delivered the opinion of the Court.
This action was brought by respondent Vermont dairy farmers, “country” milk producers, seeking a judgment invalidating as contrary to the Agricultural Marketing Agreement Act of 1937, as amended, 50 Stat. 246, 7 U. S. C. §601 et seq. (1964 ed. and Supp. IV), the so-called farm location differential provided for by order *171of the Secretary of Agriculture.1 The effect of that order is to require milk distributors to pay to milk producers situated at certain distances from milk marketing areas, “nearby” farmers, higher prices than are paid to producers located at greater distances from such areas. The District Court issued a preliminary injunction on January 16, 1967, against further payments and on respondents’ motion for summary judgment transformed its decree into a permanent injunction on June 15, 1967. The Court of Appeals for the District of Columbia Circuit affirmed. 131 U. S. App. D. C. 109, 402 F. 2d 660 (1968). We granted certiorari to resolve the important issue of statutory construction involved in this aspect of the administration of the federal milk regulation program. 394 U. S. 958 (1969).
*172I
BACKGROUND
Once again this Court must traverse the labyrinth of the federal milk marketing regulation provisions.2 While previous decisions have outlined the operation of the statute and the pertinent regulations, a brief odyssey through the economic and regulatory background is essential perspective for focusing the issue now before the Court.
A. The Economics of the Milk Industry
The two distinctive and essential phenomena of the milk industry are a basic two-price structure that permits a higher return for the same product, depending on its ultimate use, and the cyclical characteristic of production.
Milk has essentially two end uses: as a fluid staple of daily consumer diet, and as an ingredient in manufactured dairy products such as butter and cheese. Milk used in the consumer market has traditionally commanded a premium price, even though it is of no higher quality than milk used for manufacture. While cost differences account for part of the discrepancy in price, they do not explain the entire gap. At the same time the milk industry is characterized by periods of seasonal overproduction. The winter months are low in yield and *173conversely the summer months are fertile. In order to meet fluid demand which is relatively constant, sufficiently large herds must be maintained to supply winter needs. The result is oversupply in the more fruitful months. The historical tendency prior to regulation was for milk distributors, “handlers,” to take advantage of this surplus to obtain bargains during glut periods. Milk can be obtained from distant sources and handlers can afford to absorb transportation costs and still pay more to outlying farmers whose traditional outlet is the manufacturing market.3 To maintain income farmers increase production and the disequilibrium snowballs.
To protect against market vicissitudes, farmers in the early 1920’s formed cooperatives. These cooperatives were effective in eliminating the self-defeating overproduction by pooling the milk supply and refusing to deal with handlers except on a collective basis.4 During *174the 1920’s era of relative market stability the nearby farmers enjoyed premium prices for their product. These favorable prices were apparently attributable to reduced transportation costs and also the nearby farmer’s historic position as a fluid supplier.5
B. The First Federal Program
The drop in commodity prices during the depression years destroyed the equilibrium of the 1920’s and utter chaos ensued. Congress, in an effort to restore order to the market and boost the purchasing power of farmers, enacted the licensing provisions of the Agricultural Adjustment Act, 48 Stat. 31, 35. Under § 8 (3) the Secretary of Agriculture was empowered
“[t]o issue licenses permitting processors, associations of producers, and others to engage in the handling, in the current of interstate or foreign commerce, of any agricultural commodity or product thereof, or any competing commodity or product thereof. Such licenses shall be subject to such terms and conditions, not in conflict with existing Acts of *175Congress or regulations pursuant thereto, as may be necessary to eliminate unfair practices or charges that prevent or tend to prevent the effectuation of the declared policy and the restoration of normal economic conditions in the marketing of such commodities or products and the financing thereof. The Secretary of Agriculture may suspend or revoke any such license, after due notice and opportunity for hearing, for violations of the terms or conditions thereof. . .
Under the licensing system base-rating plans not unlike the private arrangements that obtained in the 1920’s were adopted.6 Producers were assigned bases which fixed the percent of their output that they would be permitted to sell at the Class I price that was paid for fluid milk.7 The viability of the licensing scheme was jeopardized, however, by judicial decisions disapproving a similarly broad delegation of power under the National Industrial Recovery Act provisions, 48 Stat. 195. Schechter Poultry Corp. v. United States, 295 U. S. 495 (1935). With its agricultural marketing program resting on quicksand, Congress moved swiftly to eliminate the defect of overbroad delegation and to shore up the void in the agricultural marketing provisions. Section 8 (3) of the 1933 Act was amended in 1935 and the pertinent language has been carried forward without significant *176change into § 8c of the present Act. Agricultural Marketing Agreement Act of 1937, 60 Stat. 246, as amended, 7 U. S. C. § 608c (1964 ed. and Supp. IV).8
*177C. The Present Regulatory Scheme
The present system, which differs little in substance from the scheme conceived in 1937 for regulating the Boston market,9 provides for a uniform market price payable to all producers by all handlers.10 Prices are established for Class I and Class II uses. The total volume of milk channeled into the market in each category is multiplied by the appropriate coefficient price and the two results are totaled and then divided by the total number of pounds sold. The result represents the average value of milk sold in the marketing area and is the basic “uniform” price. Were all producers to receive this price they would share on an equal basis *178the profits of Class I marketing and assume equally the costs of disposing of the economic surplus in the Class II market. The actual price to the producer is, however, the “blended” price which is computed by adding and subtracting certain special differentials provided for by statute and order. See 7 CFR § 1001.64 (1969). The deduction for differential payments withheld for the benefit of nearby producers reduces the uniform “blended” price to those producers ineligible to collect this particular adjustment.11 The provision is contained in § 1001.72 of the order and provides:
“In making the payments to producers . . . each handler shall add any applicable farm location differential specified in this section.
