Universal Camera Corp. v. National Labor Relations Board

Mr. Justice Frankfurter delivered the opinion of the Court. The essential issue raised by this case and its companion, Labor Board v. Pittsburgh Steamship Co., post, p. 498, is the effect of the Administrative Procedure Act and the legislation colloquially known as the Taft-Hartley Act on the duty of Courts of Appeals when called upon to review orders of the National Labor Relations Board. The Court of Appeals for the Second Circuit granted enforcement of an order directing, in the main, that petitioner reinstate with back pay an employee found to have been discharged because he gave testimony under the Wagner Act and cease and desist from discriminating against any employee who files charges or gives testimony under that Act. The court below, Judge Swan dissenting, decreed full enforcement of the order. 179 F. 2d 749. Because the views of that court regarding the effect of the new legislation on the relation between the Board and the Courts of Appeals in the enforcement of the Board’s orders conflicted with those of the Court of Appeals for the Sixth Circuit 1 we brought both cases here. 339 U. S. 951 and 339 U. S. 962. The clash of opinion obviously required settlement by this Court. *477 I. Want of certainty in judicial review of Labor Board decisions partly reflects the intractability of any formula to furnish definiteness of content for all the impalpable factors involved in judicial review. But in part doubts as to the nature of the reviewing power and uncertainties in its application derive from history, and to that extent an elucidation of this history may clear them away. The Wagner Act provided: “The findings of the Board as to the facts, if supported by evidence, shall be conclusive.” Act of July 5, 1935, § 10 (e), 49 Stat. 449, 454, 29 U. S. C. § 160 (e). This Court read “evidence” to mean “substantial evidence,” Washington, V. & M. Coach Co. v. Labor Board, 301 U. S. 142, and we said that “[substantial evidence is more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Consolidated Edison Co. v. Labor Board, 305 U. S. 197, 229. Accordingly, it “must do more than create a suspicion of the existence of the fact to be established. ... it must be enough to justify, if the trial were to a jury, a refusal to direct a verdict when the conclusion sought to be drawn from it is one of fact for the jury.” Labor Board v. Columbian Enameling & Stamping Co., 306 U. S. 292, 300. The very smoothness of the “substantial evidence” formula as the standard for reviewing the evidentiary validity of the Board’s findings established its currency. But the inevitably variant applications of the standard to conflicting evidence soon brought contrariety of views and in due course bred criticism. Even though the whole record may have been canvassed in order to determine whether the evidentiary foundation of a determination by the Board was “substantial,” the phrasing of this Court’s process of review readily lent itself to the notion *478 that it was enough that the evidence supporting the Board’s result was “substantial” when considered by itself. It is fair to say that by imperceptible steps regard for the fact-finding function of the Board led to the assumption that the requirements of the Wagner Act were met when the reviewing court could find in the record evidence which, when viewed in isolation, substantiated the Board’s findings. Compare Labor Board v. Waterman Steamship Corp., 309 U. S. 206; Labor Board v. Bradford Dyeing Assn., 310 U. S. 318; and see Labor Board v. Nevada Consolidated Copper Corp., 316 U. S. 105. This is not to say that every member of this Court was consciously guided by this view or that the Court ever explicitly avowed this practice as doctrine. What matters is that the belief justifiably arose that the Court had so construed the obligation to review. 2 Criticism of so contracted a reviewing power reinforced dissatisfaction felt in various quarters with the Board’s administration of the Wagner Act in the years preceding the war. The scheme of the Act was attacked as an inherently unfair fusion of the functions of prosecutor and judge. 3 Accusations of partisan bias were not wanting. 4 The “irresponsible admission and weighing of hearsay, opinion, and emotional speculation in place of factual evidence” was said to be a “serious menace.” 5 No doubt *479 some, perhaps even much, of the criticism was baseless and some surely was reckless. 6 What is here relevant, however, is the climate of opinion thereby generated and its effect on Congress. Protests against “shocking injustices” 7 and intimations of judicial “abdication” 8 with which some courts granted enforcement of the Board’s orders stimulated pressures for legislative relief from alleged administrative excesses. The strength of these pressures was reflected in the passage in 1940 of the Walter-Logan Bill. It was vetoed by President Roosevelt, partly because it imposed unduly rigid limitations on the administrative process, and partly because of the investigation into the actual operation of the administrative process then being conducted by an experienced committee appointed by the Attorney General. 9 It is worth noting that despite its aim to tighten control over administrative determinations of fact, the Walter-Logan Bill contented itself with the conventional formula that an agency’s decision could be set aside if “the findings of fact are not supported by substantial evidence.” 10 *480 The final report of the Attorney General’s Committee was submitted in January, 1941. The majority concluded that “[dissatisfaction with the existing standards as to the scope of judicial review derives largely from dissatisfaction with the fact-finding procedures now employed by the administrative bodies.” 