Dickinson v. ZurkoAnnotate this Case
527 U.S. 150 (1999)
OCTOBER TERM, 1998
DICKINSON, ACTING COMMISSIONER OF PATENTS AND TRADEMARKS v. ZURKO ET AL.
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE FEDERAL CIRCUIT
No. 98-377. Argued March 24, 1999-Decided June 10, 1999
In reviewing a Patent and Trademark Office (PTO) decision to deny respondents' patent application, the Federal Circuit analyzed the PTO's factual finding using a "clearly erroneous" standard of review, which generally governs appellate review of district court findings of fact (court/court review), rather than the less stringent standards set forth in the Administrative Procedure Act (APA), which permit a court to set aside agency findings of fact found to be arbitrary, capricious, an abuse of discretion, or unsupported by substantial evidence (court/agency review), 5 U. S. C. § 706. The court found the PTO's factual finding to be clearly erroneous.
Held: The Federal Circuit must use the framework set forth in § 706 when reviewing PTO findings of fact. Pp. 154-165.
(a) Absent an exception, a reviewing court must apply the APA's court/agency review standards to agency factual findings. The Federal Circuit bases such an exception on 5 U. S. C. § 559, which provides that the APA does "not limit or repeal additional requirements ... recognized by law." In its view, at the time the APA was adopted in 1946, the Court of Customs and Patent Appeals (CCPA), a Federal Circuit predecessor, applied a court/court standard that was stricter than ordinary court/agency review standards, and this special tradition of strict review amounted to an "additional requirement" that trumps § 706's requirements. However, a close examination of the CCPA's cases reviewing PTO decisions do not reflect a well-established court/court standard. The presence of the phrases "clear case of error," "clearly wrong," and "manifest error" in those cases does not conclusively signal such review. The relevant linguistic conventions were less firmly established before the APA's adoption than they are today, with courts sometimes using words such as "clearly erroneous" to describe less strict court/agency review and words such as "substantial evidence" to describe stricter court/court review. The absence of the words "substantial evidence" in the CCPA's cases is not especially significant, since standardization of that term began to take hold only after Congress started using it in various federal statutes. Further, not one of the CCPA's opinions actually uses the words "clear error" or "clearly erroneous," which are terms
of art signaling court/court review. Most of them use "manifest error," which is not now such a term of art. At the same time, this Court's precedent undermines the claim that "clearly wrong" or "manifest error" signal court/court review. Although the Court in Morgan v. Daniels, 153 U. S. 120, used language that could be read as setting forth a court/court standard, the Court's reasoning makes clear that it meant its words to stand for a court/agency standard. The CCPA's cases reveal a similar pattern, using words such as "clearly wrong" and "manifest error" with explanations indicating that they had court/agency, not court/court, review in mind. Pp. 154-161.
(b) Several policy reasons that the Federal Circuit believes militate against using APA review standards-that a change will be disruptive to the bench and bar; that the change will create an anomaly in which a disappointed patent applicant who seeks review directly in the Federal Circuit will be subject to court/agency review, while one who first seeks review in a district court will have any further appeal reviewed under a court/court standard; and that stricter review produces better agency factfinding-are unconvincing. Pp. 161-165.
142 F.3d 1447, reversed and remanded.
BREYER, J., delivered the opinion of the Court, in which STEVENS, O'CONNOR, SCALIA, SOUTER, and THOMAS, JJ., joined. REHNQUIST, C. J., filed a dissenting opinion, in which KENNEDY and GINSBURG, JJ., joined, post, p. 170.
Deputy Solicitor General Wallace argued the cause for petitioner. With him on the briefs were Solicitor General Waxman, Assistant Attorney General Hunger, Edward C. DuMont, William Kanter, Bruce G. Forrest, Albin F. Drost, Karen A. Buchanan, and Kenneth R. Corsello.
Ernest Gellhorn argued the cause for respondents. With him on the brief were Jeffrey S. Lubbers, Ann G. Weymouth, Janice M. Mueller, and Russell Wong. *
*Briefs of amici curiae urging reversal were filed for Intellectual Property Professors by John F. Duffy and Thomas G. Field, Jr.; and for Theis Research, Inc., by Paul R. Johnson.
Briefs of amici curiae urging affirmance were filed for the Biotechnology Industry Organization by Scott F. Partridge, Bob E. Shannon, and Scott K. Field; for the Houston Intellectual Property Law Association by Jeffrey W Tayon; for the International Trademark Association by Albert Robin; for the New York Intellectual Property Law Association by Bruce
JUSTICE BREYER delivered the opinion of the Court.
The Administrative Procedure Act (APA) sets forth standards governing judicial review of findings of fact made by federal administrative agencies. 5 U. S. C. § 706. We must decide whether § 706 applies when the Federal Circuit reviews findings of fact made by the Patent and Trademark Office (PTO). We conclude that it does apply, and the Federal Circuit must use the framework set forth in that section.
Section 706, originally enacted in 1946, sets forth standards that govern the "Scope" of court "review" of, e. g., agency factfinding (what we shall call court/agency review). It says that a
"reviewing court shall-
"(2) hold unlawful and set aside agency ... findings ... found to be-
"(A) arbitrary, capricious, [or] an abuse of discretion,
"(E) unsupported by substantial evidence in a case subject to sections 556 and 557 of this title or otherwise reviewed on the record of an agency hearing provided by statute; ...
