Critical Mass Energy Project v. Nuclear Regulatory Commission

Opinion for the court en banc filed by Circuit Judge BUCKLEY, in which Circuit Judges SILBERMAN, WILLIAMS, D.H. GINSBURG, SENTELLE, HENDERSON, and RANDOLPH join.

Concurring opinion filed by Circuit Judge RANDOLPH, in which Circuit Judges SILBERMAN and SENTELLE join.

Dissenting opinion filed by Circuit Judge RUTH BADER GINSBURG, in which Chief Judge MIKVÁ and Circuit Judges WALD and HARRY T. EDWARDS join.

BUCKLEY, Circuit Judge, with whom SILBERMAN, STEPHEN F. WILLIAMS, D.H. GINSBURG, SENTELLE, KAREN LeCRAFT HENDERSON, and RANDOLPH, Circuit Judges, join:

Appellant seeks the release of certain reports that have been provided to the Nuclear Regulatory Commission by the Institute of Nuclear Power Operations on the understanding that they will be treated as confidential. In granting the petition to rehear the case en banc, we agreed to reconsider a seventeen-year-old decision, National Parks and Conservation Ass’n v. Morton, 498 F.2d 765 (D.C.Cir.1974), in which we established a two-part test for determining when financial or commercial information in the Government’s possession is to be treated as confidential under Exemption 4 of the Freedom of Information Act. We reaffirm the test but confine it to information that persons are required to provide the Government. We hold that where, as here, the information sought is given to the Government voluntarily, it will be treated as confidential under Exemption 4 if it is of a kind that the provider would not customarily make available to the public.

I. BACKGROUND

A. Legal Framework

When Congress enacted the Freedom of Information Act (“FOIA”), 5 U.S.C. § 552 (1988), it “set[] forth a policy of broad disclosure of Government documents in order to ensure ‘an informed citizenry, vital to the functioning of a democratic society.’ ” FBI v. Abramson, 456 U.S. 615, 621, 102 S.Ct. 2054, 2059, 72 L.Ed.2d 376 (1982) (citation omitted). At the same time, however, “Congress realized that legitimate governmental and private interests could be harmed by release of certain types of information.” Id. Balancing these private and public interests, Congress enacted nine exemptions to FOIA. Exemption 4, the provision at issue here, shields from disclosure “trade secrets and commercial or financial information obtained from a person and privileged or confidential.” 5 U.S.C. § 552(b)(4).

In the years immediately following the enactment of FOIA, we had frequent occasion to put flesh on the exemption. See, e.g., Sterling Drug, Inc. v. FTC, 450 F.2d 698 (D.C.Cir.1971); Soucie v. David, 448 F.2d 1067, 1078 (D.C.Cir.1971); Grumman Aircraft Engineering Corp. v. The Renegotiation Board, 425 F.2d 578, 580-81 (D.C.Cir.1970); Bristol-Myers Co. v. FTC, 424 F.2d 935, 938-39 (D.C.Cir.), cert. denied, 400 U.S. 824, 91 S.Ct. 46, 27 L.Ed.2d 52 (1970). In Sterling Drug, we considered a FOIA request for records submitted to the Federal Trade Commission. In order to determine the purpose of Exemption 4, we examined the following passage from a Senate committee report:

This exception is necessary to protect the confidentiality of information which is obtained by the Government through *873questionnaires or other inquiries, but which would customarily not be released to the public by the person from whom it was obtained.

Sterling Drug, 450 F.2d at 709 (quoting S.Rep. No. 813, 89th Cong., 2d Sess. 9 (1964) U.S.Code Cong. & Admin.News 1966, p. 2418 (“Senate Report”)). We then concluded that the requested information was protected by the exemption because it was of “the type ‘which would customarily not be released to the public by the person from whom it was obtained.’ ” Id.

Three years later, we decided National Parks. The plaintiff in that case sought disclosure of audits and other financial materials that the National Park Service required its concessioners to submit. We reversed the district court’s grant of summary judgment in favor of the agency based on Exemption 4. After observing that the statute did not define “confidential,” we noted that we had been guided, in earlier decisions, by the “would customarily not be released to the public” passage from the Senate Report. National Parks, 498 F.2d at 766. We found, however, that whether the person providing the information would customarily not have released it “is not the only relevant inquiry in determining whether that information is ‘confidential.’ ... A court must also be satisfied that non-disclosure is justified by the legislative purpose which underlies the exemption.” Id. at 767. We then stated that “[i]n general, the various [FOIA] exemptions ... serve two interests — that of the Government in efficient operation and that of persons supplying certain kinds of information in maintaining its secrecy.” Id.