“(a) With respect to milk received from a producer whose farm is located within any of the places specified in this paragraph, the differential shall be 46 cents per hundredweight, unless the addition of 46 cents gives a result greater than the Class I price determined under §§ 1001.60, 1001.62, and 1001.63 which is effective at the plant at which the milk is received. In that event there shall be added a rate which will produce that price.”
A differential of 23$ is provided for deliveries from farms in intermediate nearby zones. § 1001.72 (b).
The foregoing provisions appear in the so-called 1964 Massachusetts-Rhode Island Order, which consolidated into one region the four sub-markets which were pre*179viously regulated separately under the so-called four “New England” orders: the 1951 Boston order which carried forward the order adopted for the Boston area in 1937; the Springfield order promulgated in 1949; and the Southeastern New England order of 1958. Each order included a provision for a nearby differential payment to farmers within a stated radius of a designated market center. For example the differential under the Boston order was payable to farmers located within a 40-mile radius of the State House in Boston; a slightly lower differential was paid to farmers within an 80-mile radius. Under the 1964 order there is no central point for the computation of the radius for payment of the differential; the Secretary has retained the differential provisions as they appeared in the previous four orders. Farmers who would have been entitled to the differential under any one of the previous four marketing regulations continue to receive those payments under the present order. These nearby farmers are eligible for the differential on any shipments within the New England marketing area, even though their milk may actually be used outside the radius of their particular nearby zone.
II
THE STATUTORY SCHEME
The foundation of the statutory scheme is to provide uniform prices to all producers in the marketing area, subject only to specifically enumerated adjustments. The question before the Court, stated most simply, is whether payment of farm location differentials, set forth above, is a permissible adjustment under § 8c (5) (B) to the general requirement of uniformity of price.12
*180The Secretary has in the past labeled the “nearby” differential a “location” differential and defended its inclusion in his orders on that ground. The justification and argument are now, however, pitched in a different key. The Government has apparently abandoned all but one of the numerous theories advanced below, and pressed most vigorously in the Blair v. Freeman litigation (125 U. S. App. D. C. 207, 370 P. 2d 229 (1966)), and it now stresses the provision in § 8c (5) (B) for “volume, market, and production differentials customarily applied by the handlers subject to such order.”
While the proper resolution of the issue is by no means self-evident, we are persuaded that “market . . . differentials customarily applied” contemplates cost adjustments. The plain thrust of the federal statute was to remove ruinous and self-defeating competition among *181the producers and permit all farmers to share the benefits of fluid milk profits according to the value of goods produced and services rendered. The Government’s proposed reading of the Act, bottomed, as it is, on the historical payment of a premium to nearby farmers during the monopolistic era of the cooperative pools, would come to perpetuate economic distortion and freeze the milk industry into the competitive structure that prevailed during the 1920’s.
Without the benefit of government muscle to eliminate crippling price warfare in the summer months, neither nearby nor country producers could share in the monopoly-type profits that accrue from fluid milk sales. Absent regulation only the handlers, if anyone, would stand to benefit from the “fluid” monopoly. While we cannot project what would be the case today if a free market prevailed, we might well anticipate that the nearby producers’ winter advantages would be negligible in view of reduced transportation costs and more reliable refrigeration. Thus even in winter handlers might be free to play nearby and outlying farmers against each other since handlers would be free of the leverage exercised by the nearby cooperatives during the 1920’s. Nearby producers now seek the best of both worlds. Having achieved the security that comes with regulation, they seek under a regulatory umbrella to appropriate monopoly profits that were never secure in the unregulated market.
We are reluctant to attribute such intent to Congress and, simply in the name of administrative expertise, to follow a path not marked by the language of the statute. Indeed, such signposts as may be discerned from the legislative history point in a very different direction. The legislative history strongly suggests that “market differentials,” as well as all the other differentials, contemplated particular understood economic adjustments. The House Report, in discussing the allowable adjust*182ments characterizes the market differential as a payment over and above the transportation costs, i. e., a location differential, for delivery to the primary market.13 Thus farmers would share with handlers the savings from bypassing country-station processing and handling the milk only at the city plant.
The significance of the legislative history emerges upon study of the subsequent administrative practice. The original Boston order obscures the market differential payment by providing in place of a labeled adjustment a two-price structure which allowed an additional 180 per cwt. for city-delivered milk over and above the costs of transporting the milk from the country plant. However, the testimony of Mr. Aplin for the Market Administrator erases any doubt that those responsible for administering the Act fully understood the meaning of the Committee’s explanation of market differential.14
*183Subsequent orders have combined the country station handling adjustment, properly the market differential, and the location-transportation differential into the so-called zone differential.15
The statute before us does not contain a mandate phrased in broad and permissive terms. Congress has spoken with particularity and provided specifically enumerated differentials, which negatives the conclusion that it was thinking only in terms of historical considerations. The prefatory discussion in the House Report emphasizes the congressional purpose to confine the boundaries of the Secretary’s delegated authority.16 In these circumstances an administrator does not have “broad dispensing power.” See Addison v. Holly Hill Co., 322 U. S. 607, 617 (1944). The congressional purpose is further illumined by the character of the other statutory differentials for “volume,” *184“grade or quality,” “location,” and “production,”17 all of which compensate or reward the producer for providing an economic service of benefit to the handler.18
The general language of the committee report indicating that Congress intended to carry forward the basic regulatory approach adopted under the 1933 Act, following the precedent of the 1920’s, is stressed by the dissent to this opinion. This committee language, it is argued, reinforces the continuity connotations of the “customarily applied” language, a thrust that is not blunted *185by any specific language indicating a legislative purpose to treat all farmers equally.