M. Wexler and Howard B. Barnaby; for the Patent, Trademark & Copyright Section of the Bar Association of the District of Columbia by Lynn Eccleston, David W Long, and Harold Wegner; for Pharmaceutical Research and Manufacturers of America by Gerald J. Mossinghoff; and for John P. Sutton, pro se.
Briefs of amici curiae were filed for the Dallas-Fort Worth Intellectual Property Law Association by D. Scott Hemingway; and for Intellectual Property Creators et al. by David Roy Pressman, pro se.
"In making the foregoing determinations, the court shall review the whole record or those parts of it cited by a party .... "
Federal Rule of Civil Procedure 52 (a) sets forth standards that govern appellate court review of findings of fact made by a district court judge (what we shall call court/court review). It says that the appellate court shall set aside those findings only if they are "clearly erroneous." Traditionally, this court/court standard of review has been considered somewhat stricter (i. e., allowing somewhat closer judicial review) than the APA's court/agency standards. 2 K. Davis & R. Pierce, Administrative Law Treatise § 11.2,
The Court of Appeals for the Federal Circuit believes that it should apply the "clearly erroneous" standard when it reviews findings of fact made by the PTO. In re Zurko, 142 F.3d 1447, 1459 (1998) (case below). The Commissioner of Patents, the PTO's head, believes to the contrary that ordinary APA court/agency standards apply. See, e. g., In re Kemps, 97 F.3d 1427, 1430-1431 (CA Fed. 1996); In re Napier, 55 F.3d 610, 614 (CA Fed. 1995); In re Brana, 51
The case before us tests these two competing legal views.
Respondents applied for a patent upon a method for increasing computer security. The PTO patent examiner concluded that respondents' method was obvious in light of prior art, and so it denied the application. See 35 U. S. C. § 103 (1994 ed., Supp. III). The PTO's review board (the Board of Patent Appeals and Interferences) upheld the examiner's decision. Respondents sought review in the Federal Circuit, where a panel treated the question of what the prior art teaches as one of fact, and agreed with respondents that the PTO's factual finding was "clearly erroneous." In re Zurko, 111 F.3d 887, 889, and n. 2 (1997).
The Federal Circuit, hoping definitively to resolve the review-standard controversy, then heard the matter en banco
After examining relevant precedents, the en banc court concluded that its use of the stricter court/court standard was legally proper. The Solicitor General, representing the Commissioner of Patents, sought certiorari. We granted the writ in order to decide whether the Federal Circuit's review of PTO factfinding must take place within the framework set forth in the AP A.
The parties agree that the PTO is an "agency" subject to the AP A's constraints, that the PTO's finding at issue in this case is one of fact, and that the finding constitutes "agency action." See 5 U. S. C. § 701 (defining "agency" as an "authority of the Government of the United States"); § 706 (applying AP A "Scope of review" provisions to "agency action"). Hence a reviewing court must apply the AP A's court/agency review standards in the absence of an exception.
The Federal Circuit rests its claim for an exception upon § 559. That section says that the APA does "not limit or repeal additional requirements ... recognized by law." In the Circuit's view: (1) at the time of the APA's adoption, in 1946, the Court of Customs and Patent Appeals (CCPA), a Federal Circuit predecessor, applied a court/court "clearly erroneous" standard; (2) that standard was stricter than ordinary court/agency review standards; and (3) that special tradition of strict review consequently amounted to an "additional requirement" that under § 559 trumps the requirements imposed by § 706.
Recognizing the importance of maintaining a uniform approach to judicial review of administrative action, see, e. g., Universal Camera Corp. v. NLRB, 340 U. S. 474, 489 (1951); 92 Congo Rec. 5654 (1946) (statement of Rep. Walter), we have closely examined the Federal Circuit's claim for an exception to that uniformity. In doing so, we believe that respondents must show more than a possibility of a height-
ened standard, and indeed more than even a bare preponderance of evidence in their favor. Existence of the additional requirement must be clear. This is suggested both by the phrase "recognized by law" and by the congressional specification in the APA that "[n]o subsequent legislation shall be held to supersede or modify the provisions of this Act except to the extent that such legislation shall do so expressly." § 12, 60 Stat. 244, 5 U. S. C. § 559. A statutory intent that legislative departure from the norm must be clear suggests a need for similar clarity in respect to grandfathered common-law variations. The APA was meant to bring uniformity to a field full of variation and diversity. It would frustrate that purpose to permit divergence on the basis of a requirement "recognized" only as ambiguous. In any event, we have examined the 89 cases which, according to respondents and supporting amici, embody the pre-APA standard of review. See App. to Brief for New York Intellectual Property Law Association as Amicus Curiae 1a-6a (collecting cases), and we conclude that those cases do not reflect a well-established stricter court/court standard of judicial review for PTO factfinding, which circumstance fatally undermines the Federal Circuit's conclusion.