We determined that Exemption 4 serves both of these purposes. As to the governmental interest, we noted that “[ujnless persons having necessary information can be assured that it will remain confidential, they may decline to cooperate with officials^] and the ability of the Government to make intelligent, well informed decisions will be impaired.” Id. After discussing the legislative history, we concluded that “[t]his exemption is intended to encourage individuals to provide certain kinds of confidential information to the Government[.]” Id. at 768 (quoting Soucie, 448 F.2d at 1078).

Turning to the private interests, we noted that the exemption “protects persons who submit financial or commercial data to government agencies from the competitive disadvantages which would result from its publication.” Id. We also recognized

a twofold justification for the exemption of commercial material: (1) encouraging cooperation by those who are not obliged to provide information to the government and (2) protecting the rights of those who must.

Id. at 769. We concluded that the legislative history “firmly supports the inference that section 552(b)(4) is intended for the benefit of persons who supply information as well as the agencies which gather it.” Id. at 770. We then offered what has come to be known as the National Parks “two-part test”:

[Cjommercial or financial matter is “confidential” ... if disclosure of the information is likely ... either ... (1) to impair the Government’s ability to obtain necessary information in the future; or (2) to cause substantial harm to the competitive position of the person from whom the information was obtained.

Id. at 770 (footnote omitted). At the same time, we reserved the question whether the impairment of other governmental interests might be implicated in the exemption. See id. at 770 n. 17 (“We express no opinion as to whether other governmental interests are embodied in this exemption.”).

On the facts of National Parks, we determined that, although the information being sought was customarily withheld from the public, “the Government ha[d] no apparent interest in preventing disclosure of the matter” as it was “supplied to the Park Service pursuant to statute.” Id. at 770. We then held that

“[s]ince the concessioners are required to provide this financial information to the government, there is presumably no danger that public disclosure will impair the ability of the Government to obtain this information in the future[,]”

*874id. (emphasis in original), and remanded for further findings on the question of competitive harm. See id. at 770-71.

B. Facts and Proceedings Below

This case involves a dispute between Critical Mass Energy Project (“CMEP”) and the Nuclear Regulatory Commission (“NRC”) over access to safety reports prepared by the Institute for Nuclear Power Operations (“INPO”) and voluntarily transmitted to the NRC on the condition that the agency will not release the information to other parties without INPO’s consent. INPO was formed after the 1979 Three Mile Island accident to promote safety and reliability in the operation of nuclear power plants. INPO is a nonprofit corporation whose membership includes all operators of nuclear power plants in the United States.

One of INPO’s principal programs is the Significant Event Evaluation and Information Network (“SEE-IN”), a system for collecting, analyzing, and distributing information concerning the construction and operation of nuclear facilities. Compilation of these reports requires the solicitation of candid comments and evaluations from nuclear power plant employees. The reports are distributed on a voluntary basis to INPO members, certain other participants in the nuclear industry, and the NRC pursuant to the explicit understanding that they are not to be disclosed to additional persons without INPO’s consent.

In 1984, CMEP asked the NRC, pursuant to FOIA, to provide it with copies of the INPO reports. The NRC denied the request, finding that they contained confidential commercial information and were therefore protected from disclosure by Exemption 4. CMEP then brought suit in district court challenging the NRC’s determination.

District Judge Thomas Penfield Jackson granted the NRC’s motion for summary judgment, concluding that the documents were both commercial and confidential and therefore exempt under section 552(b)(4) as interpreted by this court in National Parks. See Critical Mass Energy Project v. NRC, 644 F.Supp. 344, 346-47 (D.D.C. 1986). The judge acknowledged that while INPO itself was not subject to NRC regulation, its members were; therefore, the agency could compel them to produce the reports. Nevertheless, deeming it “preferable to have the INPO reports furnished to the NRC voluntarily, rather than delivered up under compulsion in circumstances less conducive to candor, accuracy, and timeliness,” Judge Jackson held that “the NRC’s ability to acquire the information imparted by the INPO reports will likely be impaired if it is not permitted to honor its commitment to INPO to keep them confidential.” Id. at 347.