Legislative silence is a poor beacon to follow in discerning the proper statutory route. For here the light illumines two different roads. If nearby payments had the notoriety and significance in the milk distribution industry attributed to them by the dissent, Congress could have given its blessing by carving out another specific exception to the uniform price requirement. In an Act whose very purpose was to avoid the infirmity of overbroad delegation and to set forth with particularity the details for a comprehensive regulatory scheme, it would have been a simple matter to include in a list of enumerated differentials, “nearby” payments, or at least allude to them in the report of the draftsmen. It is clear that Congress was not conferring untrammeled discretion on the Secretary and authorizing him to proceed in a vacuum. This was the very evil condemned by the courts that the 1935 amendments sought to eradicate.19 It would be perverse to assume that congressional drafters, in eliminating ambiguity from the old Act,20 were careless in listing their exceptions and selecting the illustrations from the committee report from which their words would ultimately derive content.21
*186We consider our conclusions in no way undermined by the colloquy on the floor between Senator Copeland and Senator Murphy upon which the dissent places such emphasis. A committee report represents the considered and collective understanding of those Congressmen involved in drafting and studying proposed legislation. Floor debates reflect at best the understanding of individual Congressmen. It would take extensive and thoughtful debate to detract from the plain thrust of a committee report in this instance. There is no indication, however, that the question of nearby differentials and the meaning of “market . . . differentials customarily applied” were precisely considered in the floor dialogue. The exchange is not only brief but also inconclusive as to meaning.22 Indeed, Senator Murphy apparently acqui*187esced in Senator Copeland’s implied criticism of the statute for providing uniform prices for distant and nearby producers within the marketing region. When Senator Copeland pursued his inquiry, asking whether the Act recognized the higher cost for taxes on nearby lands, Senator Murphy merely recited the differential provisions of the Act and suggested that they “adopt the present practice of business,” but conspicuously lacking is an affirmative statement that any specific differential covered these costs. This is not impressive legislative history especially in light of Senator Murphy’s earlier agreement with Senator Copeland’s statement that “[t]he provisions of the equalization . . . provide that a producer who is producing his milk on farms near to cities would receive the same price for his product as a farmer who produces his milk, say, 40 or 50 miles away from the same community,” and the specific business illustrations of the House Report.
*188Ill
SCOPE OF MARKET DIFFERENTIAL
While market differentials customarily applied need not be restricted to the sole illustration in the House Report, that illustration, taken in conjunction with the discussion of all the statutory differentials, suggests that the permissible adjustments are limited to compensation for rendering an economic service.23 The challenged nearby differentials do not fall into this category.24
Nor has the Secretary advanced any economic justification for these differential payments. It is plain from the administrative record that the nearby differential was included in the original Boston order as a recognition of the favored position of nearby producers in the fluid market and as an inducement to nearby farmers to approve the Secretary’s order. (J. A. 237.25) The only sense *189in which the handler may be said to gain economically is by virtue of the elimination of the nearby producer as a potential competitor. While this factor is mentioned in the findings accompanying the 1937 order, it has not *190been emphasized in the 1964 findings and the testimony at the 1963 hearings suggests that support in the record is indeed scant. That entry of the nearbys into the distribution market would bring unwanted competition, is irrelevant if it does not jeopardize market stability. We think the analysis of the court below was correct: if there is any economic benefit here, producers should receive their compensation directly from the handlers and not out of the marketwide pool. 131 U. S. App. D. C., at 114, 402 F. 2d, at 665.
While petitioner nearby farmers do not concede so readily the absence of economic foundation for the differential, no justifications are advanced that find any substantial support in the record. The allusion to the evenness of production on nearby farms would not justify the exclusive payment of this differential to nearby farmers. If the Secretary intended a production differential, all producers who qualify would be eligible. Some amici and petitioners point to higher taxes on nearby lands and opportunity costs as reason for retaining the differential. These are, admittedly, additional costs of nearby production, but they are of no concern to handlers who seek only to obtain reliably milk at the cheapest price. See Kessel, Economic Effects of Federal Regulation of Milk Markets, 10 J. Law & Econ. 51 (1967). This Court has been slow to attribute to Congress an intent to compensate for inefficient allocation of economic resources. Cf. West Ohio Gas Co. v. Comm’n, 294 U. S. 63, 72 (1935). While petitioners argue that the differential is a necessary inducement to keep the nearby farmers in business, the record does not reveal that the Secretary acted out of concern that the nearby farmers would quit the market, nor is there any evidence demonstrating the present necessity for nearby producers. In an era where efficient transportation is *191available this may be of nominal concern. At most this may have been an unspoken consideration in 1937.26
Since the Secretary made no findings to that effect, the Court need not consider whether they would justify payment of the nearby differential in view of the legislative history indicating that the statute contemplates adjustments primarily for economic costs to handlers that are absorbed or reduced by the producers. Further if the representations of respondents are correct — and they are not without support in the record — it appears that the elimination of the 40-mile zone nearby differential payments of 460, even with the suspension of the intermediate differential payments of 230, would result in a higher uniform price to those farmers now receiving the 230 differential.27
IV
PRIOR DECISIONS
Our holding does not represent a departure from this Court’s precedents. No opinion of this Court has ever explicitly approved the nearby differential. Reliance on United States v. Rock Royal Co-op., 307 U. S. 533 (1939), is misplaced. This Court’s refusal to invalidate the payment of a nearby differential to farmers in certain counties named in the New York order must be taken in the context of that action which was initiated by the Government against handlers who refused to obey the regulations. That decision did not repudiate the District Court’s finding that the provision was “discriminatory as between producers.” Id., at 567. The narrow reach of our Rock Royal holding was recognized in Stark v. *192Wickard, 321 U. S. 288 (1944), where we noted that Rock Royal held the handlers without standing "to object to the operation of the producer settlement fund,” id., at 308, except as it affected handlers. The Court in Rock Royal went on to reject Rock Royal’s contention that the payments placed those handlers without customers in the nearby counties at a competitive disadvantage.