11 Departure from the “substantial evidence” test, it thought, would either create unnecessary uncertainty or transfer to courts the responsibility for ascertaining and assaying matters the significance of which lies outside judicial competence. Accordingly, it recommended against legislation embodying a general scheme of judicial review. 12 *481 Three members of the Committee registered a dissent. Their view was that the "present system or lack of system of judicial review” led to inconsistency and uncertainty. They reported that under a "prevalent” interpretation of the "substantial evidence” rule “if what is called ‘substantial evidence’ is found anywhere in the record to support conclusions of fact, the courts are said to be obliged to sustain the decision without reference to how heavily the countervailing evidence may preponderate — unless indeed the stage of arbitrary decision is reached. Under this interpretation, the courts need to read only one side of the case and, if they find any evidence there, the administrative action is to be sustained and the record to the contrary is to be ignored.” 13 Their view led them to recommend that Congress enact principles of review applicable to all agencies not excepted by unique characteristics. One of these principles was expressed by the formula that judicial review could extend to “findings, inferences, or conclusions of fact unsupported, upon the whole record, by substantial evidence.” 14 So far as the *482 history of this movement for enlarged review reveals, the phrase “upon the whole record” makes its first appearance in this recommendation of the minority of the Attorney General’s Committee. This evidence of the close relationship between the phrase and the criticism out of which it arose is important, for the substance of this formula for judicial review found its way into the statute books when Congress with unquestioning — we might even say uncritical — unanimity enacted the Administrative Procedure Act. 15 *483 One is tempted to say “uncritical” because the legislative history of that Act hardly speaks with that clarity of purpose which Congress supposedly furnishes courts in order to enable them to enforce its true will. On the one hand, the sponsors of the legislation indicated that they were reaffirming the prevailing “substantial evidence” test. 16 But with equal clarity they expressed disapproval of the manner in which the courts were applying *484 their own standard. The committee reports of both houses refer to the practice of agencies to rely upon “suspicion, surmise, implications, or plainly incredible evidence,” and indicate that courts are to exact higher standards “in the exercise of their independent judgment” and on consideration of “the whole record.” 17 Similar dissatisfaction with too restricted application of the “substantial evidence” test is reflected in the legislative history of the Taft-Hartley Act. 18 The bill as reported to the House provided that the “findings of the Board as to the facts shall be conclusive unless it is made to appear to the satisfaction of the court either (1) that the findings of fact are against the manifest weight of the *485 evidence, or (2) that the findings of fact are not supported by substantial evidence.” 19 The bill left the House with this provision. Early committee prints in the Senate provided for review by “weight of the evidence” or “clearly erroneous” standards. 20 But, as the Senate Committee Report relates, “it was finally decided to conform the statute to the corresponding section of the Administrative Procedure Act where the substantial evidence test prevails. In order to clarify any ambiguity in that statute, however, the committee inserted the words ‘questions of fact, if supported by substantial evidence on the record considered as a whole . . . .’ ” 21 This phraseology was adopted by the Senate. The House conferees agreed. They reported to the House: “It is believed that the provisions of the conference agree *486 ment relating to the courts’ reviewing power will be adequate to preclude such decisions as those in N. L. R. B. v. Nevada Consol. Copper Corp. (316 U. S. 105) and in the Wilson, Columbia Products, Union Pacific Stages, Hearst, Republic Aviation, and Le Tourneau, etc. cases, supra, without unduly burdening the courts.” 22 The Senate version became the law. *487 It is fair to say that in all this Congress expressed a mood. And it expressed its mood not merely by oratory but by legislation. As legislation that mood must be respected, even though it can only serve as a standard for judgment and not as a body of rigid rules assuring sameness of application. Enforcement of such broad standards implies subtlety of mind and solidity of judgment. But it is not for us to question that Congress may assume such qualities in the federal judiciary. From the legislative story we have summarized, two concrete conclusions do emerge. One is the identity of aim of the Administrative Procedure Act and the Taft-Hartley Act regarding the proof with which the Labor Board must support a decision. The other is that now Congress has left no room for doubt as to the kind of scrutiny which a Court of Appeals must give the record before the Board to satisfy itself that the Board’s order rests on adequate proof. It would be mischievous word-playing to find that the scope of review under the Taft-Hartley Act is any different from that under the Administrative Procedure Act. The Senate Committee which reported the review clause of the Taft-Hartley Act expressly indicated that the two standards were