The 89 pre-APA cases all involve CCPA review of a PTO administrative decision, which either denied a patent or awarded priority to one of several competing applicants. See 35 U. S. C. § 59a (1934 ed.) (granting CCP A review authority over PTO decisions); 35 U. S. C. § 141 (current grant of review authority to the Federal Circuit). The major consideration that favors the Federal Circuit's view consists of the fact that 23 of the cases use words such as "clear case of error" or "clearly wrong" to describe the CCPA's review standard, while the remainder use words such as "manifest error," which might be thought to mean the same thing. See App. to Brief for New York Intellectual Property Law Association as Amicus Curiae 1a-6a. When the CCP A decided many of these cases during the 1930's and early 1940's,
legal authorities had begun with increasing regularity to use the term "clearly erroneous" to signal court/court review, Fed. Rule Civ. Proc. 52(a) (adopted in 1937), and the term "substantial evidence" to signal less strict court/agency review. Stern, Review of Findings of Administrators, Judges and Juries: A Comparative Analysis, 58 Harv. L. Rev. 70, 88 (1944) (describing congressional debates in which members argued for and against applying the "clearly erroneous" standard to agency review "precisely because it would give administrative findings less finality than they enjoyed under the 'substantial evidence' rule").
Yet the presence of these phrases is not conclusive. The relevant linguistic conventions were less firmly established before adoption of the AP A than they are today. At that time courts sometimes used words such as "clearly erroneous" to describe less strict court/agency review standards. See, e. g., Polish National Alliance v. NLRB, 136 F.2d 175, 181 (CA7 1943); New York Trust Co. v. SEC, 131 F.2d 274, 275 (CA2 1942), cert. denied, 318 U. S. 786 (1943); Hall v. Commissioner, 128 F.2d 180, 182 (CA71942); First National Bank of Memphis v. Commissioner, 125 F.2d 157 (CA6 1942) (per curiam); NLRB v. Algoma Plywood & Veneer Co., 121 F.2d 602, 606 (CA7 1941). Other times they used words such as "substantial evidence" to describe stricter court/ court review (including appeals in patent infringement cases challenging district court factfinding). See, e. g., Cornell v. Chase Brass & Copper Co., 142 F.2d 157, 160 (CA2 1944); Dow Chemical Co. v. Halliburton Oil Well Cementing Co., 139 F.2d 473, 475 (CA6 1943), aff'd, 324 U. S. 320 (1945); Gordon Form Lathe Co. v. Ford Motor Co., 133 F.2d 487, 496-497 (CA6), aff'd, 320 U. S. 714 (1943); Electro Mfg. Co. v. Yellin, 132 F.2d 979, 981 (CA7 1943); Ajax Hand Brake Co. v. Superior Hand Brake Co., 132 F.2d 606,609 (CA7 1943); Galion Iron Works & Mfg. Co. v. Beckwith Machinery Co., 105 F.2d 941, 942 (CA3 1939). Indeed, this Court itself on at least one occasion used the words "substantial evidence"
to explain why it would not disturb a trial court's factual findings. Borden's Farm Products Co. v. Ten Eyck, 297 U. S. 251, 261 (1936); see also Great Atlantic & Pacific Tea Co. v. Grosjean, 301 U. S. 412, 420 (1937) (accepting trial court's findings of fact because they have "substantial support in the record").
Nor is the absence of the words "substantial evidence" in the CCP A's cases especially significant. Before the AP A, the use of that term to describe court/agency review proceeded by fits and starts, with the standardization of the term beginning to take hold only after Congress began using it (or the like) in various federal statutes. For example, this Court first used the phrase "substantial evidence" in the agency context to describe its approach to the Interstate Commerce Commission's (ICC's) factual findings, ICC v. Union Pacific R. Co., 222 U. S. 541, 548 (1912), even though the underlying statute simply authorized a court of competent jurisdiction to suspend or set aside orders of the Commission, § 12, 36 Stat. 551. The Court did not immediately grant the Federal Trade Commission the same leeway it granted the ICC, see FTC v. Curtis Publishing Co., 260 U. S. 568, 580 (1923), even though the underlying Act used language to which the phrase "substantial evidence" might have applied, see § 5, 38 Stat. 720 (the "findings of the commission as to the facts, if supported by testimony, shall be conclusive"). As the words "substantial evidence" began to appear more often in statutes, the Court began to use those same words in describing review standards, sometimes supplying the modifier "substantial" when Congress had left it out. See, e. g., Consolidated Edison Co. v. NLRB, 305 U. S. 197,229 (1938); see Stason, "Substantial Evidence" in Administrative Law, 89 U. Pa. L. Rev. 1026, 1026-1028 (1941) (collecting statutes); see also Dobson v. Commissioner, 320 U. S. 489, 499 (1943) (speaking generally of the "theoretical and practical reason[s] for ... [crediting] administrative decisions"). The patent statutes, however, did not and do not
use the term "substantial evidence" or any other term to describe the standard of court review. 35 U. S. C. §§ 61, 62 (1934 ed.). Indeed, it apparently remains disputed to this day (a dispute we need not settle today) precisely which APA standard-"substantial evidence" or "arbitrary, capricious, abuse of discretion"-would apply to court review of PTO factfinding. See 5 U. S. C. § 706(2)(E) (applying the term "substantial evidence" where agency factfinding takes place "on the record"); see also Association of Data Processing Service Orgs., Inc. v. Board of Governors of Federal Reserve System, 745 F.2d 677, 683-684 (CADC 1984) (Scalia, J.) (finding no difference between the APA's "arbitrary, capricious" standard and its "substantial evidence" standard as applied to court review of agency factfinding.)
