep. No. 713, supra note 23, at 7. Although unemployment is once again high, the legislative history as a whole — and particularly § 101 of the Act — makes clear that increased employment was only a desirable side effect and not a controlling purpose of the legislation. No party to this proceeding argues that the Act should be viewed as a public works project.
Some Congressmen apparently felt that having a large United States merchant marine would somehow lower transportation rates and thus increase our sales abroad. E. g., House Hearings, supra note 6, at 1237 (radio address of Representative Walsh); id. at 833-834 (testimony of Shipping Division Chief Haag). The fear of exorbitant rates seems to contradict the fear that foreign-flag ships could price below our costs and thus drive United States-flag ships out of the market. Compare H.Doc. No. 118, supra note 23, at 1 (message from President Roosevelt) (“subsidies granted by other nations, shipping combines, and other restrictive or rebating methods may well be used to the detriment of American shippers”). The apprehension was that foreign-flag ships whose own countries’ products are in competition with ours might delay shipments and set high rates for our cargo. 80 Cong.Rec. 10575 (1936) (Representative Sirovich); see H.R.Rep. No. 1277, supra note 6, at 8.
Neither of these concerns seems particularly relevant here because we deal with inbound goods. As long as there is domestic competition for these imports, high rates charged by foreign lines would not seem to hurt participants in our economy. As to outbound goods, our containerized vessel capacity actually at least matches projected cargo for all of Trade Route 29. Board Opinion, J.App. 115. In any event, the effect of subsidies on freight rates may be almost nil owing to the existence of a shipping conference governing rates on the route in question. See Initial Decision, J.App. 22. See generally The Ocean Freight Industry, supra note 9, at 99-100; D. Marx, International Shipping Cartels: A Study of Industry Self-Regulation by Shipping Conferences 264-267 (1953).
. S.Rep. No. 898, supra note 23, at 2; House Hearings, supra note 6, at 63 (Shipping Division Chief Haag).
. See G. Gilmore & C. Black, supra note 23, at 970 n. 65 (“[t]he final goal perhaps ought to be envisioned in terms of result — a going-concern merchant marine of the kind required by strategic and other considerations — rather than of ‘fair competition’ ” (emphasis in original)).
. See House Hearings, supra note 6, at 832-833 (Shipping Division Chief Haag) (“[i]f a national emergency arises, even if it does not take place within 5, 10, or 20 years, and those ships are merely kept in spot condition, it would still be a sound investment for the Nation”); id. at 64-65 (United States would have saved billions of dollars by having an adequate merchant marine available at the outset of World War I rather than having to attempt to construct one after the outbreak of war).
. See note 6 supra and accompanying text.
. It is not at all clear from the record that subsidization would remain at the same level in terms of either absolute dollars or effective benefit. The administrative law judge believed that “[n]o greater subsidy payments would be due [the] applicant if the proposals are approved (these may be less without approval) . .” Initial Decision, J.App. 4. The Board characterized APL’s arguments as “including the point that its application would require no increase in subsidy payments . . .” Board Opinion, J.App. 126. These conclusions are basically unexplained in the record before us. APL’s subsidy application states that the yearly subsidy after approval might be less than without approval due to the decrease in the number of seamen needed for container-ships. Record Exhibit 2, at 4. The existing contract does not contain any specific formula other than the general provision that the *123amount of subsidy will be sufficient to result in parity of costs with foreign vessels. Id. Exhibit 3, at 8-10. See generally C. McDowell & H. Gibbs, Ocean Transportation 268 (1954) (“[b]ecause of the difficulties in determining rates for subsidy payments, most of the postwar contracts have not embodied approved rates”).
The federal appellants reason that because “no increase in the number of vessels was sought, no increase in the amount of [operating differential subsidy] was expected or requested.” Brief for Federal Appellants at 13. Yet some components of the subsidy vary with the number of trips, even if the number of vessels subsidized per year remains constant. See, e. g„ 46 C.F.R. § 252.32 (1976) (subsidy formula for maintenance and repair).
In any event, even if APL would not receive more dollars, it cannot be doubted that it would effectively receive an increase in subsidization. If the more efficient configuration of APL’s ships allows them to carry the base-demand amount of cargo in less time, the continuation of what seems to be the same subsidy is actually a larger subsidy for that portion of the subsidy allocable to costs that are time-related and not trip-related. For example, the operating subsidy is aimed to a large extent at wage differentials. See note 6 supra. Thus, if an APL ship previously made, for instance, ten trips per year and now can make fifteen, the crew is now only working two-thirds of the year on base-demand cargo, and that would seem to justify only two-thirds of the wage-related subsidy. See, e. g., 46 C.F.R. § 252.31 (1976) (wage-related subsidization based on per diem rate multiplied by number of voyage days). See also House Hearings, supra note 6, at 40-43 (testimony of Solicitor Trimble of Commerce Department) (amount of subsidy for items such as wages and insurance necessarily varies with time needed for voyage). The rest of the subsidy thus should be allocated to increased-demand cargo, and that would give APL something of a competitive advantage over unsubsidized lines with respect to carrying the increased demand. APL implicitly admits this, arguing that if its application is denied it will be forced to “inefficiently lengthen its voyages, operate partially unsubsidized on a marginal profit basis, or reduce its fleet.” Board Opinion, J.App. 126 (emphasis supplied).