to conform in this regard, and the wording of the two Acts is for purposes of judicial administration identical. And so we hold that the standard of proof specifically required of the Labor Board by the Taft-Hartley Act is the same as that to be exacted by courts reviewing every administrative action subject to the Administrative Procedure Act. Whether or not it was ever permissible for courts to determine the substantiality of evidence supporting a Labor Board decision merely on the basis of evidence which in and of itself justified it, without taking into account contradictory evidence or evidence from which conflicting inferences could be drawn, the new legislation *488 definitively precludes such a theory of review and bars its practice. The substantiality of evidence must take into account whatever in the record fairly detracts from its weight. This is clearly the significance of the requirement in both statutes that courts consider the whole record. Committee reports and the adoption in the Administrative Procedure Act of the minority views of the Attorney General’s Committee demonstrate that to enjoin such a duty on the reviewing court was one of the important purposes of the movement which eventuated in that enactment. To be sure, the requirement for canvassing “the whole record” in order to ascertain substantiality does not furnish a calculus of value by which a reviewing court can assess the evidence. Nor was it intended to negative the function of the Labor Board as one of those agencies presumably equipped or informed by experience to deal with a specialized field of knowledge, whose findings within that field carry the authority of an expertness which courts do not possess and therefore must respect. Nor does it mean that even as to matters not requiring expertise a court may displace the Board’s choice between two fairly conflicting views, even though the court would justifiably have made a different choice had the matter been before it de novo. Congress has merely made it clear that a reviewing court is not barred from setting aside a Board decision when it cannot conscientiously find that the evidence supporting that decision is substantial, when viewed in the light that the record in its entirety furnishes, including the body of evidence opposed to the Board’s view. There remains, then, the question whether enactment of these two statutes has altered the scope of review other than to require that substantiality be determined in the light of all that the record relevantly presents. A formula for judicial review of administrative action may afford grounds for certitude but cannot assure certainty of appli *489 cation. Some scope for judicial discretion in applying the formula can be avoided only by falsifying the actual process of judging or by using the formula as an instrument of futile casuistry. It cannot be too often repeated that judges are not automata. The ultimate reliance for the fair operation of any standard is a judiciary of high competence and character and the constant play of an informed professional critique upon its work. Since the precise way in which courts interfere with agency findings cannot be imprisoned within any form of words, new formulas attempting to rephrase the old are not likely to be more helpful than the old. There are no talismanic words that can avoid the process of judgment. The difficulty is that we cannot escape, in relation to this problem, the use of undefined defining terms. Whatever changes were made by the Administrative Procedure and Taft-Hartley Acts are clearly within this area where precise definition is impossible. Retention of the familiar “substantial evidence” terminology indicates that no drastic reversal of attitude was intended. But a standard leaving an unavoidable margin for individual judgment does not leave the judicial judgment at large even though the phrasing of the standard does not wholly fence it in. The legislative history of these Acts demonstrates a purpose to impose on courts a responsibility which has not always been recognized. Of course it is a statute and not a committee report which we are interpreting. But the fair interpretation of a statute is often “the art of proliferating a purpose,” Brooklyn National Corp. v. Commissioner, 157 F. 2d 450, 451, revealed more by the demonstrable forces that produced it than by its precise phrasing. The adoption in these statutes of the judicially-constructed “substantial evidence” test was a response to pressures for stricter and more uniform practice, not a reflection of approval of all existing practices. *490 To find the change so elusive that it cannot be precisely defined does not mean it may be ignored. We should fail in our duty to effectuate the will of Congress if we denied recognition to expressed Congressional disapproval of the finality accorded to Labor Board findings by some decisions of this and lower courts, or even of the atmosphere which may have favored those decisions. We conclude, therefore, that the Administrative Procedure Act and the Taft-Hartley Act direct that courts must now assume more responsibility for the reasonableness and fairness of Labor Bo'ard decisions than some courts have shown in the past. Reviewing courts must be influenced by a feeling that they are not to abdicate the conventional judicial function. Congress has imposed on them responsibility for assuring that the Board keeps within reasonable grounds. That responsibility is not less real because it is limited to enforcing the requirement that evidence appear substantial when viewed, on the record as a whole, by courts invested with the authority and enjoying the prestige of the Courts of