On appeal, a divided panel reversed and remanded for further proceedings. Critical Mass Energy Project v. NRC, 830 F.2d 278 (D.C.Cir.1987) (“Critical Mass /”). We agreed that the information sought was commercial. With respect to the question of confidentiality, we held that the agency must meet two requirements: First, it must establish that the information sought “would customarily not be released to the public by the person from whom it was obtained” — a condition that the NRC had met in this case; and second, it must show that “disclosure [would] harm a specific interest that Congress sought to protect by enacting the exemption.” Id. at 281 (citation omitted). We observed that National Parks had identified two of those interests: (1) the Government’s need to have access to commercial and financial data and (2) the provider’s need to be protected from competitive harm. Id. at 281-82.

Applying the National Parks test, we agreed that the INPO reports were not “customarily released to the public,” id. at 282, but found insufficient evidence that the Government’s ability to secure the information would be impaired. See id. at 282-86. We concluded that to prove impairment, the NRC would have to show either that cessation of INPO’s voluntary submission of the reports would deprive the NRC of the information they contained or that the employment of alternative means of obtaining the reports would create a significant risk that their value would *875be diminished. Id. at 283. We held that the NRC could not establish the first point, as the agency had ample authority to compel production of the INPO reports from its licensees, id. at 283-84, and remanded for further findings on the second. Id. at 286. We also held that the interests identified in National Parks (protection against governmental impairment and competitive injury) were not exclusive and instructed the district court to consider efficiency and effectiveness concerns in conducting its Exemption 4 analysis. Id. at 286-87.

After remand, INPO intervened in the litigation. Cross-motions for summary judgment were filed, and the district court granted the defendants’ motion based on a determination that they had shown that release of the reports would harm the governmental interest in efficiency and cooperation with INPO. See Critical Mass Energy Project v. NRC, 731 F.Supp. 554, 557 (D.D.C.1990).

On appeal, we rejected the district court’s rationale but remanded for further findings on the effect of disclosure on the quality of the reports. Critical Mass Energy Project v. NRC, 931 F.2d 939, 943-44, 947 (D.C.Cir.1991) (“Critical Mass II ”). In particular, we asked for “testimony from working-level employees regarding the importance of confidential treatment of INPO reports to their willingness to speak to INPO analysts.” Id. at 946.

Judge Randolph, joined by Judge Williams, wrote a separate concurring opinion. While recognizing that the court was bound by them, he questioned our decisions in National Parks and Critical Mass I. He was particularly critical of the two-part test announced in National Parks for having “added — or, as has been said, ‘fabricated, out of whole cloth’ — an additional requirement that must be met before confidential information is exempt from disclosure.” Id. at 947 (quoting Note, Trade Secrets and the Fifth Amendment, 54 U.Chi.L.Rev. 334, 364 (1987)).

In response to petitions filed by the NRC and INPO, we vacated the opinion in Critical Mass II and ordered that the case be reheard en banc. See Critical Mass Energy Project v. NRC, 942 F.2d 799 (D.C.Cir.1991). In granting the petitions, we undertook not only to reexamine the decisions in Critical Mass I and II, but also to “reconsider the definition of ‘confidential’ set forth in National Parks ... for purposes of applying Freedom of Information Act Exemption 4.” Critical Mass Energy Project v. NRC, No. 90-5120, Order (D.C.Cir. Oct. 4, 1991).

II. DISCUSSION

A. National Parks Reexamined

In challenging the definition of “confidential” presented in National Parks, the NRC and INPO ask us to set aside circuit precedent of almost twenty years’ standing. In obedience to the principle of stare decisis, we reaffirm the definition but correct some misunderstandings as to its scope and application.

