delivered the opinion of the Court.
This case originated eleven years ago. As a result of proceedings begun in April, 1930 under the Packers and Stockyards Act, 42 Stat. 159, 7 U. S. C. § 181 et seq., the Secretary of Agriculture in June, 1933, issued an order setting maximum rates to be charged by market agencies for their services at the Kansas City Stockyards. The market agencies brought suit to set aside his order. The district court issued a temporary restraining order, under which amounts charged in excess of the rates fixed by the order were impounded, and later it upheld the order. 8 F. Supp. 766. On appeal here, 7 U. S. C. § 217; 28 U. S. C. §§ 44, 47a, the case was sent back to the district court in order to determine on the issues raised by the pleadings whether the agencies had been denied the “full hearing” demanded by § 310 of the Act. 298 U. S. 468. The district court thereupon decided that this requirement of the statute had been satisfied. 23 F. Supp. 380. The case was again brought here and the order of the Secretary *414was held invalid because of procedural defects. 304 U. S. 1. Prior to this decision, the Secretary and the market agencies had agreed upon a higher schedule of rates to become effective on December 1, 1937. However, under the impounding order, which had continued in effect until that date, over half a million dollars had been deposited. The disposition of this fund was made a ground for a petition for rehearing after the second Morgan decision, but the petition was denied because that question was for the district court. 304 U. S. 23, 26. The Secretary- then reopened the original proceedings to determine reasonable rates during the impounding period. Before the Secretary had made a new order, the district court directed that ■the impounded moneys be turned over to the market agencies. 24 F. Supp. 214. The case came here for the third time, and we reversed the district court and required its retention of the fund “until such time as the Secretary, proceeding with due expedition, shall have entered a final order in the proceedings before him.” 307 U. S. 183,198. This decision was rendered oh May 15, Í939: A month later, the. Secretary issued a new schedule of rates for the impounding period based on elaborate findings. Accordingly, the Government moved the district court to distribute the funds in accordance with the Secretary’s order, but that court, with one of its three judges dissenting, held the order invalid and directed that the funds be given to the market agencies. 32 F. Supp. 546. The case is now here for the fourth time.
The validity of the Secretary’s order has undergone the closest scrutiny in elaborate briefs and extended oral arguments. Nothing has been overlooked. However, in the final stage of this long drawn out litigation, critical examination reveals only a few issues demanding attention.
When the matter was last here we defined the duty of the Secretary. He was to determine reasonable rates for the impounding period so that there could be just dis*415tribution of the funds which the court below had taken into its registry. The nature of the problem before the Secretary was a guide to its solution. The Secretary’s task was not the usual enterprise of fixing rates for the future, so largely an exercise in prophecy. Unique circumstances made him, in 1939, the arbiter of rates for a period between 1933 and 1937. But even such a retrospective determination does not present a mathematical problem. Doubts and difficulties incapable of exact resolution confront judgment. More than that, since the Secretary is the guardian of the public interest in regulating a business of public concern it is not for him merely to reflect the items on a profit and loss statement. He must consider whether these represent services which properly should be charged to the public. While, therefore, the Secretary in determining rates for the past could not deny himself the benefit of hindsight, he was not merely a bookkeeper posting items into a ledger. Rates to which these public agencies were entitled were not to be derived merely from their expenditures and actual income.
This Court defined the duty of the Secretary in its decision in the 307th U. S. The record leaves no doubt that the Secretary, when he filed his order a month, after that decision, appropriately discharged the duty. He served upon the market agencies the order of June 14, 1933, and the findings underlying it as the starting point of the inquiry. The market agencies protested against any order “nunc pro tunc as of June 14,1933,” alleged that conditions had changed much since 1933, and asked for the appointment of an examiner to take new evidence. Because he deemed the earlier findings illuminating and helpful “as a working basis for this hearing,” the Secretary refused to withdraw them. But he appointed an examiner to hear new evidence and denied “any intention of depriving the respondents of the opportunity of offering evidence concerning conditions affecting the reasonableness of their *416rates during the period subsequent to June 14,1933.” He further stated that the “forecasts of conditions” in the 1933 order “can nowise checked in light, of subsequent events.” He neither purported to make nor did he make a nunc pro tune order. The Secretary thus adopted a procedure which admitted whatever light was shed by change of circumstances after' 1933. The market agencies freely availed themselves of this procedure; and the Secretary’s findings leave no room for doubt that his conclusions represent a judgment of 1939 and not a prophecy of 1933. IJaving overruled the contention of Government counsel that evidence of conditions after 1933 was irrelevant, he took note of the fact that fewer livestock came to the market after 1933; that a larger number came by truck, thereby causing á decrease in the number of animals in an average consignment; that specific as well as general economic factors touching the market at Kansas City had changed; that statistics relevant in 1933 had become outmoded; and that he had before him evidence of expenses for “business getting and maintaining” and salesmanship not before him in 1933. • The Secretary thus unequivocally avowed his intention to consider conditions after 1933 and his findings carry out his purpose.1 We must therefore reject the claim that thé Secretary’s judgment was founded on the misconception that he must shut his mind to everything that happened after 1933 and in 1939 fix rates in the imaginary world of 1933.