Appeals. The' Board’s findings are entitled to respect; but they must nonetheless be set aside when the record before a Court of Appeals clearly precludes the Board’s decision from being justified, by a fair estimate of the worth of the testimony of witnesses or its informed judgment on matters within its special competence or both. From this it follows that enactment of these statutes does not require every Court of Appeals to alter its practice. Some — perhaps a majority — have always applied the attitude reflected in this legislation. To explore whether a particular court should or should not alter its practice would only divert attention from the application of the standard now prescribed to a futile inquiry into the nature of the test formerly used by a particular court. Our power to review the correctness of application of the present standard ought seldom to be called into action. *491 Whether on the record as a whole there is substantial evidence to support agency findings is a question which Congress has placed in the keeping of the Courts of Appeals. This Court will intervene only in what ought to be the rare instance when the standard appears to have been misapprehended or grossly misapplied. II. Our disagreement with the view of the court below that the scope of review of Labor Board decisions is unaltered by recent legislation does not of itself, as we have noted, require reversal of its decision. The court may have applied a standard of review which satisfies the present Congressional requirement. The decision of the Court of Appeals is assailed on two grounds. It is said (1) that the court erred in holding that it was barred from taking into account the report of the examiner on questions of fact insofar as that report was rejected by the Board, and (2) that the Board’s order was not supported by substantial evidence on the record considered as a whole, even apart from the validity of the court’s refusal to consider the rejected portions of the examiner’s report. The latter contention is easily met. It is true that two of the earlier decisions of the court below were among those disapproved by Congress. 23 But this disapproval, we have seen, may well have been caused by unintended intimations of judicial phrasing. And in any event, it is clear from the court’s opinion in this case that it in fact did consider the “record as a whole,” and did not deem itself merely the judicial echo of the Board’s conclusion. The testimony of the company’s witnesses was inconsistent, and there was clear evidence that the complaining *492 employee had been discharged by an officer who was at one time influenced against him because of his appearance at the Board hearing. On such a record we could not say that it would be error to grant enforcement. The first contention, however, raises serious questions to which we now turn. III. The Court of Appeals deemed itself bound by the Board's rejection of the examiner’s findings because the court considered these findings not “as unassailable as a master’s.” 24 179 F. 2d at 752. They are not. Section 10 (c) of the Labor Management Relations Act provides that “If upon the preponderance of the testimony taken the Board shall be of the opinion that any person named in the complaint has engaged in or is engaging in any such unfair labor practice, then the Board shall state its findings of fact . . . 61 Stat. 147, 29 U. S. C. (Supp. Ill) § 160 (c). The responsibility for decision thus placed on the Board is wholly inconsistent with the notion that it has power to reverse an examiner’s findings only when they are “clearly erroneous.” Such a limitation would make so drastic a departure from prior administrative practice that explicitness would be required. The Court of Appeals concluded from this premise “that, although the Board would be wrong in totally disregarding his findings, it is practically impossible for a *493 court, upon review of those findings which the Board itself substitutes, to consider the Board’s reversal as a factor in the court’s own decision. This we say, because we cannot find any middle ground between doing that and treating such a reversal as error, whenever it would be such, if done by a judge to a master in equity.” 179 F. 2d at 753. Much as we respect the logical acumen of the Chief Judge of the Court of Appeals, we do not find ourselves pinioned between the horns of his dilemma. We are aware that to give the examiner’s findings less finality than a master’s and yet entitle them to consideration in striking the account, is to introduce another and an unruly factor into the judgmatical process of review. But we ought not to fashion an exclusionary rule merely to reduce the number of imponderables to be considered by reviewing courts. The Taft-Hartley Act provides that “The findings of the Board with respect to questions of fact if supported by substantial evidence on the record considered as a whole shall be conclusive.” 