Further, not one of the 89 opinions actually uses the precise words "clear error" or "clearly erroneous," which are terms of art signaling court/court review. Most of the 89 opinions use words like "manifest error," which is not now such a term of art.
At the same time, precedent from this Court undermines the Federal Circuit's claim that the phrases "clearly wrong" or "manifest error" signal court/court review. The Federal Circuit traced its standard of review back to Morgan v. Daniels, 153 U. S. 120 (1894), which it characterized as the foundation upon which the CCP A later built its review standards. 142 F. 3d, at 1453-1454. We shall describe that case in some detail.
Morgan arose out of a Patent Office interference proceeding-a proceeding to determine which of two claimants was the first inventor. The Patent Office decided the factual question of "priority" in favor of one claimant; the Circuit Court, deciding the case "without any additional testimony," 153 U. S., at 122, reversed the Patent Office's factual finding and awarded the patent to the other claimant. This Court in turn reversed the Circuit Court, thereby restoring the Patent Office decision.
"What," asked Justice Brewer for the Court, "is the rule which should control the [reviewing] court in the determination of this case?" Ibid. Is it that the Patent Office decision "should stand unless the testimony shows beyond any reasonable doubt that the plaintiff was the first inventor"? I d., at 123. The Court then cited two cases standing for such a "reasonable doubt" standard. Ibid. (citing Cantrell v. Wallick, 117 U. S. 689, 695 (1886), and Coffin v. Ogden, 18 Wall. 120, 124 (1874)). The Court found the two cases "closely in point." 153 U. S., at 123. Justice Brewer wrote that a person "challenging the priority awarded by the Patent Office ... should ... be held to as strict proof." Ibid. (emphasis added). The Court, pointing out that the Circuit Court had used language "not quite so strong" (namely, "a clear and undoubted preponderance of proof"), thought that the Circuit Court's standard sounded more like the rule used by "an appellate court in reviewing findings of fact made by the trial court." Ibid. The Court then wrote:
"But this is something more than a mere appeal. It is an application to the court to set aside the action of one of the executive departments of the government . . . . A new proceeding is instituted in the courts ... to set aside the conclusions reached by the administrative department .... It is ... not to be sustained by a mere preponderance of evidence .... It is a controversy between two individuals over a question of fact which has once been settled by a special tribunal, entrusted with full power in the premises. As such it might be well argued, were it not for the terms of this statute, that the decision of the patent office was a finality upon every matter of fact." Id., at 124 (emphasis added).
The Court, in other words, reasoned strongly that a court/ court review standard is not proper; that standard is too strict; a somewhat weaker standard of review is appropriate.
We concede that the Court also used language that could be read as setting forth a court/court standard of review. It said, for example, that the
"Patent Office [decision] must be accepted as controlling upon that question of fact ... unless the contrary is established by testimony which ... carries thorough conviction .... [I]f doubtful, the decision of the Patent Office must control." Id., at 125 (emphasis added).
It added that the testimony was "not ... sufficient to produce a clear conviction that the Patent Office made a mistake." I d., at 129 (emphasis added). But the Court did not use the emphasized words today; it used those words more than 100 years ago. And its reasoning makes clear that it meant those words to stand for a court/agency review standard, a standard weaker than the standard used by "an appellate court in reviewing findings of fact made by the trial court." Id., at 123.
The opinions in the 89 CCP A cases, cataloged in the Appendix to this opinion, reveal the same pattern. They use words such as "manifest error" or "clearly wrong." But they use those words to explain why they give so much, not so little, deference to agency factfinding. And, their further explanations, when given, indicate that they had court! agency, not court/court, review in mind.
In nearly half of the cases, the CCP A explains why it uses its "manifest error" standard by pointing out that the PTO is an expert body, or that the PTO can better deal with the technically complex subject matter, and that the PTO consequently deserves deference. In more than three-fourths of the cases the CCP A says that it should defer to PTO factfinding because two (and sometimes more) PTO tribunals had reviewed the matter and agreed about the factual finding. These reasons are reasons that courts and commentators have long invoked to justify deference to agency factfinding. See Universal Camera, 340 U. S., at 496-497 (intraagency
agreement); NLRB v. Link-Belt Co., 311 U. S. 584, 597 (1941) (expertise); Rochester Telephone Corp. v. United States, 307 U. S. 125, 145-146 (1939) (expertise); ICC v. Louisville & Nashville R. Co., 227 U. S. 88, 98 (1913) (expertise); Stern, 58 Harv. L. Rev., at 81-82 (expertise); 2 Davis & Pierce § 11.2, at 178-181 (intraagency agreement). They are not the reasons courts typically have given for deferring to factfinding made by a lower court judge. See, e. g., Concrete Pipe & Products of Cal., Inc. v. Construction Laborers Pension Trust for Southern Cal., 508 U. S. 602, 623 (1993); Stern, supra, at 82-83 (trial court advantages lie in, e. g., evaluation of witness, not comparative expertise). And we think it also worth noting, in light of the pre-APA movement toward standardization discussed above, supra, at 157, that the CCPA began to refer more frequently to technical complexity and agency expertise as time marched closer to 1946. Out of the 45 cases in our sample decided between 1929 and 1936, 40% (18 of 45) specifically referred to technical complexity. That percentage increased to 57% (25 of 44) for the years 1937 to 1946.