Our attention is also drawn to the First Circuit’s decision in Green Valley Creamery v. United States, 108 F. 2d 342 (1939). As in Rock Royal, supra, the parties did not have standing to raise the invalidity of the nearby differential. To the extent the First Circuit’s view is contrary to our present holding, we disapprove it.
Y
SIGNIFICANCE OF DEPARTMENTAL CONSTRUCTION
While this Court has announced that it will accord great weight to a departmental construction of its own enabling legislation, especially a contemporaneous construction, see Udall v. Tollman, 380 U. S. 1, 16 (1965); Power Reactor Co. v. Electricians, 367 U. S. 396, 408 (1961); it is only one input in the interpretational equation. Its impact carries most weight when the administrators participated in drafting and directly made known their views to Congress in committee hearings. See Power Reactor Co. v. Electricians, supra; United States v. American Trucking Assns., 310 U. S. 534, 539 (1940). In such circumstances, absent any indication that Congress differed with the responsible department, a court should resolve any ambiguity in favor of the administrative construction, if such construction enhances the general purposes and policies underlying the legislation. *193See American Power & Light Co. v. SEC, 329 U. S. 90, 112-114 (1946).
The Court may not, however, abdicate its ultimate responsibility to construe the language employed by' Congress. Those props that serve to support a disputable administrative construction are absent here. There is no suggestion in the findings, nor have the parties explained, how the present differential contributes to the broad, general purpose of eliminating crippling competition. Nor in the present case has the Court’s attention been drawn to any hearings that suggest that Congress acted with the particular administrative construction before it in either 1935 or 1937. And if those administrators who participated in drafting the 1935 Act understood market differentials to encompass the farm location differential, they obviously failed to communicate their understanding to the drafters of the committee report. It is also evident that the 1937 re-enactment of the 1935 amendments was routine and did not follow a comprehensive review of the issues that had been explored in detail by the 1935 draftsmen who wrote the committee reports.28
It is true that a report from the Federal Trade Commission set forth the computations employed under the 1936 Boston order which apparently provided for a *194nearby differential.29 But the stark figures, set forth in the appendix to the report without explication, can hardly be said to have given the administrative construction the “notoriety” that this Court found persuasive in Udall v. Tollman, 380 U. S., at 18. In Udall the Court was impressed by the fact that the Secretary’s interpretation had “been a matter of public record and discussion.” Id,., at 17. Even despite active congressional involvement in reviewing certain administrative action in connection with particular leases, the Court noted that it would not attribute ratification to Congress. Udall v. Tallman, supra. Nor can petitioners put flesh on this argument by citing § 4 of the 1937 re-enactment, 50 Stat. 249,30 and the committee report, H. R. Rep. No. 468, 75th Cong., 1st Sess., 4 (1937), which merely states in the language of the Act that § 4 purports to ratify, legalize, and confirm all action taken pursuant to the agreement and order provisions under the 1935 statute.31
*195VI
RELEVANCE OF PRODUCER APPROVAL
Petitioners allude to the fact that the orders in question have been specifically approved by the farmers concerned as required by §§ 8c (9) (B) (i) and (ii) of the Act.32 While the contention is adumbrated, the argument appears to run as follows: since provision is made for approval of orders by the regulated subjects, the Secretary’s discretion should be generously inter*196preted. If provision for such approval could ever legitimize a regulation not authorized by statute, the provision has no significance in the case before us, in light of the considerations already discussed. It is the Secretary, not the farmers, who is responsible for administering the statute and initiating orders.33
YII
PROPRIETY OP SUMMARY JUDGMENT
Although the Secretary does not press the point, the private petitioners argue that this Court should at the very least reverse for a trial on the merits or alternatively reverse with instructions to remand to the Secretary for further consideration.
This is not a case where a department has acted without a formal record. In such instances a trial might be appropriate to afford the department an opportunity to develop those facts which underpin its action. When action is taken on a record the department cannot then present testimony in court to remedy the gaps in the record, any more than arguments of counsel on review can substitute for an agency’s failure to make findings or give reasons. A remand to the Secretary is inappropriate in the absence of a request by the Government. Counsel for the Department has advanced no new theory for sustaining the order. Cf. SEC v. Chenery Corp., 318 U. S. 80, 92 (1943).
Unlike Addison v. Holly Hill Co., 322 U. S. 607 (1944), we do not have before us a definition in a regulation that is necessary to give meaning and content to the admin*197istrative scheme. Nor does our decision have the effect of engrafting a definition on a particular statutory term, a function that should, in the first instance, be left to the appropriate administrative body. The 1964 order, moreover, expressly provides for severance of any provision that is found invalid. See 7 CFR § 1001.96.