61 Stat. 148, 29 U. S. C. (Supp. Ill) § 160 (e). Surely an examiner’s report is as much a part of the record as the complaint or the testimony. According to the Administrative Procedure Act, “All decisions (including initial, recommended, or tentative decisions) shall become a part of the record . . . .” § 8 (b), 60 Stat. 242, 5 U. S. C. § 1007 (b). We found that this Act’s provision for judicial review has the same meaning as that in the Taft-Hartley Act. The similarity of the two statutes in language and purpose also requires that the definition of “record” found in the Administrative Procedure Act be construed to be applicable as well to the term “record” as used in the Taft-Hartley Act. It is therefore difficult to escape the conclusion that the plain language of the statutes directs a reviewing court to determine the substantiality of evidence on the record including the examiner’s report. The conclusion *494 is confirmed by the indications in the legislative history that enhancement of the status and function of the trial examiner was one of the important purposes of the movement for administrative reform. This aim was set forth by the Attorney General’s Committee on Administrative Procedure: “In general, the relationship upon appeal between the hearing commissioner and the agency ought to a considerable extent to be that of trial court to appellate court. Conclusions, interpretations, law, and policy should, of course, be open to full review. On the other hand, on matters which the hearing commissioner, having heard the evidence and seen the witnesses, is best qualified to decide, the agency should be reluctant to disturb his findings unless error is clearly shown.” 25 Apparently it was the Committee’s opinion that these recommendations should not be obligatory. For the bill which accompanied the Final Report required only that hearing officers make an initial decision which would become final in the absence of further agency action, and that agencies which differed on the facts from their examiners give reasons and record citations supporting their conclusion. 26 This proposal was further moderated by the Administrative Procedure Act. It permits agencies to use examiners to record testimony but not to evaluate it, and contains the rather obscure provision that an agency which reviews an examiner’s report has “all the powers which it would have in making the initial decision.” 27 *495 But this refusal to make mandatory the recommendations of the Attorney General’s Committee should not be construed as a repudiation of them. Nothing in the statutes suggests that the Labor Board should not be influenced by the examiner’s opportunity to observe the witnesses he hears and sees and the Board does not. Nothing suggests that reviewing courts should not give to the examiner’s report such probative force as it intrinsically commands. To the contrary, § 11 of the Administrative Procedure Act contains detailed provisions designed to maintain high standards of independence and competence in examiners. Section 10 (c) of the Labor Management Relations Act requires that examiners "shall issue ... a proposed report, together with a recommended order.” Both statutes thus evince a purpose to increase the importance of the role of examiners in the administrative process. High standards of public administration counsel that we attribute to the Labor Board’s examiners both due regard for the responsibility which Congress imposes on them and the competence to discharge it. 28 *496 The committee reports also make it clear that the sponsors of the legislation thought the statutes gave significance to the findings of examiners. Thus, the Senate Committee responsible for the Administrative Procedure Act explained in its report that examiners’ decisions “would be of consequence, for example, to the extent that material facts in any case depend on the determination of credibility of witnesses as shown by their demeanor or conduct at the hearing.” 29 The House Report reflects the same attitude; 30 and the Senate Committee Report on the Taft-Hartley Act likewise indicates regard for the responsibility devolving on the examiner. 31 We do not require that the examiner’s findings be given more weight than in reason and in the light of judicial experience they deserve. The “substantial evidence” standard is not modified in any way when the Board and its examiner disagree. We intend only to recognize that evidence supporting a conclusion may be less substantial when an impartial, experienced examiner who has observed the witnesses and lived with the case has drawn conclusions different from the Board’s than when he has reached the same conclusion. The findings of the examiner are to be considered along with the consistency and inherent probability of testimony. The significance of his report, of course, depends largely on the importance of credibility in the particular case. To give it this significance does not seem to us materially more difficult *497 than to heed the other factors which in sum determine whether evidence is “substantial.” The direction in which the law moves is often a guide for decision of particular cases, and here it serves to confirm our conclusion. However halting its progress, the trend in litigation is toward a rational inquiry into truth, in which the tribunal considers everything “logically probative of some matter requiring to be proved.” Thayer, A Preliminary Treatise on Evidence, 530; Funk v. United States, 290 U. S. 371. This Court has refused to accept assumptions of fact which are demonstrably false, United States v. Provident Trust Co., 291 U. S. 272, even when agreed to by the parties, Swift & Co. v. Hocking Valley R. Co., 243 U. S. 281. Machinery for discovery of evidence has been strengthened; the boundaries of judicial notice have been slowly but perceptibly enlarged. It would reverse this process for courts to deny examiners’ findings the probative force they would have in the conduct of affairs outside a courtroom. We therefore remand the cause to the Court of Appeals. On reconsideration of the record it should accord the findings of the trial examiner the relevance that they reasonably command in answering the comprehensive question whether the evidence supporting the Board’s order is substantial. But the court need not limit its reexamination of the case to the effect of that report on its decision. We leave it free to grant or deny