Another attack upon the Secretary’s order is the con*417ventional objection that the findings were not rooted in proof. To reexamine here with particularity the extensive findings made by the Secretary, and to test them by a record of 1340 printed pages and thousands of pages of additional exhibits, would in itself go a long way to convert a contest before the Secretary into one before the courts. Compare Litchfield v. Register and Receiver,-§ Wall. 575, 578. We have canvassed too fully in the past the duties respectively allotted to the Secretary of Agriculture and the courts in the enforcement of the Packers and Stockyards Act to justify extended discussion of the governing principles. Tagg Bros. & Moorhead v. United States, 280 U. S. 420; Acker v. United States, 298 U. S. 426; see also United States v. Morgan, 307 U. S. 183, 190-91. We are in the legislative realm of fixing rates. This is a task of striking a balance and reaching a judgment on factors beset with doubts and difficulties, uncertainty and speculation. On ultimate analysis the. real question is whether the Secretary or a court should make an appraisal of elements having delusive certainty. Congress has put the responsibility on the Secretary and the Constitution does not deny the assignment.
The objection that the proof does not support the findings is really a repetition in disguise of the unfounded claim that the Secretary misconceived his duty and made his order in 1939 as though he were acting in 1933. The bedrock of these variously phrased attacks upon the order is the. contention that the Secretary was indifferent to events occurring after 1933. The short answer is that he was not. The conclusion which he drew from these events is another matter.2
*418Specifically, it is urged that by the increase of rates for the future, to which the market agencies and the Secretary agreed in 1937, changes in circumstances were recognized, while the present order ignored these changes because its rates are at the same level as the original order. But the Secretary did not disregard changed market conditions during the impounding period. Evidence showing these changes was submitted by the market agencies.3 He was thus duly apprised of the changes and *419they entered into the findings. To be sure, in ascertaining the reasonable rates for the impounding period he did not attach to them the significance which the market agencies drew from them. As a result of an elaborate study of conditions prior to 1933 and evidence indicating no essential changé in those conditions for the purpose at hand during the later years, the Secretary concluded that the market w overstaffed and that in the competitive setting of the Dusiness amounts had been spent not justified by that public interest which he is charged to protect. Actual expenses for salesmen’s salaries and “business getting,” the items chiefly in controversy, he found, did not furnish an adequate guide to the ascertainment of reasonable rates. Had the lower rates originally set by the Secretary in 1933 been tested by experience, audits of the market agencies under these rates would have reflected the practical operation of the policy of lowering costs under controlled conditions. But this source of experience was unavailable because the agencies throughout the impounding period continued to operate under the higher rates. Quite different considerations. may properly have influenced the Secretary in fixing rates for the impounding period from those by which he determined a schedule of rates for the future. The existence of the differences is recognized in the agreement between the Secretary and the market agencies whereby the h'igher.rates of the 1937 schedule were to be “without prejudice” either to the Government or to the agencies *420in the present litigation. It was further agreed in 1937 that after six months, and unless the rate order of 1933 was found invalid, the Secretary could at any time “without further hearing” reduce the rates for the future to the 1933 level. There were very great complexities in determining rates for an industry affected by the unstable conditions which surrounded the Kansas City market in 1937. And the expert tribunal charged with the task may well have felt a need for flexibility in the prophecy involved in setting future rates which did not enter the judgment required in fixing rates for a past period. It is not for us to try to penetrate the precise course of the Secretary’s reasoning. Our duty is at an end when we find, as we do find, that the Secretary was responsibly conscious of conditions at the market during the years following 1933, that he duly weighed them, and nevertheless concluded that rates similar to those in the. 1933 order were proper.
But the market agencies go beyond saying that the record did not warrant what the Secretary found. They say that bias disqualified him. This serious charge derives from a letter written by the Secretary to the New York Times immediately following the decision of this Court in the second Morgan case, 304 IJ. S. 1. By that decision, the Court had upset the order of 1933 because of procedural defects. Largely because of his assumption that this meant the return of the impounded funds to the market agencies, the Secretary in his letter vigorously criticized the decision. The market agencies in due course moved to disqualify the Secretary in the proceedings started by him to fix new rates. In denying their motion the Secretary wrote a patently sincere denial of bias. He stated that he had complained against a return of the impounded funds to the market agencies prior to a determination of the rates on the merits, that the denial of the petition for rehearing, 304 U. S. 23, 26, had shown him the error of his assumption, that in his letter of criticism he *421made no prejudgment about the rates to be fixed, and that his only concern was to “see that the substantive rights of the parties are fairly determined.” He added that “as a matter of expediency” he might have disqualified himself but for the fact that, while the market agencies were pressing his disqualification, they were simultaneously urging that none other than the Secretary had legal authority to make the rate order. Plainly enough, when it was thus suggested that he create a situation in which no order could be made, the Secretary was offered no escape from his duty even had he preferred to consult the comforts of personal convenience.