VIII
DISPOSITION OF THE ESCROW FUND
Petitioner farmers’ last line of retreat is their contention that they are entitled to escrow monies that have been accruing since the District Court’s entry of the order granting the respondents’ motion for a preliminary injunction. The court below struck an equitable balance in awarding to petitioners, nearby farmers, all escrow monies collected prior to the entry of final judgment by the District Court. This is a fair solution, and one this Court will not disturb. Petitioners have been on notice since Blair v. Freeman, 125 U. S. App. D. C. 207, 370 F. 2d 229 (1966), that nearby differentials were bottomed on a shaky statutory premise. Lest losing parties be encouraged to prolong litigation by frivolous appeals in order to reap a windfall, we think respondents deserve the fruits of their victory as of the date of final judgment at trial.
The judgment below is
Affirmed.
The Chief Justice and Mr. Justice Marshall took no part in the consideration or decision of these cases.The Secretary has promulgated comprehensive regulations to govern the marketing of milk, 7 CFR § 1002.1 et seq. (1969), pursuant to the Agricultural Marketing Agreement Act. The provisions relevant to this cause are set forth in Part I of this opinion, at 178, infra.
The action was originally brought against the Secretary only. Petitioners Zuber et al., nearby farmers, unsuccessfully sought leave to intervene before the District Court in support of the Secretary’s regulations. When judgment was rendered against the Secretary, petitioners sought leave to intervene for the purposes of appeal. Leave was granted and the Secretary also decided to take an appeal. The parties have devoted a good deal of energy to disputing what constitutes the record in this litigation. Petitioners at various times have referred us to the testimony and record compiled in an action brought in the Northern District of New York, Cranston v. Freeman, 290 F. Supp. 785 (1968). Respondents have objected, noting that the record in Cranston is not formally before this Court, and have included in the appendix various materials that were not of record below. The Court need not pause over the controversy since none of the materials in respondents’ appendix is decisive of the action before us. As for the references to the Cranston record, they too are not decisive of the dispute.
See, e. g., Lehigh Valley Cooperative v. United States, 370 U. S. 76 (1962); Brannan v. Stark, 342 U. S. 451 (1952); Stark v. Wickard, 321 U. S. 288 (1944); United States v. Rock Royal Co-op., 307 U. S. 533 (1939); H. P. Hood & Sons v. United States, 307 U. S. 588 (1939). Tlie lower courts have also been plagued by the milk problem. See especially Judge Frank’s lament, Queensboro Farm Prods. v. Wickard, 137 F. 2d 969 (C. A. 2d Cir. 1943); see also Blair v. Freeman, 125 U. S. App. D. C. 207, 370 F. 2d 229 (1966); Green Valley Creamery v. United States, 108 F. 2d 342 (C. A. 1st Cir. 1939).
For fluid use, milk must be transported in its natural state and as such is a bulky and highly perishable commodity. Thus cost of shipment to a consumer market is greater than transporting an equal supply to a manufacturing plant. These factors, combined with more rigid sanitary requirements for plants distributing the fluid product, see Agricultural Adjustment Administration Report, May 1933-Feb. 1934, p. 154, explain part of the disparity between the price for Class I (fluid milk) and Class II (other uses) milk. Nearby producers, given equilibrium of supply and demand, are logical fluid suppliers to the urban areas. See generally J. Cassels, A Study of Fluid Milk Prices (1937).
The cooperative system amounted to a pooling arrangement wherein participating producers would bargain collectively with the handlers and threaten to withhold their milk if the handlers refused to agree to purchase a certain minimum percentage of their Class I fluid milk from the pool. Without this supply the handlers would be unable to meet their winter requirements.
Essential to this arrangement of course was a sufficiently wide membership to insure no alternative source of supply to recalcitrant handlers.
The second aspect of the arrangement was the division of the profits among the producer members of the cooperative. Frequently *174employed was a base-rating plan whereby each producer would be assigned a percentage of his milk for which he could claim payment at the Class I fluid price. For the remaining production he would be paid at the Class II rate. Apparently bases were assigned according to the anticipated participation of the producer in the fluid market. As a result, nearby producers received more favorable bases in view of their historical role as fluid suppliers in an equilibrium market. For descriptions of the cooperative systems see Cassels, swpra, n. 3, at 56-70; J. Black, The Dairy Industry and the AAA 49-51 (1935).
Because they were historically fluid suppliers the nearby producers apparently maintained at all times production sufficient to service the consumer fluid market. In addition their close proximity enabled them to deliver to small retailers. As such they were potential competitors.
See Agricultural Adjustment Administration Report, supra, n. 3, at 159-161; G. Barnhart, The Development of the Licenses and Order Regulating the Handling of Milk in the Greater Boston, Massachusetts, Marketing Area, Nov. 3, 1933-June 1, 1946 (unpublished dissertation on file with Department of Agriculture and Harvard University).
License 38 for the Boston area provided more favorable bases for the nearby producers. See Barnhart, supra, n. 6, at 95-96.
“(5) Milk and its products; terms and conditions of orders.
“In the case of milk and its products, orders issued pursuant to this section shall contain one or more of the following terms and conditions, and (except as provided in subsection (7) of this section) no others:
“(A) Classifying milk in accordance with the form in which or the purpose for which it is used, and fixing, or providing a method for fixing, minimum prices for each such use classification which all handlers shall pay, and the time when payments shall be made, for milk purchased from producers or associations of producers. Such prices shall be uniform as to all handlers, subject only to adjustments for (1) volume, market, and production differentials customarily applied by the handlers subject to such order, (2) the grade or quality of the milk purchased, and (3) the locations at which delivery of such milk, or any use classification thereof, is made to such handlers:
“(B) Providing:
“(i) for the payment to all producers and associations of producers delivering milk to the same handler of uniform prices for all milk delivered by them: Provided, That except in the case of orders covering milk products only, such provision is approved or favored by at least three-fourths of the producers who, during a representative period determined by the Secretary of Agriculture, have been engaged in the production for market of milk covered in such order or by producers who, during such representative period, have produced at least three-fourths of the volume of such milk produced for market during such period; the approval required hereunder shall be separate and apart from any other approval or disapproval provided for by this section; or
“(ii) for the payment to all producers and associations of producers delivering milk to all handlers of uniform prices for all milk so delivered, irrespective of the uses made of such milk by the individual handler to whom it is delivered;
“subject, in either case, only to adjustments for (a) volume, market, and production differentials customarily applied by the handlers subject to such order, (b) the grade or quality of the milk delivered, (c) the locations at which delivery of such milk is made, and (d) a further adjustment, equitably to apportion the total value of the *177milk purchased by any handler, or by all handlers, among producers and associations of producers, on the basis of their marketings of milk, which may be adjusted to reflect sales of such milk by any handler or by all handlers in any use classification or classifications, during a representative period of time which need not be limited to one year.”