Given the CCPA's explanations, the review standard's origins, and the nondeterminative nature of the phrases, we cannot agree with the Federal Circuit that in 1946, when Congress enacted the AP A, the CCP A "recognized" the use of a stricter court/court, rather than a less strict court/ agency, review standard for PTO decisions. Hence the Federal Circuit's review of PTO findings of fact cannot amount to an "additional requiremen[t] ... recognized by law." 5 U. S. C. § 559.
The Federal Circuit also advanced several policy reasons which in its view militate against use of AP A standards of review. First, it says that both bench and bar have now become used to the Circuit's application of a "clearly erroneous" standard that implies somewhat stricter court/court review. It says that change may prove needlessly disrup-
tive. 142 F. 3d, at 1457-1458. Supporting amici add that it is better that the matter remain "'settled than that it be settled right.'" Brief for Patent, Trademark & Copyright Section of the Bar Association of the District of Columbia as Amicus Curiae 23 (quoting Square D Co. v. Niagara Frontier Tariff Bureau, Inc., 476 U. S. 409, 424 (1986)).
This Court, however, has not previously settled the matter. The Federal Circuit's standard would require us to create § 559 precedent that itself could prove disruptive by too readily permitting other agencies to depart from uniform AP A requirements. And in any event we believe the Circuit overstates the difference that a change of standard will mean in practice.
This Court has described the AP A court/agency "substantial evidence" standard as requiring a court to ask whether a "reasonable mind might accept" a particular evidentiary record as "adequate to support a conclusion." Consolidated Edison, 305 U. S., at 229. It has described the court/court "clearly erroneous" standard in terms of whether a reviewing judge has a "definite and firm conviction" that an error has been committed. United States v. United States Gypsum Co., 333 U. S. 364, 395 (1948). And it has suggested that the former is somewhat less strict than the latter. Universal Camera, 340 U. S., at 477, 488 (analogizing "substantial evidence" test to review of jury findings and stating that appellate courts must respect agency expertise). At the same time the Court has stressed the importance of not simply rubber-stamping agency factfinding. Id., at 490. The AP A requires meaningful review; and its enactment meant stricter judicial review of agency factfinding than Congress believed some courts had previously conducted. Ibid.
The upshot in terms of judicial review is some practical difference in outcome depending upon which standard is used. The court/agency standard, as we have said, is somewhat less strict than the court/court standard. But the dif-
ference is a subtle one-so fine that (apart from the present case) we have failed to uncover a single instance in which a reviewing court conceded that use of one standard rather than the other would in fact have produced a different outcome. Cf. International Brotherhood of Electrical Workers v. NLRB, 448 F.2d 1127, 1142 (CADC 1971) (Leventhal, J., dissenting) (wrongly believing-and correcting himself-that he had found the "case dreamed of by law school professors" where the agency's findings, though "clearly erroneous," were "nevertheless" supported by "substantial evidence").
The difficulty of finding such a case may in part reflect the basic similarity of the reviewing task, which requires judges to apply logic and experience to an evidentiary record, whether that record was made in a court or by an agency. It may in part reflect the difficulty of attempting to capture in a form of words intangible factors such as judicial confidence in the fairness of the factfinding process. Universal Camera, supra, at 489; Jaffe, Judicial Review: "Substantial Evidence on the Whole Record," 64 Harv. L. Rev. 1233, 1245 (1951). It may in part reflect the comparatively greater importance of case-specific factors, such as a finding's dependence upon agency expertise or the presence of internal agency review, which factors will often prove more influential in respect to outcome than will the applicable standard of review.
These features of review underline the importance of the fact that, when a Federal Circuit judge reviews PTO factfinding, he or she often will examine that finding through the lens of patent-related experience-and properly so, for the Federal Circuit is a specialized court. That comparative expertise, by enabling the Circuit better to understand the basis for the PTO's finding of fact, may play a more important role in assuring proper review than would a theoretically somewhat stricter standard.
Moreover, if the Circuit means to suggest that a change of standard could somehow immunize the PTO's fact-related "reasoning" from review, 142 F. 3d, at 1449-1450, we disagree. A reviewing court reviews an agency's reasoning to determine whether it is "arbitrary" or "capricious," or, if bound up with a record-based factual conclusion, to determine whether it is supported by "substantial evidence."