enforcement as it thinks the principles expressed in this opinion dictate. Judgment vacated and cause remanded. Mr. Justice Black and Mr. Justice Douglas concur with parts I and II of this opinion but as to part III agree with the opinion of the court below, 179 F. 2d 749, 753. Footnotes [1] Labor Board v. Pittsburgh Steamship Co., 180 F. 2d 731, affirmed, post, p. 498. The Courts of Appeals of five circuits have agreed with the Court of Appeals for the Second Circuit that no material change was made in the reviewing power. Eastern Coal Corp. v. Labor Board, 176 F. 2d 131, 134-36 (C. A. 4th Cir.); Labor Board v. La Salle Steel Co., 178 F. 2d 829, 833-834 (C. A. 7th Cir.); Labor Board v. Minnesota Mining & Mfg. Co., 179 F. 2d 323, 325-326 (C. A. 8th Cir.); Labor Board v. Continental Oil Co., 179 F. 2d 552, 555 (C. A. 10th Cir.); Labor Board v. Booker, 180 F. 2d 727, 729 (C. A. 5th Cir.); but see Labor Board v. Caroline Mills, Inc., 167 F. 2d 212, 213 (C. A. 5th Cir.). [2] See the testimony of Dean Stason before the Subcommittee of the Senate Committee on the Judiciary in 1941. Hearings on S. 674, 77th Cong., 1st Sess. 1355-1360. [3] See, for example, the remarks of Laird Bell, then Chairman of the Committee on Administrative Law of the Chicago Bar Association, writing in 1940 in the American Bar Association Journal. 26 A. B. A. J. 552. [4] See Gall, The Current Labor Problem: The View of Industry, 27 Iowa L. Rev. 381, 382. [5] This charge was made by the majority of the Special Committee of the House appointed in 1939 to investigate the National Labor Relations Board. H. R. Rep. No. 1902, 76th Cong., 3d Sess. 76. [6] Professor Gellhorn and Mr. Linfield reached the conclusion in 1939 after an extended investigation that “the denunciations find no support in fact.” Gellhorn and Linfield, Politics and Labor Relations, 39 Col. L. Rev. 339, 394. See also Millis and Brown, From the Wagner Act to Taft-Hartley, 66-75. [7] Wilson & Co. v. Labor Board, 126 F. 2d 114, 117. [8] In Labor Board v. Standard Oil Co., 138 F. 2d 885, 887, Judge Learned Hand said, “We understand the law to be that the decision of the Board upon that issue is for all practical purposes not open to us at all; certainly not after we have once decided that there was 'substantial’ evidence that the ‘disestablished’ union was immediately preceded by a period during which there was a ‘dominated’ union. . . . “[W]e recognize how momentous may be such an abdication of any power of review ....’’ [9] 86 Cong. Rec. 13942-13943, reprinted as H. R. Doc. No. 986, 76th Cong., 3d Sess. [10] S. 915, H. R. 6324, 76th Cong., 1st Sess., § 5 (a). [11] Final Report, 92. [12] Referring to proposals to enlarge the scope of review to permit inquiry whether the findings are supported by the weight of the evidence, the majority said: “Assuming that such a change may be desirable with respect to special administrative determinations, there is serious objection to its adoption for general application. “In the first place there is the question of how much change, if any, the amendment would produce. The respect that courts have for the judgments of specialized tribunals which have carefully considered the problems and the evidence cannot be legislated away. The line between ‘substantial evidence’ and ‘weight of evidence’ is not easily drawn — particularly when the court is confined to a written record, has a limited amount of time, and has no opportunity further to question witnesses on testimony which seems hazy or leaves some lingering doubts unanswered. ‘Substantial evidence’ may well be equivalent to the ‘weight of evidence’ when a tribunal in which one has confidence and which had greater opportunities for accurate determination has already so decided. “In the second place the wisdom of a general change to review of the 'weight of evidence’ is questionable. If the change would require the courts to determine independently which way the evidence preponderates, administrative tribunals would be turned into little more than media for transmission of the evidence to the courts. It would destroy the values of adjudication of fact by experts or specialists in the field involved. It would divide the responsibility for administrative adjudications.” Final Report, 91-92. [13] Id., 210-211. [14] The minority enumerated four “existing deficiencies” in judicial review. These were (1) “the haphazard, uncertain, and variable results of the present system or lack of system of judicial review,” (2) the interpretation permitting substantiality to be determined without taking into account conflicting evidence, (3) the failure of existing formulas “to take account of differences between the various types of fact determinations,” and (4) the practice of determining standards of review by “case-to-case procedure of the courts.” They recommended that “Until Congress finds it practicable to examine into the situation of particular agencies, it should provide more definitely by general legislation for both the availability and scope of judicial review in order to reduce uncertainty and variability. As the Committee recognizes in its report, there are several principal subjects of judicial review— including constitutional questions, statutory interpretation, procedure, and the support of findings of fact by adequate evidence. The last of these should, obviously we think, mean support of all findings of *482 fact, including inferences and conclusion of fact, upon the whole record. Such a legislative provision should, however, be qualified by a direction to the courts to respect the experience, technical competence, specialized knowledge, and discretionary authority of each agency. We have framed such