The Boston order of 1937, 2 Fed. Reg. 1331, established uniform prices for all producers at $3.19 and $3.01 per cwt. of milk, depending on the place of delivery, with a further adjustment for transportation to the handler's plant in the marketing area. Article VIII, §4 (1) also provided for an adjustment based on the cost of transporting milk from outlying plants to the primary Boston market. The present regulations calculate price with reference to the purchasing power of milk based on the 1958 cost-of-living index. No transportation adjustment is provided for in calculation of the uniform price under § 1001.62 of the order. Differentials to compensate for zone of delivery are retained as separate adjustments. See infra.
10 The Secretary has three alternative modes of proceeding under the Act. He may establish “use” prices which all handlers must pay to all producers according to the actual amount of milk used in each category, § 8c (5) (A); individual handler pools where all producers or cooperatives selling to an individual handler shall be paid a uniform price for milk delivered to that handler; or a marketwide pool where all handlers must pay all producers a uniform price for all milk delivered irrespective of end use.
Also included is an adjustment for delivery to a nearby plant. The location of handler plants is classified by zones. 7 CFR § 1001.62. Delivery to a plant located nearby the consumer market is, of course, advantageous to the handler and the producer is compensated for this service. The handler also saves the cost of handling and processing at his country plant in addition to saving transportation cost. Conversely, depositing milk at handlers’ plants in outlying districts results in a negative adjustment.
Section 8c (5) (B) (ii) requires all uniform prices to be paid “irrespective of the uses made of such milk by the individual handler to whom it is delivered.” Respondents contend that the nearby differ*180ential is merely a disguised payment for the nearby suppliers’ greater share of fluid milk sales. Such was apparently the case in the New Jersey order invalidated by the Court of Appeals in Blair v. Freeman, supra, where the payment of the differential was explicitly linked to the percentage of nearby milk actually supplied to the fluid market. We share respondents’ skepticism and our doubts are reinforced by the explicit connection of differential payments with the share of fluid milk supplied in the 1936 Boston order. Further cause for skepticism is found in the present zone differential structure which undercompensates the handlers for transportation from outlying districts and thus encourages them to buy from nearby farmers. See Kessel, Economic Effects of Federal Regulation of Milk Markets, 10 J. Law & Econ. 51, 64 — 65 (1967). Here, however, unlike the situation in Blair v. Freeman, supra, the producer receives the differential irrespective of the use to which his milk is ultimately put. Since the nearby differential in the present order is not directly tied to the percentage of fluid milk sales, although the order limits differential payments to 460 or the Class I price, whichever is higher, we accept the Government’s contention that, as a matter of strict logic, the payment of differentials based on the historical position of nearby producers as fluid suppliers, is not inconsistent with the irrespective-of-end-use requirement.
“The market differential is a differential which is given to the producer to compensate him for delivering his mill!; to a city market instead of to a country plant. These differentials vary with the markets and cannot be qualified as a ‘location’ differential, because of the fact that location is usually determined on the distance from a primary market whereas market differentials are usually paid in secondary markets.” H. R. Rep. No. 1241, 74th Cong., 1st Sess., 10 (1935).
The relevant excerpts from the hearing are included in the Joint Appendix and appear at 258-259:
“Section 4 . . . provides for location differentials. . . . Now, the price which is arrived at from the calculation of the pool is a blended price for all milk f. o. b. the market with country station allowances deducted. Now, Paragraph 1 [of § 4] here provides that there shall be deducted from that blended price in the case of milk delivered to a plant more than 40 miles from the State House an amount equal to the carlot freight rate from that plant to Boston, so that that deduction would be different for each freight zone, and the price would be smaller by the amount of difference in freight from each zone as we go out from the market. Now, Paragraph 2 [of § 4] provides that in the case of milk delivered from a producer *183to a plant located within forty miles of the State House there should be added 18 cents per hundredweight. That,is added for the reason that in the ease of country stations there is allowed the dealers on Class I milk 20 cents a hundredweight as a country station charge, and we are allowing for containers in which to ship the milk three cents in the case of milk received at city plants, instead of having a 20 cent and a three cent deduction, which would be 23 cents. There is a receiving station allowance of only five cents. The difference is 18 cents per hundredweight. We add back in here 18 cents to the producer whose milk does not pass through a country station.” (Emphasis added.)
See Barnhart, supra, n. 6, at 620.
“To eliminate questions of improper delegation of legislative authority raised by the decisions in Schechter et al. v. United States, the provisions relating to orders enumerate the commodities to which orders issued by the Secretary of Agriculture may be applicable, prescribe fully the administrative procedure to be followed by the Secretary in issuing, enforcing, and terminating orders, and specify the terms which may be included in orders dealing with the enumerated commodities.” H. R. Rep. No. 1241, supra, at 8. See Brannan v. Stark, 342 U. S. 451, 465 (1952).