Second, the Circuit and its supporting amici believe that a change to AP A review standards will create an anomaly. An applicant denied a patent can seek review either directly in the Federal Circuit, see 35 U. S. C. § 141, or indirectly by first obtaining direct review in federal district court, see § 145. The first path will now bring about Federal Circuit court/agency review; the second path might well lead to Federal Circuit court/court review, for the Circuit now reviews federal district court factfinding using a "clearly erroneous" standard. Gould v. Quigg, 822 F.2d 1074, 1077 (1987). The result, the Circuit claims, is that the outcome may turn upon which path a disappointed applicant takes; and it fears that those applicants will often take the more complicated, timeconsuming indirect path in order to obtain stricter judicial review of the PTO's determination.
We are not convinced, however, that the presence of the two paths creates a significant anomaly. The second path permits the disappointed applicant to present to the court evidence that the applicant did not present to the PTO. Ibid. The presence of such new or different evidence makes a factfinder of the district judge. And nonexpert judicial factfinding calls for the court/court standard of review. We concede that an anomaly might exist insofar as the district judge does no more than review PTO factfinding, but nothing in this opinion prevents the Federal Circuit from adjusting related review standards where necessary. Cf. Fregeau v. Mossinghoff, 776 F.2d 1034, 1038 (CA Fed. 1985) (harmonizing review standards).
Finally, the Circuit reasons that its stricter court/court review will produce better agency factfinding. It says that the standard encourages the creation of "administrative records that more fully describe the metes and bounds of the patent grant" and "help avoid situations where board fact finding on matters such as anticipation or the factual inquiries underlying obviousness become virtually unreviewable." 142 F. 3d, at 1458. Neither the Circuit nor its supporting amici, however, have explained convincingly why direct review of the PTO's patent denials demands a stricter factrelated review standard than is applicable to other agencies. Congress has set forth the appropriate standard in the AP A. For the reasons stated, we have not found circumstances that justify an exception.
For these reasons, the judgment of the Federal Circuit is reversed. We remand the case for further proceedings consistent with this opinion.
APPENDIX TO OPINION OF THE COURT
Review of 89 Pre-APA CCPA Patent Cases Reciting "Clear" or "Manifest" Error Standard
Cases Referring to both Technical Complexity/Agency Expertise and the Agreement (Disagreement) Within the Agency
Stern v. Schroeder, 17 C. C. P. A. 670, 674, 36 F.2d 515, 517 (1929)
In re Ford, 17 C. C. P. A. 893, 894, 38 F.2d 525, 526 (1930) In re Demarest, 17 C. C. P. A. 904, 906, 38 F.2d 895, 896 (1930)
In re Wietzel, 17 C. C. P. A. 1079, 1082, 39 F.2d 669, 671 (1930)
In re Anhaltzer, 18 C. C. P. A. 1181, 1184, 48 F.2d 657, 658 (1931)
Appendix to opinion of the Court
Dorer v. Moody, 18 C. C. P. A. 1188, 1190, 48 F.2d 388, 389 (1931)
In re Hornsey, 18 C. C. P. A. 1222, 1224, 48 F.2d 911, 912 (1931)
Rowe v. Holtz, 19 C. C. P. A. 970, 974, 55 F.2d 468, 470471 (1932)
In re Fessenden, 19 C. C. P. A. 1048, 1050-1051, 56 F.2d 669, 670 (1932)
Martin v. Friendly, 19 C. C. P. A. 1181, 1182-1183,58 F.2d 421, 422 (1932)
In re Dubilier, 20 C. C. P. A. 809, 815, 62 F.2d 374, 377 (1933)
In re Alden, 20 C. C. P. A. 1083, 1084-1085, 65 F.2d 136, 137 (1933)
Farmer v. Pritchard, 20 C. C. P. A. 1096, 1101, 65 F.2d 165, 168 (1933)
In re Pierce, 20 C. C. P. A. 1170, 1175, 65 F.2d 271, 274 (1933)
Angell v. Morin, 21 C. C. P. A. 1018, 1024, 69 F.2d 646, 649 (1934)
Daley v. Trube, 24 C. C. P. A. 964, 971, 88 F.2d 308, 312 (1937)
Coast v. Dubbs, 24 C. C. P. A. 1023, 1031-1032,88 F.2d 734, 739 (1937)
Bryson v. Clarke, 25 C. C. P. A. 719, 721, 92 F.2d 720, 722 (1937)
Brand v. Thomas, 25 C. C. P. A. 1053, 1055, 96 F.2d 301, 302 (1938)
Creed v. Potts, 25 C. C. P. A. 1084, 1089, 96 F.2d 317, 321 (1938)
In re Cassidy, 25 C. C. P. A. 1282, 1285, 97 F.2d 93, 95 (1938)
