Affected non-party Meyer, Suozzi, English & Klein, P.C. ("Meyer, Suozzi"), a law firm, appeals from Judge Patterson's order releasing a redacted version of a sealed investigative report filed with the district court. We reverse in part and remand in part.

The report in question was prepared by Mary Shannon Little, a Court Officer appointed pursuant to a consent decree to investigate allegations of corruption in Local 100 of the Hotel Employees & Restaurant Employees International Union AFL-CIO. The consent decree granted the Court Officer broad powers, including authority to subpoena witnesses and to take testimony under oath and "all of the powers, privileges and immunities of a person appointed pursuant to Rule 66, Fed.R.Civ.P., and which are customary for court appointed offices [sic] performing similar assignments." The Court Officer subpoenaed hundreds of documents and interviewed hundreds of witnesses in the course of her investigation.

As part of her inquiry, the Court Officer examined the activities of various firms and individuals who provided services to Local 100. Included was Meyer, Suozzi, Local 100's legal counsel for the period 1983-1991. Her investigation of Meyer, Suozzi's representation of Local 100 necessarily involved Harold Ickes, a member of the firm principally responsible for representing Local 100 during this period. Mr. Ickes has since been appointed Deputy Chief of Staff to President Clinton.

The Court Officer deemed it advisable to submit periodic reports to the court. Usually, these reports were simply filed with the underlying litigation's docket number in the district court. On occasion, however, she delivered material that she considered confidential directly to Judge Patterson. One such report concerned the Court Officer's inquiry into Meyer, Suozzi's relationship with Local 100 ("the Report"). Presumably because of Ickes' position in the White House, appellee Newsday perceived the Report to be of public interest and intervened to seek its unsealing.

This matter is before us for the second time. In the first appeal, familiarity with which is assumed, we held that the Report was a "judicial document" and thus "presumptively" subject to public inspection. United States v. Amodeo, 44 F.3d 141, 146 (2d Cir.1995) ("Amodeo I "). We also held, however, that the district court could redact the Report, if necessary, to accommodate concerns of the Court Officer, who objected to the unsealing on the ground that it might interfere with her ongoing investigation. Id. at 147. In particular, the Court Officer was concerned that release of the Report would disclose the identity of confidential informants who had provided her with leads and would perhaps deter sources in the future. As a result of those concerns, the district court had, prior to the appeal in Amodeo I, allowed the Court Officer to redact the report. Id. at 144. We remanded with regard to these redactions to insure that they were based on the court's independent balancing of interests and were not the result of an improper delegation of judicial authority to the Court Officer. Id. at 147. We also stated that the district court could redact the report to accommodate the privacy interests of Meyer, Suozzi. Id. at 147-48.

On remand, the district court redacted the Report so as to accommodate the wishes of the Court Officer, see United States v. Amodeo, No. 92-7744 (S.D.N.Y. Feb. 23, 1995), but rejected additional redactions--virtually the entire Report--proposed by Meyer, Suozzi, see United States v. Amodeo, No. 92-7744, 1995 WL 261517 (S.D.N.Y. May 1, 1995). The firm then took this expedited appeal.

Our disposition of this matter and our reasons for it are somewhat complex. Our previous decision established a "presumption favoring access" to the Report, Amodeo I, 44 F.3d at 146, and stated that Meyer, Suozzi had the burden of overcoming that presumption. Id. at 148. We now address the standards to be used in balancing the presumption of access to the Report against Meyer, Suozzi's objections. We agree that the redactions made to the two-part Report by the district court at the suggestion of the Court Officer were appropriate. However, we believe that release of the redacted version of Part 1 of the Report either would provide little meaningful information to the public because the redactions are so extensive or might, if responded to, cause the confidential sources to be identified. It would, moreover, subject Ickes and Meyer, Suozzi to the public airing of accusations that are anonymous, unverified, and, to a degree, of doubtful veracity. We therefore reverse as to the unsealing of Part 1 of the Report. We remand for a reconsideration of the unsealing of the remaining portion, Part 2 of the Report, in light of the ensuing discussion.

Courts have given various descriptions of the weight to be given to the presumption of access, ranging from an "especially strong" presumption requiring "extraordinary circumstances to justify restrictions," United States v. Myers (In re Nat'l Broadcasting Co.), 635 F.2d 945, 952 (2d Cir.1980), to merely "one of the interests" that may bow before "good reasons" to deny the requested access. Belo Broadcasting Corp. v. Clark, 654 F.2d 423, 434 (5th Cir. Unit A 1981).