In this connection it should be noted that the production differential authorized for maintaining an adequate supply for fluid use during the lean winter months is not, strictly speaking, a handler cost but a general cost of the market. It is, however, an essential cost that cannot be eliminated by looking to an alternative supplier. Viewed in this context, it is of course a cost to the handler; for in a nonregulated equilibrium market, a handler would be forced to pay a premium during the winter months when supply is limited and demand constant.
“The volume differential is a differential which is paid when the operations of several country plants are consolidated into one plant. The inconvenience which is caused to producers by closing up plants to which they have been delivering and requiring that all of their milk be handled by one plant, is compensated by an additional payment to the producers. The production differential is the differential which is paid to a producer, compensating him for keeping his farm and milk qualified for a city market even though his milk may actually be going into manufactured use. . . . The production differential is a payment to the farmer for performing this junction in the market.” (Emphasis supplied.) H. It. Rep. No. 1241, supra, at 9-10.
In Brannan v. Stark, supra, this Court invalidated regulations providing certain payments to cooperatives that had the effect of reducing the blended price to nonmember producers. The premise underlying our holding was that these payments would have to represent compensation for rendering of economic services of benefit to all producers. Even the dissenters took as a point of departure the proposition that the payments could be sustained only if justified in terms of services rendered.
See Brannon v. Stark, supra.
“The proposed amendments, insofar as they relate to marketing agreements and orders, are primarily intended to implement and spell out in more detail and with greater freedom from ambiguity the powers which were provided for in the original act. The present language of the statute is, unfortunately, subject to serious misconstruction. This has given rise to obstacles in connection with the enforcement of the marketing agreements and licenses which have seriously endangered their successful operation.” H. R. Rep. No. 1241, supra, at 7.
The verdict of quiescent years cannot be invoked to baptize a statutory gloss that is otherwise impermissible. This Court has many times reconsidered statutory constructions that have been passively abided by Congress. Congressional inaction frequently *186betokens unawareness, preoccupation, or paralysis. “It is at best treacherous to find in congressional silence alone the adoption of a controlling rule of law.” Girouard v. United States, 328 U. S. 61, 69 (1946). Its significance is greatest when the area is one of traditional year-by-year supervision, like tax, where watchdog committees are considering and revising the statutory scheme. Even less deference is due silence in the wake of unsuccessful attempts to eliminate an offending interpretation by amendment. See, e. g., Girouard v. United States, supra. Where, as in the ease before us, there is no indication that a subsequent Congress has addressed itself to the particular problem, we are unpersuaded that silence is tantamount to acquiescence, let alone the approval discerned by the dissent.
The floor exchange is reported at 79 Cong. Rec. 11139-11140.
“Mr. COPELAND. What has the Senator to say to the suggestion that in a number of communities in up-State New York there is not a sufficient supply of milk surrounding the market to take care of the demand; therefore, milk must be brought into the market from more distant points ? The provisions of the equalization which we are now discussing provide that a producer who is producing his milk on farms near to cities would receive the same price for his product as a farmer who produces his milk, say, 40 or 50 miles away from the same community.
“Mr. MURPHY. If they were embraced in the same marketing area, that would be true. Let us keep in mind what the situation is. There is a deficiency of consumer demand. There is a surplus of *187milk. The price is greatly depressed, and has been for 5 years. The only way in which one can determine how each one of the producers included in the plan provided here shall bear his share of the cost of effecting a higher price is to divide the milk by classification uses.
“Mr. COPELAND. I do not think the Senator has quite stated all the conditions. He does not take into consideration the difference in the cost of production. Taxes and values of property near the city are very much higher than in the case of property farther away from the city. The transportation differential does not compensate for the difference in cost, as I see it.
“Mr. MURPHY. If the Senator will refer to page 12, line 13, he will see that there is this qualification:
“ 'Such prices shall be uniform as to all handlers, subject only to adjustments for (1) volume, market, and production differentials customarily applied by the handlers subject to such order — ’
“They adopt the present practice of business—
“'(2) the grade or quality of the milk purchased, and (3) the locations at which delivery of such milk, or any use classification thereof, is made to such handlers.’ ”
The market differential does not, strictly speaking, compensate the producer for absorbing a cost to the handler for it may be no additional cost to the producer to deliver to a city plant. A nearby farmer, for example, would not incur additional costs by delivering to a preferred city plant as opposed to a country station. The savings to the handler are nevertheless plain and the market differential should properly be viewed as an adjustment that permits the producer to share in the handler’s profits resulting from reduced costs.
See Kessel, swpra, n. 12,-at 65-66 (1967). After criticizing the present, undercompensation for transportation costs from faraway zones as a disguised subsidy to nearby producers, resulting in an inefficient allocation of economic resources, the author draws a comparison with the nearby differential lamenting, “However weak the case for zone differentials that fail to depict transportation costs, it is infinitely stronger than the case for location differentials.”
The Secretary’s 1964 findings include the following: “The farm location differential provisions under the present New England orders should be continued under the Massachusetts-Rhode Island order and the Connecticut order.
“A group of nine cooperative associations, which represents principally producers whose farms are located outside any of the sped-*189fled farm location differential areas, proposed that farm location differentials be eliminated under the New England orders. Three other cooperative associations proposed that a producer whose farm is located within New England and who is presently eligible to receive a farm location differential . . . under any New England order be eligible to receive the same differential irrespective of the New England order under which his milk is pooled. Another cooperative association proposed that the farm location differentials be increased ....