Krebs v. Melicharek, 25 C. C. P. A. 1362, 1365-1366, 97 F.2d 477, 479 (1938)
Parker v. Ballantine, 26 C. C. P. A. 799,804, 101 F.2d 220, 223 (1939) (disagreement)
Reed v. Edwards, 26 C. C. P. A. 901, 904, 101 F.2d 550, 552 (1939)
Hill v. Casler, 26 C. C. P. A. 930, 932, 102 F.2d 219, 221 (1939)
Tears v. Robinson, 26 C. C. P. A. 1391, 1392, 104 F.2d 813, 814 (1939)
In re Bertsch, 27 C. C. P. A. 760, 763-764, 107 F.2d 828, 831 (1939)
In re Wuertz, 27 C. C. P. A. 1039, 1046, 110 F.2d 854, 857 (1940)
In re Kaplan, 27 C. C. P. A. 1072, 1075, 110 F.2d 670, 672 (1940)
Prahl v. Redman, 28 C. C. P. A. 937, 940, 117 F.2d 1018, 1021 (1941)
In re Bertsch, 30 C. C. P. A. 813, 815-816, 132 F.2d 1014, 1016 (1942)
In re Stacy, 30 C. C. P. A. 972, 974, 135 F.2d 232,233 (1943) Poulsen v. McDowell, 31 C. C. P. A. 1006, 1011, 142 F.2d 267, 270 (1944)
Pinkerton v. Stahly, 32 C. C. P. A. 723, 728, 144 F.2d 881, 885 (1944)
Cases Referring to Technical Complexity/Agency Expertise In re Engelhardt, 17 C. C. P. A. 1244, 1251, 40 F.2d 760, 764 (1930)
In re McDonald, 18 C. C. P. A. 1099, 1102, 47 F.2d 802, 804 (1931)
In re Hermans, 18 C. C. P. A. 1211, 1212, 48 F.2d 386, 387 (1931)
In re Batcher, 19 C. C. P. A. 1275, 1278, 59 F.2d 461, 463 (1932)
In re Carlton, 27 C. C. P. A. 1102, 1105, 111 F.2d 190, 192 (1940)
Farnsworth v. Brown, 29 C. C. P. A. 740, 749,124 F.2d 208, 214 (1941)
Appendix to opinion of the Court
In re Ubbelhode, 29 C. C. P. A. 1042, 1046, 128 F.2d 453, 456 (1942)
In re Cohen, 30 C. C. P. A. 876, 880, 133 F.2d 924, 926 (1943)
In re Ruzicka, 32 C. C. P. A. 1165, 1169, 150 F.2d 550, 553 (1945)
In re Allbright, 33 C. C. P. A. 760, 764, 152 F.2d 984, 986 (1946)
Cases Referring to Agreement Within the Agency
Beidler v. Caps, 17 C. C. P. A. 703, 705, 36 F.2d 122, 123 (1929)
Stern v. Schroeder, 17 C. C. P. A. 690, 696-697, 36 F.2d 518, 521-522 (1929)
Janette v. Folds, 17 C. C. P. A. 879, 881, 38 F.2d 361, 362 (1930)
In re Moulton, 17 C. C. P. A. 891, 892, 38 F.2d 359, 360 (1930)
In re Banner, 17 C. C. P. A. 1086, 1090, 39 F.2d 690, 692 (1930)
In re Walter, 17 C. C. P. A. 982, 983, 39 F.2d 724 (1930) Pengilly v. Copeland, 17 C. C. P. A. 1143, 1145, 40 F.2d 995, 996 (1930)
Thompson v. Pettis, 18 C. C. P. A. 755, 757, 44 F.2d 420, 421 (1930)
In re Kochendorfer, 18 C. C. P. A. 761, 763, 44 F.2d 418, 419 (1930)
In re Dickerman, 18 C. C. P. A. 766, 768, 44 F.2d 876, 877 (1930)
Bennett v. Fitzgerald, 18 C. C. P. A. 1201, 1202, 48 F.2d 917, 918 (1931)
In re Doherty, 18 C. C. P. A. 1278, 1280, 48 F.2d 952, 953 (1931)
In re Murray, 19 C. C. P. A. 766, 767-768, 53 F.2d 540, 541 (1931)
In re Breer, 19 C. C. P. A. 929, 931, 55 F.2d 485, 486 (1932)
Robbins v. Steinbart, 19 C. C. P. A. 1069, 1072, 57 F.2d 378, 379 (1932)
Henry v. Harris, 19 C. C. P. A. 1092, 1096-1097, 56 F.2d 864, 866 (1932)
Fageol v. Midboe, 19 C. C. P. A. 1117, 1122, 56 F.2d 867, 870 (1932)
Gamble v. Church, 19 C. C. P. A. 1145, 1146, 57 F.2d 761, 762 (1932)
Thompson v. Fawick, 20 C. C. P. A. 953, 956, 64 F.2d 125, 127 (1933)
Evans v. Clocker, 20 C. C. P. A. 956, 960, 64 F.2d 137, 139 (1933)
In re Bloch, 20 C. C. P. A. 1180, 1183, 65 F.2d 268, 269 (1933)
In re Snyder, 21 C. C. P. A. 720, 722, 67 F.2d 493, 495 (1933) Osgood v. Ridderstrom, 21 C. C. P. A. 1176, 1182, 71 F.2d 191, 195 (1934)
Urschel v. Crawford, 22 C. C. P. A. 727, 730, 73 F.2d 510, 511 (1934)
Marine v. Wright, 22 C. C. P. A. 946, 948-949, 74 F.2d 996, 997 (1935)
Berman v. Rondelle, 22 C. C. P. A. 1049, 1052, 75 F.2d 845, 847 (1935)
Tomlin v. Dunlap, 24 C. C. P. A. 1108, 1114, 88 F.2d 727, 731 (1937)
Lasker v. Kurowski, 24 C. C. P. A. 1253, 1256, 90 F.2d 132, 134 (1937)
In re Taylor, 25 C. C. P. A. 709, 711, 92 F.2d 705, 706 (1937) In re Adamson, 25 C. C. P. A. 726, 729-730, 92 F.2d 717, 720 (1937)
Adams v. Stuller, 25 C. C. P. A. 865, 870, 94 F.2d 403, 406 (1938)
Ellis v. Maddox, 25 C. C. P. A. 1045, 1053, 96 F.2d 308, 314 (1938)
Kauffman v. Etten, 25 C. C. P. A. 1127, 1134, 97 F.2d 134, 139 (1938)
Kindelmann v. Morsbach, 25 C. C. P. A. 1344, 1349, 97 F.2d 796, 799-800 (1938)
King v. Young, 26 C. C. P. A. 762, 771, 100 F.2d 663, 670 (1938)
Meuer v. Schellenger, 26 C. C. P. A. 1430, 1434, 104 F.2d 949, 952 (1939)