a provision in the appendix to this statement.” Id., 210-212. The text of the recommended provision is as follows: “(e) Scope of review. — As to the findings, conclusions, and decisions in any case, the reviewing court, regardless of the form of the review proceeding, shall consider and decide so far as necessary to its decision and where raised by the parties, all relevant questions of: (1) constitutional right, power, privilege, or immunity; (2) the statutory authority or jurisdiction of the agency; (3) the lawfulness and adequacy of procedure; (4) findings, inferences, or conclusions of fact unsupported, upon the whole record, by substantial evidence; and (5) administrative action otherwise arbitrary or capricious. Provided, however, That upon such review due weight shall be accorded the experience, technical competence, specialized knowledge, and legislative policy of the agency involved as well as the discretionary authority conferred upon it.” Id., 246-247. [15] 60 Stat. 237, 5 U. S. C. § 1001 et seq. The form finally adopted reads as follows: “Sec. 10. Except so far as (1) statutes preclude judicial review or . (2) agency action is by law committed to agency discretion— “(e) Scope op review. — So far as necessary to decision and where presented the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of any agency action. It shall (A) compel agency action unlawfully withheld or unreason *483 ably delayed; and (B) hold unlawful and set aside agency action, findings, and conclusions found to be (1) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law; (2) contrary to constitutional right, power, privilege, or immunity; (3) in excess of statutory jurisdiction, authority, or limitations, or short of statutory right; (4) without observance of procedure required by law; (5) unsupported by substantial evidence in any case subject "to the requirements of sections 7 and 8 or otherwise reviewed on the record of an agency hearing provided by statute; or (6) unwarranted by the facts to the extent that the facts are subject to trial de novo by the reviewing court. In making the foregoing determinations the court shall review the whole record or such portions thereof as may be cited by any party, and due account shall be taken of the rule of prejudicial error.” 60 Stat. 243-244, 5 U. S. C. § 1009 (e). (Italics ours.) In the form in which the bill was originally presented to Congress, clause (B) (5) read, “unsupported by competent, material, and substantial evidence upon the whole agency record as reviewed by the court in any case subject to the requirements of sections 7 and 8.” H. R. 1203, 79th Cong., 1st Sess., quoted in S. Doc. No. 248, 79th Cong., 2d Sess. 155, 160. References to competency and materiality of evidence were deleted and the final sentence added by the Senate Committee. S. Rep. No. 752, 79th Cong., 1st Sess. 28; S. Doc. No. 248, supra, 39-40, 214. No reason was given for the deletion. [16] A statement of the Attorney General appended to the Senate Report explained that the bill “is intended to embody the law as declared, for example, in Consolidated Edison Co. v. National Labor Relations Board (305 U. S. 197).” Section 10 (e) of Appendix B to S. Rep. No. 752, supra, reprinted in S. Doc. No. 248, supra, 230. Mr. McFarland, then Chairman of the American Bar Association Committee on Administrative Law, testified before the House Judiciary Committee to the same effect. Id., 85-86. [17] The following quotation from the report of the Senate Judiciary Committee indicates the position of the sponsors. “The ‘substantial evidence’ rule set forth in section 10 (e) is exceedingly important. As a matter of language, substantial evidence would seem to be an adequate expression of law. The difficulty comes about in the practice of agencies to rely upon (and of courts to tacitly approve) something less — to rely upon suspicion, surmise, implications, or plainly incredible evidence. It will be the duty of the courts to determine in the final analysis and in the exercise of their independent judgment, whether on the whole record the evidence in a given instance is sufficiently substantial to support a finding, conclusion, or other agency action as a matter of law. In the first instance, however, it will be the function of the agency to determine the sufficiency of the evidence upon which it acts — and the proper performance of its public duties will require it to undertake this inquiry in a careful and dispassionate manner. Should these objectives of the bill as worded fail, supplemental legislation will be required.” S. Rep. No. 752, swpra, 30-31. The House Committee Report is to substantially the same effect. H. R. Rep. No. 1980, 79th Cong., 2d Sess. 45. The reports are reprinted in S. Doc. No. 248, supra, 216-217, 279. See also the response of Senator McCarran in debate, to the effect that the bill changed the “rule” that courts were “powerless to interfere” when there “was no probative evidence.” Id., 322. And see the comment of Congressman Springer, a member of the House Judiciary Committee, id., 376. [18] 61 Stat. 136, 29 U. S. C. (Supp. Ill) § 141 et seq. [19] H. R. 3020, 80th Cong., 1st Sess., § 10 (e), reprinted in 1 Legislative History of the Labor Management Relations Act, 1947, p. 71. [20] The history of the evolution of the Senate provision was given by Senator Morse. 