“. . . [F]arm location differentials have been in effect under the several New England orders since the inception of the orders. The differentials were adopted to reflect in the pricing structure of the orders historical price relationships by location which prevailed in these markets. It was found that customarily somewhat higher values, above those which normally reflected transportation costs, attached to milk produced near the principal consumption centers as compared to the market value of milk produced in the more distant areas of the milkshed.
“While considerable testimony in support of removal of the provisions was received, it was not established that the farm location differential provisions are resulting in unstable or disruptive marketing conditions which warrant their deletion from the orders at this time. Although certain marketing problems in the nearby and intermediate market areas were referred to in the testimony, these problems are not the result of production increases on farms in these areas which logically might be attributable to the higher returns to producers in these areas. Such increases have not been significantly different from those on farms not eligible for the farm location differentials.” (J. A. 349-351).
There is no reason to dispute the Secretary’s finding that the differentials have no disruptive effect on the market. The issue, however, is whether the provisions are authorized by statute. The Secretary’s order is devoid of any economic justification and relies solely on the historical factor of the nearby producer’s favorable share of the fluid use market. See also Report to the Secretary of Agriculture by the Federal Milk Order Study Committee 74A75 (1962).
See Report to the Secretary of Agriculture by the Federal Milk Order Study Committee, supra, n. 25, at 75.
See J. A. 455 reporting excerpts from the Secretary’s decision of October 21, 1958, accompanying the order for the Southeastern New England marketing area.
Judge Frank expressed the view in Queensboro Farm Prods. v. Wickard, 137 F. 2d 969 (C. A. 2d Cir. 1943), that Congress intended to adopt the intervening administrative interpretation of the “use” language of § 8c (5) (A) by its 1937 re-enactment. The construction of the “use” provision may well have caused more concern than the interpretation of the § 8c (5) (B) differentials. In any event, Judge Frank’s assumption that Congress gave “careful consideration . . . in connection with a re-enactment,” 137 F. 2d, at 977, is not supported by citation to specific legislative history that would indicate that Congress had in mind specific problems in connection with the administration of the marketing provisions.
The 1936 order provided for payment of a uniform price subject to adjustments and with a special exception for “any producer, whose farm is located within forty (40) miles of the State House in Boston and who delivers milk to such handler at a plant located within forty (40) miles of the State House in Boston, at $3.30 per hundredweight for that quantity of milk delivered by such producer not in excess of the base of such producer.” (Emphasis supplied.) Art. VIII, §1 (2).
Section 4 of the Act provided:
“Nothing in this Act shall be construed as invalidating any marketing agreement, license, or order, or any regulation relating to, or any provision of, or any act of the Secretary of Agriculture in connection with, any such agreement, license, or order which has been executed, issued, approved, or done under the Agricultural Adjustment Act, or any amendment thereof, but such marketing agreements, licenses, orders, regulations, provisions, and acts are hereby expressly ratified, legalized, and confirmed.”
To the extent that Congress could be said to have acted against the background of the 1936 order, the Court must reject petitioners’ argument. The 1936 order was superseded by the 1937 order which differed in approach. The provision for nearby differentials in the *1951936 order was obscured by allowing a more favorable total price to nearby producers. See n. 29, supra. The 210 differential incorporated in the 1937 order for the benefit of intermediate nearby zones was not included in the 1936 order. The 210 differential provided in Art. VIII, § 4 (2), of the 1936 order could have been viewed as a true market differential since its payment depended on delivery to a handler within a 40-mile zone from a producer beyond a 40-mile zone. Further, as noted by the court below, § 4 is typical of statutory boilerplate traditionally included in legislative re-enactments, to avoid breaks in regulatory continuity. 131 U. S. App. D. C., at 119, 402 F. 2d, at 670.
Section 8c (9) of the Act, 7 U. S. C. § 608c (9), provides that no order shall become effective until the Secretary determines:
“(B) That the issuance of such order is the only practical means of advancing the interests of the producers of such commodity pursuant to the declared policy, and is approved or favored:
“(i) By at least two-thirds of the producers . . . who, during a representative period determined by the Secretary, have been engaged, within the production area specified in such marketing agreement or order, in the production for market of the commodity specified therein, or who, during such representative period, have been engaged in the production of such commodity for sale in the marketing area specified in such marketing agreement, or order, or
“(ii) By producers who, during such representative period, have produced for market at least two-thirds of the volume of such commodity produced for market within the production area specified in such marketing agreement or order, or who, during such representative period, have produced at least two-thirds of the volume of such commodity sold within the marketing area specified in such marketing agreement or order.”
Lower courts have, in some circumstances, permitted an agency to rely on the approval of those affected by an action as evidence that the action is in the “public interest.” Compare Citizens for Allegan County v. FPC, 134 U. S. App. D. C. 229, 414 F. 2d 1125 (1969), with Marine Space Enclosures v. FMC, 137 U. S. App. D. C. 9, 420 F. 2d 577 (1969). We need not consider what scope, if any, may be given to these principles.
S. Rep. No. 1011, ,74th Cong., 1st Sess., 9; H. R. Rep. No. 1241, 74th Cong., 1st Sess., 9. This basic purpose is reflected in the fact that Congress provided the Secretary with three different schemes of regulation, each of which followed a variety of regulations used by the milk cooperatives. See n. 10, ante, at 177. In 1965 Congress followed this same basic purpose when it amended the Act to make explicit the Secretary’s power to employ base-rating plans, described in n. 4, ante, at 173-174. See 79 Stat. 1187, 7 U. S. C. § 608c (5) (B), cl. (d) (1964 ed., Supp. IV).