1. Stare Decisis

“[T]he doctrine of stare decisis is of fundamental importance to the rule of law.” Welch v. Texas Dep’t of Highways and Public Transportation, 483 U.S. 468, 494, 107 S.Ct. 2941, 2957, 97 L.Ed.2d 389 (1987). The Supreme Court has held that “any departure from the doctrine ... demands special justification.” Arizona v. Rumsey, 467 U.S. 203, 212, 104 S.Ct. 2305, 2310, 81 L.Ed.2d 164 (1984). Furthermore, “the burden borne by the party advocating the abandonment of an established precedent is greater where the Court is asked to overrule a point of statutory construction ... [f]or here, unlike in the context of constitutional interpretation, ... Congress remains free to alter what we have done.” Patterson v. McLean Credit Union, 491 U.S. 164, 172-73, 109 S.Ct. 2363, 2370, 105 L.Ed.2d 132 (1989) (citations omitted).

In Patterson, the Supreme Court articulated in some detail the circumstances in which it may find sufficient justification for overturning a statutory precedent:

In cases where statutory precedents have been overruled, the primary reason for the Court’s shift in position has been the intervening development of the law, *876through either the growth of judicial doctrine or further action taken by Congress. Where such changes have removed or weakened the conceptual underpinnings from the prior decision, or where the later law has rendered the decision irreconcilable with competing legal doctrines or policies, the Court has not hesitated to overrule an earlier decision. ...
Another traditional justification for overruling a prior case is that a precedent may be a positive detriment to coherence and consistency in the law, either because of inherent confusion created by an unworkable decision, or because the decision poses a direct obstacle to the realization of important objectives embodied in other laws.

Id. at 173, 109 S.Ct. at 2370-71 (citations omitted).

Circuit courts of appeal, of course, play a different role in the federal system than the Supreme Court, and this is reflected in certain differences in the manner in which the principle of stare decisis is applied to circuit precedent. See United States v. Aguon, 851 F.2d 1158, 1172-76 (9th Cir.1988) (Reinhardt, J., concurring). Circuit courts do not establish the ultimate judicial precedent for the application of a federal statute; and because there are thirteen of them, they will at times interpret the same statute in different ways. Moreover, circuit precedent is generally established by the majority vote of just three circuit judges. Their decisions, nevertheless, bind the circuit “unless and until overturned by the court en banc or by Higher Authority.” Save Our Cumberland Mountains, Inc. v. Hodel, 826 F.2d 43, 54 (D.C.Cir.1987) (Ginsburg, Ruth B., J., concurring).

Thus, in addition to considering the factors outlined in Patterson, a circuit court may reexamine its own established interpretation of a statute if it finds that other circuits have persuasively argued a contrary construction. An en banc court may also set aside its own precedent if, on reexamination of an earlier decision, it decides that the panel’s holding on an important question of law was fundamentally flawed. For example, in Save Our Cumberland Mountains, Inc. v. Hodel, 857 F.2d 1516 (D.C.Cir.1988) (en banc), we overturned a four-year-old precedent because we concluded that the rule established in that case was inconsistent with the intent of Congress and that the panel had misapplied a Supreme Court decision. Id. at 1524. See also Groves v. Ring Screw Works, Ferndale Fastener Div., — U.S.-,-n. 8, 111 S.Ct. 498, 501 n. 8, 112 L.Ed.2d 508 (1990) (given intercircuit conflict and panel’s doubt about correctness of circuit precedent, “it might have been appropriate for the panel to request a rehearing en banc ”); Aguon, 851 F.2d at 1160-72 (en banc) (overruling circuit precedent in light of conflict with subsequent Second Circuit opinion).

Appellees NRC and INPO have failed to demonstrate any of the considerations that would justify our overturning the National Parks test. We note, first, the widespread acceptance of National Parks by other circuits. To date, seven have adopted its test of confidentiality; none has rejected it. See, e.g., 9 to 5 Org. for Women Office Workers v. Bd. of Governors of the Fed. Res. System, 721 F.2d 1, 7-10 (1st Cir.1983); American Airlines, Inc. v. Nat’l Mediation Bd., 588 F.2d 863, 871 (2d Cir.1978); Acumenics Research & Technology v. Dep’t of Justice, 843 F.2d 800, 807 (4th Cir.1988); Sharyland Water Supply Corp. v. Block, 755 F.2d 397, 399 (5th Cir.), cert. denied, 471 U.S. 1137, 105 S.Ct. 2678, 86 L.Ed.2d 697 (1985); General Elec. Co. v. NRC, 750 F.2d 1394, 1402-03 (7th Cir.1984); Pacific Architects and Engineers Inc. v. Dep’t of State, 906 F.2d 1345, 1347 (9th Cir.1990); Anderson v. HHS, 907 F.2d 936, 946 (10th Cir.1990). Far from being overtaken by the tide of recent judicial developments, National Parks, it seems, has ridden its crest.