But, intrinsically, the letter did not require the Secretary’s dignified denial of bias. That he not merely held, but expressed, strong views on matters believed by him to have been in issue, did not unfit him for exercising his duty in subsequent proceedings ordered by this Court. As well might it be argued that the judges below, who had three times heard this case, had disqualifying convictions. In publicly criticizing this Court’s opinion the Secretary merely indulged in a practice familiar in the long history of Anglo-American litigation, whereby unsuccessful litigants and lawyers give vent to their disappointment in tavern or press. ' Cabinet officers charged by Congress with adjudicatory functions are not assumed to be flabby creatures any more than judges are. Both may have an underlying philosophy in approaching a specific case. But both are assumed to be men of conscience and intellectual discipline, capable of judging a particular controversy fairly on the basis of its own circumstances. Nothing in this record disturbs such an assumption.
And so we conclude that the order of the Secretary furnishes “the appropriate basis for action in the district court in making distribution of the fund in its custody.” United States v. Morgan, 307 U. S. 183, 198. But, finally, a matter not touching the validity of the order requires consideration. Over the Government’s objection the dis*422trict court authorized the market agencies to take the deposition of the Secretary. The Secretary thereupon appeared in person at the trial. He was questioned at length regarding the process by which he reached the conclusions of his order, including the manner and extent of his study of the record and his consultation with subordinates. His testimony shows that he dealt with the enormous record in a manner not unlike the practice of judges in similar situations, and that he held various conferences with the examiner who heard the evidence. Much was made of his disregard of a memorandum from one of his officials who, on reading the proposed order, urged considerations favorable to the market agencies. But the short of the business is that the Secretary should never have been subjected to this examination. The proceeding before the Secretary “has a quality resembling that of a judicial proceeding.” Morgan v. United States, 298 U. S. 468, 480. Such an examination of a judge would be destructive of judicial responsibility. We have explicitly held in this very litigation that “it was not the function of the court to probe the mental processes of the Secretary.” 304 U. S. 1,18. Just as a judge cannot be subjected to such a scrutiny, compare Fayerweather v. Ritch, 195 U. S. 276, 306-07, so the integrity of.the administrative process must be equally respected. See Chicago, B. & Q. Ry. Co. v. Babcock, 204 U. S. 585, 593. It will bear repeating that although the administrative process has had a different development and pursues somewhat different ways from those of courts, they are to be deemed collaborative instru-mentalities of justice and the appropriate independence of each should be respected by the other. United States v. Morgan, 307 U. S. 183, 191.
Reversed.
Mr. Justice Reed did not participate in the consideration or decision of this case.Attention is called to the title page of the tentative findings, on which appeared, opposite the docket number of the case and the names of the formal parties, the words “Tentative Findings of Fact, Conclusions and Proposed Order, issued as of June 14, 1933.” This formal caption is not an unnatural description of the starting point of the Secretary’s new inquiry. It clearly is not descriptive of his final findings and order, let alone a denial of the proper theory on which he avowedly proceeded.
That inferences from facts and contentions regarding their significance are the real stuff of these rate determinations is well illus- • trated by the phase of the problem before the Secretary that was most strongly pressed upon us. It is undisputed that since 1933 the arrival of animals by truck has increased, thereby causing a decrease *418in the average number of animals in a consignment. And since the consignment is the unit of cost, a decrease in the number of animals results in an increase in cost per head in the consignment. Hence, formal logic concludes, the present order in setting the same rates as those of 1933 fails to reflect this increase in per head cost, and on that ground is invalid. But both the 1933 and 1939 schedules recognize that there are minimal costs unrelated to the number of animals in a consignment. Both orders, therefore, were graduated according to the number of animals in a consignment. The Secretary found that this graduated scale which “produces an increasing per head revenue as the number of head in the consignment decreases” would “give recognition to the changing method of arrival of livestock.” Moreover, the decrease in the size of consignments may well have been reflected in the increased estimate of salesmanship cost. All these considerations only illustrate that we are moving in a difficult and specialized realm of judgment which has been entrusted to the Secretary of Agriculture and not to the courts. The Secretary’s judgment must prevail since his finding had the support of inferences fairly drawn from the entire evidence, including all that the market agencies saw fit to introduce bearing on their operations after 1933.
An objection to an exclusion of evidence by the examiner requires but slight comment. Two cooperative commission companies had accepted the rates of the Secretary’s' order of 1933, and the market agencies asked that the annual reports of these companies for the impounding period be produced by the division of the Department of Agriculture with which they were filed. The examiner refused to order’ production of the reports on the ground that he had no authority to do so, basing his ruling on a section which the Packers and Stockyards Act incorporates from the Federal Trade Commission Act and which provides that it shall be a misdemeanor for any *419officer of the regulatory agency to make public any information which the agency has obtained “without its authority, unless directed by a court.” 7 U. S. C. § 222, 15 U. S. C. § 50. We need not determine whether the reports should properly have been admitted. If they should have been, the statute provides an orderly way for having this done during the course of the hearing by seeking the Secretary’s authorization. Having failed to pursue the way of the statute, the market agencies were debarred from raising the matter at a later time.