The difficulty in defining the weight to be given the presumption of access flows from the purpose underlying the presumption and the broad variety of documents deemed to be judicial. The presumption of access is based on the need for federal courts, although independent--indeed, particularly because they are independent--to have a measure of accountability and for the public to have confidence in the administration of justice. Federal courts exercise powers under Article III that impact upon virtually all citizens, but judges, once nominated and confirmed, serve for life unless impeached through a process that is politically and practically inconvenient to invoke. Although courts have a number of internal checks, such as appellate review by multi-judge tribunals, professional and public monitoring is an essential feature of democratic control. Monitoring both provides judges with critical views of their work and deters arbitrary judicial behavior. Without monitoring, moreover, the public could have no confidence in the conscientiousness, reasonableness, or honesty of judicial proceedings. Such monitoring is not possible without access to testimony and documents that are used in the performance of Article III functions.

The public's exercise of its common law access right in civil cases promotes public confidence in the judicial system.... As with other branches of government, the bright light cast upon the judicial process by public observation diminishes the possibilities for injustice, incompetence, perjury, and fraud. Furthermore, the very openness of the process should provide the public with a more complete understanding of the judicial system and a better perception of its fairness.

Leucadia, Inc. v. Applied Extrusion Technologies, Inc., 998 F.2d 157, 161 (3d Cir.1993) (quoting Republic of the Phil. v. Westinghouse Elec. Corp., 949 F.2d 653, 660 (3d Cir.1991) (citation omitted)); accord Bank of Am. Nat'l Trust & Savings Ass'n v. Hotel Rittenhouse Assocs., 800 F.2d 339, 343-44 (3d Cir.1986); see also In re Continental Ill. Sec. Litig., 732 F.2d 1302, 1308 (7th Cir.1984) ("These policies relate to the public's right to monitor the functioning of our courts, thereby insuring quality, honesty and respect for our legal system.").

Having said that, it must be recognized that an abundance of statements and documents generated in federal litigation actually have little or no bearing on the exercise of Article III judicial power. The relevance or reliability of a statement or document generally cannot be determined until heard or read by counsel, and, if necessary, by the court or other judicial officer. As a result, the temptation to leave no stone unturned in the search for evidence material to a judicial proceeding turns up a vast amount of not only irrelevant but also unreliable material.

Unlimited access to every item turned up in the course of litigation would be unthinkable. Reputations would be impaired, personal relationships ruined, and businesses destroyed on the basis of misleading or downright false information. Newsday stresses that litigants can accomplish similar ends by including such information in pleadings, which are both public and privileged. There is some truth in this assertion, particularly with regard to vexatious pro se litigants. See In re Martin-Trigona, 737 F.2d 1254, 1264-68 (2d Cir.1984) (Appendix A, containing typical complaint). However, the public is typically neither familiar with nor interested in pleadings by such litigants.

Where lawyers are involved in scurrilous allegations, professional sanctions may be available. See generally Model Code of Professional Responsibility DR 7-102(A)(1) (1980); Model Rules of Professional Conduct Rule 3.1 (1983) (professional rules upon which sanctions may be based); see also ABA Standards for Imposing Lawyer Sanctions, Standard 6-2 (1986) (outlining appropriate circumstances for disbarment, suspension, reprimand, and admonition). Monetary sanctions are also available under Federal Rule of Civil Procedure 11. An action for wrongful civil proceedings may also be available against the party making the allegations. See, e.g., Restatement (Second) of Torts Sec. 674 (1976); Conn.Gen.Stat. Sec. 52-99. Moreover, a court can strike a pleading as scurrilous under Federal Rule of Civil Procedure 12(f) whereas it is powerless to discredit every statement or document turned up in the course of litigation. A court cannot even prevent the use by the media of the somewhat misleading term "court records" in referring to such items.

We believe that the weight to be given the presumption of access must be governed by the role of the material at issue in the exercise of Article III judicial power and the resultant value of such information to those monitoring the federal courts. Generally, the information will fall somewhere on a continuum from matters that directly affect an adjudication to matters that come within a court's purview solely to insure their irrelevance.