Nor have appellees pointed to any subsequent action in Congress that would tend to “remove[] or weaken[] the conceptual underpinnings” of National Parks. Patterson, 491 U.S. at 173, 109 S.Ct. at 2370. To the contrary, Congress has taken cognizance of the case in enacting subsequent *877legislation. See, e.g., CNA Fin. Corp. v. Donovan, 830 F.2d 1132, 1153 n. 146 (D.C.Cir.1987) (“[W]hen both Houses of Congress provided an exemption [to the Government in the Sunshine Act] ... they accepted the National Parks standard as appropriate.”), cert. denied, 485 U.S. 977, 108 S.Ct. 1270, 99 L.Ed.2d 481 (1988).

Finally, we find that National Parks has not proven so “unworkable” in practice as to constitute a “positive detriment to coherence and consistency in the law” or a “direct obstacle to the realization of important objectives embodied in other laws.” Patterson, 491 U.S. at 173, 109 S.Ct. at 2371. Nor has there been a showing that the test, as applied in National Parks, will frustrate Congress’s purposes in enacting Exemption 4. Appellees argue, nevertheless, that the decision has proven unworkable in its everyday application by agencies and the courts. They suggest that the impracticability of the “governmental impairment” prong is evidenced by the history of this case. They claim similar difficulties with the “competitive injury” prong. See, e.g., Greenberg v. Food & Drug Admin., 803 F.2d 1213 (D.C.Cir.1986) (reversing summary judgment for further proceedings on the economic effectiveness of other means of compiling private customer lists); Worthington Compressors, Inc. v. Costle, 662 F.2d 45 (D.C.Cir.1981) (reversing summary judgment for further proceedings on question of discoverability of secret information sought by reverse engineering).

While these decisions illustrate the difficulties that can arise wherever judicial lines are drawn, they fall short of demonstrating that National Parks is so flawed that we would be justified in setting it aside. Moreover, we believe that the categorical rule we announce today will greatly simplify the application of Exemption 4 in a significant number of cases.

Having examined these and other arguments in favor of overturning National Parks, we conclude that none justifies the abandonment of so well established a precedent. Whatever our individual opinions as to the merits of its two-part test, we accept the wisdom of Justice Brandeis’s observation, some sixty years ago, that “[sjtare decisis is usually the wise policy, because in most matters it is more important that the applicable rule of law be settled than that it be settled right.” Burnet v. Coronado Oil & Gas Co., 285 U.S. 393, 406, 52 S.Ct. 443, 447, 76 L.Ed. 815 (1932) (Brandeis, J., dissenting).

2. The National Parks test

The National Parks two-part test was presented as a summary of its discussion of Exemption 4. See National Parks, 498 F.2d at 770 (“To summarize_”). The test must therefore be read in a manner consistent with the observations and conclusions that preceded it. We quote the most pertinent of these:

[O]ur decisions concerning this exemption have been guided by the following passage from the Senate Report, particularly the italicized portion:
This exception is necessary to protect the confidentiality of information which is obtained by the Government ..., but which would customarily not be released to the public by the person from whom it was obtained.

Id. at 766 (emphasis in original).

The “financial information” exemption recognizes the need of government policymakers to have access to commercial and financial data. Unless persons having necessary information can be assured that it will remain confidential, they may decline to cooperate with officials and the ability of the Government to make intelligent, well informed decisions will be impaired.

Id. at 767 (emphasis added).

Apart from encouraging cooperation with the Government by persons having information useful to officials, [Exemption 4] serves another distinct but equally important purpose. It protects persons who submit financial or commercial data to government agencies from the competitive disadvantages which would result from its publication.

Id. at 768 (emphasis added).

[There is] a twofold justification for the exemption of commercial material: (1) *878encouraging cooperation by those who are not obliged to provide information to the government and (2) protecting the rights of those who must.