93 Cong. Rec. 5108, reprinted in 2 Legislative History 1504-1505. The prints were not approved by the Committee. [21] S. Rep. No. 105, 80th Cong., 1st-Sess. 26-27, reprinted in 1 Legislative History 432-433. The Committee did not explain what the ambiguity might be; and it is to be noted that the phrase it italicized is indistinguishable in content from the requirement of § 10 (e) of the Administrative Procedure Act that “the court shall review the whole record or such portions thereof as may be cited by any party . . . .” Senator Taft gave this explanation to the Senate of the meaning of the section: “In the first place, the evidence must be substantial; in the second place, it must still look substantial when viewed in the light of the entire record. That does not go so far as saying that a decision can be reversed on the weight of the evidence. It does not go quite so far as the power given to a circuit court of appeals to review a district-court decision, but it goes a great deal further than the present law, and gives the court greater opportunity to reverse an obviously unjust decision on the part of the National Labor Relations Board.” 93 Cong. Rec. 3839, reprinted in 2 Legislative History 1014. [22] H. R. Rep. No. 510, 80th Cong., 1st Sess. 56, reprinted in 1 Legislative History 560. In Labor Board v. Nevada Consolidated Copper Corp., 316 U. S. 105, 107, we reversed a judgment refusing to enforce a Board order because “upon an examination of the record we cannot say that the findings of fact of the Board are without support in the evidence.” The sufficiency of evidence to support findings of fact is not involved in the three other decisions of this Court to which reference was made. Labor Board v. Hearst Publications, Inc., 322 U. S. 111; Republic Aviation Corp. v. Labor Board and Labor Board v. Le Tourneau Co., 324 U. S. 793. The language used by the court offers a probable explanation for including two of the decisions of Courts of Appeals. In Wilson & Co. v. Labor Board, 126 F. 2d 114, 117, the Court of Appeals for the Seventh Circuit sustained a finding that the employer dominated a company union after stating that it had “recognized (or tried to) that findings must be sustained, even when they are contrary to the great weight of the evidence, and we have ignored, or at least endeavored to ignore, the shocking injustices which such findings, opposed to the overwhelming weight of the evidence, produce.” Labor Board v. Columbia Products Corp., 141 F. 2d 687, 688, is a per curiam decision of the Court of Appeals for the Second Circuit sustaining a finding of discriminatory discharge. The court said of the Board’s decision on a question of fact, “Though it may strain our credulity, if it does not quite break it down, we must accept it . . . .” The reason for disapproval of Labor Board v. Union Pacific Stages, 99 F. 2d 153, is not apparent. The Court of Appeals for the Ninth Circuit there enforced the portion of the Board’s order directing the company to disavow a policy of discrimination against union members, on the ground that there appeared “to be evidence, although disputed,” that some company officials had discouraged employees from joining. 99 F. 2d at 179. The bulk of the lengthy opinion, however, is devoted to a discussion of the facts to support the court’s conclusion that the Board’s findings of discriminatory discharges should not be sustained. [23] Labor Board v. Standard Oil Co., 138 F. 2d 885; Labor Board v. Columbia Products Corp., 141 F. 2d 687. See notes 8 and 22, supra. [24] Rule 53 (e) (2), Fed. Rules Civ. Proc., gives finality to the findings of a master unless they are clearly erroneous. The court’s ruling excluding from consideration disagreement between the Board and the examiner was in apparent conflict with the views of three other circuits. Labor Board v. Ohio Calcium Co., 133 F. 2d 721, 724 (C. A. 6th Cir.); A. E. Staley Mfg. Co. v. Labor Board, 117 F. 2d 868, 878 (C. A. 7th Cir.); Wilson & Co. v. Labor Board, 123 F. 2d 411, 418 (C. A. 8th Cir.); cf. International Assn. of Machinists v. Labor Board, 71 App. D. C. 175, 180, 110 F. 2d 29, 34 (C. A. D. C. Cir.). [25] Final Report, 51. [26] §§308 (1) and 309 (2) of the proposed bill, quoted in Final Report, 200, 201. [27] §8 (a), 60 Stat. 242, 5 U. S. C. § 1007 (a). The quoted provision did not appear in the bill in the form in which it was introduced into the Senate. S. 7, 79th Cong., 1st Sess., § 7. It was added by *495 the Senate Judiciary Committee. The Committee published its reasons for modifying the earlier draft, but gave no explanation for this particular change. See S. Doc. No. 248, supra, 32-33. It is likely that the sentence was intended to embody a clause in the draft prepared by the Attorney General’s Committee, which provided that on review of a case decided initially by an examiner an agency should have jurisdiction to remand or to “affirm, reverse, modify, or set aside in whole or in part the decision of the hearing commissioner, or itself to make any finding which in its judgment is proper upon the record.” §309 (2), Final Report, 201. The substance of this recommendation was included in bills introduced into the House. H. R. 184, 79th Cong., 1st Sess., § 309 (2), and H. R. 339, 79th Cong., 1st Sess., §7 (c), both quoted in S. Doc. No. 248, supra, 138, 143. [28] Salaries of trial examiners range from $7,600 to $10,750 per year. See Appendix to the Budget of the United States Government for the fiscal year ending June 30,1952, p. 47. [29] S. Rep. No. 752, supra, 24, reproduced in S. Doc. No. 248, supra, 210. [30] H. R. Rep. No. 1980, 79th Cong., 2d Sess. 38-39, reprinted in S. Doc. No. 248, supra, 272-273. The House Report added that “In a broad sense the agencies’ reviewing powers are to be compared with that of courts under section 10 (e) of the bill.” The language of the statute offers no support for this statement. [31] S. Rep. No. 105, 80th Cong., 1st Sess. 9, quoted in 1 Legislative History of the Labor Management Relations Act, 1947, p. 415.