In Joy v. North, 692 F.2d 880 (2d Cir.1982), cert. denied sub nom. Citytrust v. Joy, 460 U.S. 1051, 103 S. Ct. 1498, 75 L. Ed. 2d 930 (1983), we vacated a protective order that put under seal the report and accompanying materials of a corporate board's special litigation committee. Accord Continental Illinois, 732 F.2d at 1309. The report had been submitted to the court as the principal basis for a defendants' motion for summary judgment dismissing a shareholders' derivative action, which the court granted. We noted, "An adjudication is a formal act of government, the basis of which should, absent exceptional circumstances, be subject to public scrutiny. We simply do not understand the argument that derivative actions may be routinely dismissed on the basis of secret documents." Joy, 692 F.2d at 893. We noted that a denial of access to such materials might raise serious constitutional issues. Id. We have also consistently held that the public has an "especially strong" right of access to evidence introduced in trials. Myers, 635 at 952 (videotape evidence); United States v. Salerno (In re CBS, Inc.), 828 F.2d 958, 960-61 (2d Cir.1987) (videotaped deposition testimony). The First Circuit has similarly favored public access to documents filed with the court in furtherance of judicial approval of a consent decree where the district judge had unambiguously stated that "the documents now in question were material and important to my decision to approve the settlement." FTC v. Standard Fin. Management Corp., 830 F.2d 404, 408-09 & n. 5 (1st Cir.1987) (internal quotations marks omitted). In each of these cases, the strong weight to be accorded the public right of access to judicial documents was largely derived from the role those documents played in determining litigants' substantive rights--conduct at the heart of Article III--and from the need for public monitoring of that conduct. As one moves along the continuum, the weight of the presumption declines. One judge has pointed out, for example, that where a district court "denied the summary judgment motion, essentially postponing a final determination of substantive legal rights," the public interest in access "is not as pressing." In re Reporters Comm. for Freedom of the Press, 773 F.2d 1325, 1342 n. 3 (D.C.Cir.1985) (Wright, J., concurring in part and dissenting in part). Where statements or documents in the middle of the continuum are at issue, the weight to be accorded to the presumption of access must be determined by the exercise of judgment. That judgment can be informed in part by tradition. Where such documents are usually filed with the court and are generally available, the weight of the presumption is stronger than where filing with the court is unusual or is generally under seal.

Where testimony or documents play only a negligible role in the performance of Article III duties, the weight of the presumption is low and amounts to little more than a prediction of public access absent a countervailing reason. Documents that play no role in the performance of Article III functions, such as those passed between the parties in discovery, lie entirely beyond the presumption's reach, Standard Fin. Management, 830 F.2d at 408, and "stand[ ] on a different footing than ... a motion filed by a party seeking action by the court," Bank of America, 800 F.2d at 343, or, indeed, than any other document which is presented to the court to invoke its powers or affect its decisions.

Finally, we address the issue of whether the weight to be accorded the presumption is affected by the motive of the person seeking access. Although the presumption of access is based on the need for the public monitoring of federal courts, those who seek access to particular information may want it for entirely different reasons. However, we believe motive generally to be irrelevant to defining the weight accorded the presumption of access. It is true that journalists may seek access to judicial documents for reasons unrelated to the monitoring of Article III functions. Nevertheless, assessing the motives of journalists risks self-serving judicial decisions tipping in favor of secrecy. Where access is for the purpose of reporting news, moreover, those interested in monitoring the courts may well learn of, and use, the information whatever the motive of the reporting journalist.

Different considerations apply where personal motives, such as an individual vendetta or a quest for competitive economic advantage, are involved. However, we believe these considerations are best weighed as part of an assertion by a person or firm of a right of privacy based on an anticipated injury as a result of disclosure.

B. Countervailing Factors to be Balanced Against the Presumption of Access

Once the weight of the presumption is determined, a court must balance competing considerations against it. In the present matter, there are two countervailing factors: (i) the danger of impairing law enforcement or judicial efficiency and (ii) the privacy interests of those resisting disclosure.