Id. at 769 (emphasis added).

In summarizing these various purposes and justifications, we formulated the now familiar two-part test that defined as “confidential” any financial or commercial information whose disclosure would be likely either “(1) to impair the Government’s ability to obtain necessary information in the future; or (2) to cause substantial harm to the competitive position of the person from whom the information was obtained.” Id. at 770 (footnote omitted). In applying this test to the facts of National Parks, we held that because

the concessioners [were] required to provide this financial information ..., there is presumably no danger that public disclosure will impair the ability of the Government to obtain this information in the future.

Id. (emphasis in original). Then, because the record was incomplete as to the competitive harm that might be suffered by the concessioners on the release of the information, we remanded for further findings on that question.

While we indicated that the governmental interest is unlikely to be implicated where the production of information is compelled, we have since pointed out that there are circumstances in which disclosure could affect the reliability of such data. See Washington Post Co. v. HHS, 690 F.2d 252, 268-69 (D.C.Cir.1982) (possible effect of disclosure on accuracy of statements filed by consultants). Thus, when dealing with a FOIA request for information the provider is required to supply, the governmental impact inquiry will focus on the possible effect of disclosure on its quality.

When a FOIA request is made for information that is furnished on a voluntary basis, however, we have identified a different aspect of the governmental interest in securing confidential information. As is apparent from the language we have italicized in the passages quoted above, the purpose served by the exemption in such instances is that of “encouraging cooperation with the Government by persons having information useful to officials,” National Parks, 498 F.2d at 768. Moreover, we have taken note of the probable consequences of a breach of confidence by the Government:

Unless persons having necessary information can be assured that it will remain confidential, they may decline to cooperate with officials[,] and the ability of the Government to make intelligent, well informed decisions will be impaired.

Id. at 767. Thus, when information is obtained under duress, the Government’s interest is in ensuring its continued reliability; when that information is volunteered, the Government’s interest is in ensuring its continued availability.

A distinction between voluntary and compelled information must also be made when applying the “competitive injury” prong. In the latter case, there is a presumption that the Government’s interest is not threatened by disclosure because it secures the information by mandate; and as the harm to the private interest (commercial disadvantage) is the only factor weighing against FOIA’s presumption of disclosure, that interest must be significant. Where, however, the information is provided to the Government voluntarily, the presumption is that its interest will be threatened by disclosure as the persons whose confidences have been betrayed will, in all likelihood, refuse further cooperation. In those cases, the private interest served by Exemption 4 is the protection of information that, for whatever reason, “would customarily not be released to the public by the person from whom it was obtained” — to use the formulation adopted by this court in Sterling Drug, 450 F.2d at 709 (quoting Senate Report at 9). Because Critical Mass I failed to take these distinctions into account, its holding would frustrate Congress’s purpose of “encouraging cooperation with the Government by persons having information useful to officials,” National Parks, 498 F.2d at 768. See Critical Mass I, 830 F.2d at 288 (Buckley, J., concurring in part and dissenting in part); *879Critical Mass II, 931 F.2d at 948 (Randolph, J., concurring).

It should be evident from this review that the two interests identified in the National Parks test are not exclusive. Although we overrule our decision in Critical Mass I, we note that the panel there adopted the First Circuit’s conclusion that the exemption also protects a governmental interest in administrative efficiency and effectiveness. See Critical Mass I, 830 F.2d at 286 (citing 9 to 5 Org. for Women Office Workers, 721 F.2d at 11). And today, of course, we recognize a private interest in preserving the confidentiality of information that is provided the Government on a voluntary basis. We offer no opinion as to whether any other governmental or private interest might also fall within the exemption’s protection.

B. Application of Exemption 4 to Information Provided on a Voluntary Basis

The Supreme Court has encouraged the development of categorical rules whenever a particular set of facts will lead to a generally predictable application of FOIA. See United States Dep’t of Justice v. Reporters Comm. for Freedom of the Press, 489 U.S. 749, 109 S.Ct. 1468, 103 L.Ed.2d 774 (1989). The Court emphasized that

establishing a discrete category of exempt information, [will] implement ] the congressional intent to provide “workable” rules_ Only by construing the Exemption to provide a categorical rule can the Act’s purpose of expediting disclosure by means of workable rules be furthered.