McBride v. Teeple, 27 C. C. P. A. 961, 972, 109 F.2d 789, 797, cert. denied, 311 U. S. 649 (1940)
Vickery v. Barnhart, 28 C. C. P. A. 979, 982, 118 F.2d 578, 581 (1941)
Shumaker v. Paulson, 30 C. C. P. A. 1136, 1138, 136 F.2d 686, 688 (1943)
Paulson v. Hyland, 30 C. C. P. A. 1150, 1152, 136 F.2d 695, 697 (1943)
Dreyer v. Haffcke, 30 C. C. P. A. 1278, 1280, 137 F.2d 116, 117 (1943)
Cases Referring to Neither Technical Complexity/Agency Expertise nor Agreement Within the Agency
In re Schmidt, 26 C. C. P. A. 773, 777, 100 F.2d 673, 676 (1938)
Hamer v. White, 31 C. C. P. A. 1186, 1189, 143 F.2d 987, 990 (1944)
Kenyon v. Platt, 33 C. C. P. A. 748, 752, 152 F.2d 1006, 1009 (1946)
Beall v. Ormsby, 33 C. C. P. A. 959, 967, 154 F.2d 663, 668 (1946)
CHIEF JUSTICE REHNQUIST, with whom JUSTICE KENNEDY and JUSTICE GINSBURG join, dissenting.
The issue in this case is whether, at the time of the enactment of the Administrative Procedure Act (APA or Act) over 50 years ago, judicial review of factfinding by the Patent and Trademark Office (PTO) under the "clearly erroneous" standard was an "additional requiremen[t] ... recognized by law." 5 U. S. C. § 559. It is undisputed that, until to day's decision,
both the patent bench and the patent bar had concluded that the stricter "clearly erroneous" standard was indeed such a requirement placed upon the PTO. * Agency factfinding was thus reviewed under this stricter standard; in my view, properly so, since the APA by its plain text was intended to bring some uniformity to judicial review of agencies by raising the minimum standards of review and not by lowering those standards which existed at the time. Section 12 of the AP A, which was ultimately codified as § 559, provided that "[n]othing in this Act shall be held to diminish the constitutional rights of any person or to limit or repeal additional requirements imposed by statute or otherwise recognized by law." Pub. L. 404, 79th Cong., 60 Stat. 244. As a result, we must decide whether the "clearly erroneous" standard was indeed otherwise recognized by law in 1946.
This case therefore turns on whether the 89 or so cases identified by the Court can be read as establishing a requirement placed upon agencies that was more demanding than the uniform minimum standards created by the AP A. In making this determination, I would defer, not to agencies in general as the Court does today, but to the Court of Appeals for the Federal Circuit, the specialized Article III court charged with review of patent appeals. In this case the unanimous en bane Federal Circuit and the patent bar both agree that these cases recognized the "clearly erroneous" standard as an "additional requirement" placed on the PTO beyond the AP A's minimum procedures. I see no reason to reject their sensible and plausible resolution of the issue.
Nor do I agree with the Court, ante, at 154-155, that either the plain language of § 559 or the original § 12 impose any sort of "clear statement rule" on the common law. Sec-
*It appears that even the PTO acquiesced in this interpretation for almost 50 years after the enactment of the APA. See Brieffor Pharmaceutical Research and Manufacturers of America as Amicus Curiae 7, and n. 13 (the PTO first argued for the applicability of the APA's standards of review to its patentability factfinding before the Federal Circuit in 1995).
tion 12 of the AP A expressly stated that requirements which predated the AP A and were "otherwise recognized by law" were unaffected by the Act. If Congress had meant "otherwise recognized by law" to mean "clearly recognized by law," it certainly could have said so, but did not. I also reject the notion that § 559's separate textual requirement that subsequent statutes superseding or modifying the AP A must do so "expressly," 5 U. s. C. § 559, should be read to impose a nontextual clear statement rule for the antecedent commonlaw requirements that the AP A supplemented. There is no tension whatsoever between the goals of preserving more rigorous common-law requirements at the time of enactment and ensuring that future statutes would not repeal by implication the AP A's uniform supplementary procedures.
I therefore dissent for the reasons given by the Court of Appeals.