1. Law Enforcement Concerns and Judicial Efficiency

Unlimited access, while perhaps aiding the professional and public monitoring of courts, might adversely affect law enforcement interests or judicial performance. Officials with law enforcement responsibilities may be heavily reliant upon the voluntary cooperation of persons who may want or need confidentiality. If that confidentiality cannot be assured, cooperation will not be forthcoming. Although courts have compulsory process available to compel the giving of evidence or the production of documents, cooperation is also often essential to judicial efficiency. If release is likely to cause persons in the particular or future cases to resist involvement where cooperation is desirable, that effect should be weighed against the presumption of access. See United States v. Valenti, 987 F.2d 708, 714 (11th Cir.) (affirming district court's refusal to unseal transcripts of in camera proceedings on the ground that it would damage continuing law enforcement investigations), cert. denied, --- U.S. ----, 114 S. Ct. 289, 126 L. Ed. 2d 238 (1993). One consideration, therefore, is whether public access to the materials at issue is likely to impair in a material way the performance of Article III functions.

The second consideration is the privacy interest of the person resisting disclosure. We have previously held that "[t]he privacy interests of innocent third parties ... should weigh heavily in a court's balancing equation." Gardner v. Newsday, Inc. (In re Newsday, Inc.), 895 F.2d 74, 79-80 (2d Cir.) (quoting United States v. Biaggi (In re New York Times Co.), 828 F.2d 110, 116 (2d Cir.1987), cert. denied, 485 U.S. 977, 108 S. Ct. 1272, 99 L. Ed. 2d 483 (1988) (citation omitted)), cert. denied, 496 U.S. 931, 110 S. Ct. 2631, 110 L. Ed. 2d 651 (1990). Such interests, while not always fitting comfortably under the rubric "privacy," are a venerable common law exception to the presumption of access. Courts have long declined to allow public access simply to cater "to a morbid craving for that which is sensational and impure." In re Caswell, 18 R.I. 835, 836, 29 A. 259, 259 (1893). As the Supreme Court noted in Nixon v. Warner Communications, Inc., 435 U.S. 589, 98 S. Ct. 1306, 55 L. Ed. 2d 570 (1978), courts have the power to insure that their records are not "used to gratify private spite or promote public scandal," and have "refused to permit their files to serve as reservoirs of libelous statements for press consumption." Id. at 598, 98 S.Ct. at 1312 (internal quotation marks and citation omitted) (collecting cases); see also Stevenson v. News Syndicate Co., 276 A.D. 614, 96 N.Y.S.2d 751, aff'd, 302 N.Y. 81, 96 N.E.2d 187 (2d Dep't 1950).1

In determining the weight to be accorded an assertion of a right of privacy, courts should first consider the degree to which the subject matter is traditionally considered private rather than public. Financial records of a wholly owned business, family affairs, illnesses, embarrassing conduct with no public ramifications, and similar matters will weigh more heavily against access than conduct affecting a substantial portion of the public. In Joy v. North, for example, the involvement of a publicly owned company and management's relationship to the public shareholders weighed in favor of unsealing a special litigation committee's report. See 692 F.2d at 894.

The nature and degree of injury must also be weighed. This will entail consideration not only of the sensitivity of the information and the subject but also of how the person seeking access intends to use the information. Commercial competitors seeking an advantage over rivals need not be indulged in the name of monitoring the courts, and personal vendettas similarly need not be aided. The court should consider the reliability of the information. Raw, unverified information should not be as readily disclosed as matters that are verified. Similarly, a court may consider whether the nature of the materials is such that there is a fair opportunity for the subject to respond to any accusations contained therein.

We turn now to the present case. We first address the weight of the presumption of access. The Report is on the periphery of the adjudicative process. The district court did not require that such reports be filed. As the district court stated, they were filed

to make sure that the court officer is doing what she was appointed to do--that is the only reason for my having to review it--and to make sure that the case doesn't go to sleep, but letting the court officer do what the court appointed her to do.

Amodeo I, 44 F.3d at 146 (internal quotation marks omitted). The Report was not filed in connection with a request for action by the district court. Instead, the Report was, as we stated in Amodeo I, "helpful," 44 F.3d at 146, because it provided information about the comprehensiveness of the Court Officer's monitoring of the consent decree. We therefore conclude that, although the Report, as a judicial document, carries a presumption of public access, see id., the presumption is weak because the Report bears only a marginal relationship to the performance of Article III functions.

We now turn to the countervailing concerns. First, with regard to law enforcement interests or judicial efficiency, the Court Officer asserted that she would not have prepared the Report had she known that it would become public. In particular, she expressed apprehensions about her access to confidential informants helpful to her monitoring of the consent decree. If such informants in the present or future cases anticipate that their cooperation will likely become a matter of public knowledge, valuable cooperation might cease. However, Judge Patterson's redactions have satisfied those concerns.