Id. at 779, 109 S.Ct. at 1485 (quoting FTC v. Grolier, Inc., 462 U.S. 19, 27-28, 103 S.Ct. 2209, 2214-15, 76 L.Ed.2d 387 (1983) (emphasis in original)). We also take cognizance of the Court’s “practical approach when ... confronted with an issue of interpretation of [FOIA].” John Doe Agency v. John Doe Corp., 493 U.S. 146, 157, 110 S.Ct. 471, 478, 107 L.Ed.2d 462 (1989).

The circumstances of this case lend themselves to categorical treatment. It is a matter of common sense that the disclosure of information the Government has secured from voluntary sources on a confidential basis will both jeopardize its continuing ability to secure such data on a cooperative basis and injure the provider’s interest in preventing its unauthorized release. Accordingly, while we reaffirm the National Parks test for determining the confidentiality of information submitted under compulsion, we conclude that financial or commercial information provided to the Government on a voluntary basis is “confidential” for the purpose of Exemption 4 if it is of a kind that would customarily not be released to the public by the person from whom it was obtained.

We make three observations about this test. First, it is objective. As is the case with any claim under FOIA, see 5 U.S.C. § 552(a)(4)(B) (1988), the agency invoking Exemption 4 must meet the burden of proving the provider’s custom. Second, we know of no case considered by this court in the past, other than Critical Mass I and II, that would have been decided differently had this test been applied. Finally, in defining the limits of the National Parks confidentiality test, we do not repudiate any part of our holding in that case. A respect for stare decisis does not require the most expansive application of principles enunciated in a prior decision. To the contrary,

[i]t is a maxim not to be disregarded, that general expressions, in every opinion, are to be taken in connection with the ease in which those expressions are used. If they go beyond the case, they may be respected, but ought not to control the judgment in a subsequent suit when the very point is presented for decision. The reason of this maxim is obvious. The question actually before the Court is investigated with care, and considered in its full extent. Other principles which may serve to illustrate it, are considered in their relation to the case decided, but their possible bearing on all other cases is seldom completely investigated.

Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 399-400, 5 L.Ed. 257 (1821). See also Unit*880ed States v. Rubin, 609 F.2d 51, 69 n. 2 (2d Cir.1979) (Friendly, J., concurring).

Applying this rule to the INPO reports, we agree with the district court’s conclusion that the information they contain is commercial in nature; that the reports are provided to the NRC on a voluntary basis; and that INPO does not customarily release such information to the public. See Critical Mass, 644 F.Supp. at 346-47 and n. 5; see also Critical Mass I, 830 F.2d at 280-81. On the basis of these findings, we hold that the INPO reports are confidential within the meaning of Exemption 4 and therefore protected from disclosure.

CMEP asserts that the test we announce today may lead government agencies and industry to conspire to keep information from the public by agreeing to the voluntary submission of information that the agency has the power to compel. CMEP alleges that the NRC had in fact planned to impose greatly expanded mandatory reporting requirements on its licensees until INPO undertook to manage the voluntary reporting system they would have replaced and to provide the NRC with access to the information generated by it. See Brief for Appellant at 37-39.

We know of no provision in FOIA that obliges agencies to exercise their regulatory authority in a manner that will maximize the amount of information that will be made available to the public through that Act. Nor do we see any reason to interfere with the NRC’s exercise of its own discretion in determining how it can best secure the information it needs. So long as that information is provided voluntarily, and so long as it is of a kind that INPO customarily withholds from the public, it must be treated as confidential.

III. CONCLUSION

Having found none of the special circumstances that would justify overturning a precedent of such long standing, we reaffirm the two-prong test in National Parks. As a practical matter, however, we confine it to the category of cases to which it was first applied; namely, those in which a FOIA request is made for financial or commercial information a person was obliged to furnish the Government. We also hold that Exemption 4 protects any financial or commercial information provided to the Government on a voluntary basis if it is of a kind that the provider would not customarily release to the public.

Accordingly, we vacate our decision in Critical Mass I and affirm the district court’s grant of summary judgment in favor of the NRC. To the extent that any of our other precedents may be read as inconsistent with our decision today, they are to that degree overruled.

So ordered.