. Since the advent of the “container revolution,” Initial Decision, J.App. 80, many types of-cargo are shipped in containers that can more quickly be loaded and unloaded, thus allowing containerships to spend less time in port and more time between ports. See Sea-Land Serv., Inc. v. Connor, supra note 7, 135 U.S.App.D.C. at 308 n. 1, 418 F.2d at 1144 n. 1. Both the administrative law judge and the Board considered containerships to be a separate sub-market for purposes of determining adequacy because of the perceived lack of demand substitutability between containerships and break-bulk ships. Initial Decision, J.App. 24; Board Opinion, J.App. 95. This determination has not been challenged.
. Cf. American Pres. Lines, Ltd., supra note 16, 4 F.M.B. at 693 (inefficiency caused by “overly refined examination of adequacy and inadequacy of United States-flag services is inconsistent with the purposes and policy of the Act”).
. Board Opinion, J.App. 86-87. The Board first interpreted § 605(c) in this manner in 1971. American Pres. Lines, 12 S.R.R. 168 (1971).
. Brief for Appellant APL at 6 n. 4.
. I do not, of course, express any opinion as to the correctness of the Board’s interpretation of § 605(c) as applicable to this situation. It could be that the Board has good reasons for its conclusion or that in these types of cases the § 605(c) hearing should only concern itself with demand inadequacy and not with inadequacy to meet the statutory purposes. See note 19 supra.
. Maj. Op. text accompanying notes 54-55.
. Initial Decision, J.App. 9, 68 (unsubsidized service already instituted by APL); Board Opinion, J.App. 87 (same). See also 46 C.F.R. §§ 281.11, 281.13 (1976) (Board may authorize unsubsidized voyages by line receiving subsidy for other services).
. Of course, what little we know on this record indicates that there is an effective increase in subsidization. See note 44 supra. Although it seems shameful to relegate APL to inefficiency, the law is clear that a subsidized line cannot be allowed an undue advantage over a competing unsubsidized line unless necessary to achieve adequacy. And in any case, APL would not be told that it must operate its containerized ships less often than is feasible, but only that it appears that United States-flag vessels can adequately compete without subsidy with foreign-flag vessels for the increased demand and that, if it desires to be one of the lines so competing, it must renegotiate its subsidy contract so that its overall subsidization drops somewhat due to the increased efficiency of the ships serving the base demand. If the returns expected from serving the increased demand exceed the lost portion of its subsidy, APL could be expected to renegotiate. If the returns are not so large, APL probably would choose to sail less often than it would be capable of doing. But in any case, the capability of United States-flag ships to carry United States-bound goods would remain at least as high if APL’s request to provide increased service at the same subsidy were not granted. Put another way, APL might choose to make only ten trips per year, but in an emergency its ships could still make fifteen.
The obvious danger is that in a future situation of this kind APL or another line might decide not to adopt á technological change such as containerization at all. This is unlikely both because of the demands of shippers, Board Opinion, J.App. 93, and because the changeover itself received a construction-differential subsidy. See Brief for Appellant APL at 5-6.
. At 110 of 185 U.S.App.D.C., at 775 of 566 F.2d.
. Brief for Appellant APL at 20.
. At 110 of 185 U.S.App.D.C., at 775 of 566 F.2d.
. Thus, if the Board believes that denial of subsidization risks inadequate service on this subroute, it could approve APL’s request. And since the court apparently misconceives my position, Maj. Op. at note 69,1 emphasize that I do not say that 50% is a conclusive magic number. All I say is that whatever the number, and if a number is used, the Board must compute it in accord with the mandate of the statute. And as I noted earlier, it is the Board, not 1, that picked the 50% figure and utilized it in this case. Perhaps, as note 69 of the court’s opinion implies, service on this subroute is inadequate regardless of what is what percentage of what. But that is not what the Board decided, and it is not our office to replace its decisional rationale with our own independent reasoning. See note 19 supra. More importantly, by expanding the Board’s formula for adequacy in order to accommodate the court’s rationale while ostensibly relying on the Board’s reasoning, my colleagues distort the definition of adequacy and seek to justify that approach with the circular assertion that in this case it leads to an appropriate outcome.