Second, with regard to Meyer, Suozzi's privacy interests, we note that the substance of a lawyer's representation of a client generally involves a variety of matters about which privacy is expected. Much of that expectation, however, is based on the existence of the attorney-client privilege and work product immunity, neither of which applies--and thus have not been asserted--in the instant matter. The attorney-client privilege belongs to the client, see United States v. Goldberger & Dubin, P.C., 935 F.2d 501, 504 (2d Cir.1991)--here the union--and the consent decree in effect waives it. Assuming that Meyer, Suozzi has a work product confidentiality interest to be asserted in its own name, see, e.g., Rhone-Poulenc Rorer Inc. v. Home Indem. Co., 32 F.3d 851, 866 (3d Cir.1994), it was waived when Ickes opened his files and gave statements regarding that product to the Court Officer. Meyer, Suozzi does not claim otherwise. Moreover, the expectation of privacy is diminished where the client is a body such as a labor union or publicly held corporation, see Joy, 692 F.2d at 894, because a substantial number of citizens rely upon the representation but cannot control it. In such circumstances, the lawyer stands in a relationship to those citizens as something of a fiduciary. Some public scrutiny of the lawyer's role may thus be expected and even called for.

The nature of some parts of the Report militate against unsealing, however. Portions of the Report are hearsay, see United States v. Charmer Indus., 711 F.2d 1164, 1170-71 (2d Cir.1983), and may contain "misinformation," see id. at 1175. "[T]here is a strong possibility that the report will contain material which is untrustworthy or simply incorrect...." United States v. Corbitt, 879 F.2d 224, 231 (7th Cir.1989).

Turning to the Report, we draw a distinction between its two parts. Part 1 involves various accusations concerning judicial and arbitral proceedings in which Ickes represented the union and includes accusations concerning Ickes. The accusations are all unsworn, and some or all may be of doubtful veracity, possibly stemming in part from apparent personality conflicts. Most important, much of Part 1 has been rendered unintelligible as a result of redactions. Indeed, the title of Part 1 is: "1. [redacted] allegations." The names of the Court Officer's sources and of Ickes' accusers have been redacted, as have the names of the firms or other unions involved. Release of Part 1 is thus more likely to mislead than to inform the public. It would circulate accusations that cannot be tested by the interested public because the sources and much of the subject matter are shrouded by the redactions. Moreover, Meyer, Suozzi and Ickes may well know, or be able to infer, the identity of some of the accusers. Unsealing Part 1, therefore, would put Meyer, Suozzi and Ickes in the unfair position of choosing between suffering the accusations in silence or revealing redacted information.

Part 2 of the Report stands on a different footing. It has only one redaction and is otherwise intelligible. It contains little unverifiable hearsay and no material that might be described as scandalous, unfounded, or speculative. Part 2 basically describes Meyer, Suozzi's role as a legal advisor in union affairs and consists essentially of Ickes' own description of that role. Part 2 also touches upon Meyer, Suozzi's representation of the Long Island Federation of Labor when Amodeo, Local 100's President, was also the Federation's President.

The expectation of privacy concerning such matters is diminished because the clients are representative institutions with legal and ethical obligations to the membership at large. The subject matter--Meyer, Suozzi's conduct in the face of criminal elements in the union leadership--thus is such that the firm might reasonably have expected some public scrutiny. Indeed, a precedent stating that such matters may not as a matter of law be disclosed would not create desirable incentives. We add that the Report absolves Meyer, Suozzi and Ickes of wrongdoing.

We reverse the district court's unsealing of Part 1 of the Report as an abuse of discretion. We recognize that the present posture of the case has caused us to add some definition to the balancing of the presumption of access against competing considerations. We therefore believe that the district court should reconsider its decision to release Part 2 in light of this opinion, because it is in the best position to weigh the factors described above.

Partial redaction appears not to be a viable remedy. Indeed, Meyer, Suozzi's suggested redactions include virtually the entire text except for portions exonerating the firm from wrongdoing. The decision to unseal Part 2 is, therefore, an all or nothing matter. The matter need not be prolonged because we also believe that either unsealing or denying public access to Part 2 is within the court's discretion.

Reversed in part and remanded in part.