The Board's prior practice had been to file a transcript of formal proceedings with the state supreme court in some cases in which the charges were dismissed, as well as in all those in which discipline had been proposed. After receiving an opinion from counsel, however, the Board determined that under the state constitution the record was to be sent to the court, and thus made public, only in those cases where discipline was recommended. Consequently, the Larsen record remained sealed.

The First Amendment Coalition then filed a petition for mandamus with the state supreme court asking that it compel the Board to file the record with the court. The petition was denied, the court stating that it was prohibited from granting the request because the Board had not suggested suspension, removal, discipline or retirement. In the absence of a recommendation by the Board there was "no constitutional authority for [the] court to review the record and act. The matter is constitutionally closed." First Amendment Coalition v. Judicial Inquiry and Review Bd., 501 Pa. 129, 133, 460 A.2d 722, 724 (1983).

Following these developments, the district court received evidence on the history and practices of the Board. No material issues of fact were in dispute, and the plaintiffs' motion for summary judgment was granted. Preliminarily, the court rejected the Board's contention that the case was mooted by the Inquirer's publication of the Larsen transcripts. Observing that the Inquirer had "not shared its riches" with fellow members of the Coalition, the court found that the matter still presented a live dispute as to them. Moreover, the claims fell into the category of those capable of repetition yet evading review.

After surveying decisional law on the First Amendment and a right of access, the district court concluded that "a restriction on public and press access can be sustained, but only to the extent that it demonstrably advances significant governmental interests." First Amendment Coalition v. Judicial Inquiry and Review Bd., 579 F. Supp. 192, 211 (E.D.Pa.1984). Noting that only a fraction of the Board's investigations result in formal charges,1 the court found a substantial state interest "in protecting accused judges, and the judiciary itself, from the public hearing of charges, most of which will evaporate." Id. at 214. Consequently, the Coalition's "insistence on access to all charges other than those which are 'obviously unfounded or frivolous' is not persuasive." Id.

However, in instances where the Board has preferred formal charges, the court concluded that denial of access impairs the public's opportunity to appraise the work of the Board, the standards of judicial conduct it applies, and the consistency of enforcement.

Acknowledging the "trauma of public accusation," one which is "greater for an official who, due to the special constraints of the bench, is largely disabled from seeking public support," the court found a "tension between the identified public interest and the identified cost." Id. at 215. "The way of maximizing these twin interests is to permit access to all cases in which the Board prefers formal charges--but to defer the time of access until the Board's filing with the Supreme Court of a transcript which fully records the Board's proceedings." Id. As a result, the state constitutional requirement was modified by the district court's directive that the Board make public, on disposition, the record of all proceedings in which it had filed formal charges.

In discussing the contentions of the individual plaintiffs, the district court recognized a valid state interest in insisting on witness secrecy. Id. at 217. Accordingly, the court declared that the Board may "impose confidentiality upon any witness who appears and testifies ... concerning the fact of the witness' appearance and the substance of any testimony until such time as the record of the Board's proceedings are made available to the public."

All parties have appealed. Plaintiffs contend that the court erred in allowing access only to a transcript at the completion of formal proceedings and that the restrictions on nonparty witnesses violate the constitutional guarantee of free speech. The Board argues that the confidentiality requirement is appropriate to the Board's role, has only slight impact on news-gathering, and is consistent with federal constitutional standards.

In response to the need for modernization of its constitution, Pennsylvania called a convention in 1968. Although a number of changes in the organization of state government were proposed, the principal item presented to the delegates was the preparation of a new judiciary article. The convention ultimately submitted a proposal, designated Article V of the state constitution, governing the selection, retention, and tenure of judicial officers.

One of the convention's most valuable contributions to that Article was the establishment of a Judicial Inquiry and Review Board, a constitutionally independent body to oversee the conduct of the state's judiciary. The essential elements of that proposal had been recommended to the Convention's Preparatory Committee in 1967 by the Pennsylvania Bar Association. That presentation expressed dissatisfaction with the cumbersome method of impeachment as the sole procedure for grappling with the problems of the aged, infirm, irascible, or, in rare instances, corrupt judge. The Bar advocated a new method of solving these problems, including measures short of removal from office.

Favorable comments were received on the operation of the California Commission on Judicial Qualifications, which had been established some years earlier. Speaking on behalf of the state bar association, Bernard G. Segal, Esquire, devoted special attention to the fact that a number of California judges had voluntarily resigned or retired while under investigation by the Commission. He also remarked that under California practice when a complaint is filed with the Commission, its "investigations, deliberations, and conclusions on the case are completely secret, except, of course, if the Commission's decision is appealed to the Supreme Court." Statement of Bernard G. Segal on the Proposed Judiciary Article before the Preparatory Committee for the Pennsylvania Constitutional Convention.

Dean Laub of the Dickinson School of Law, a former judge, wrote an article listing the arguments for and against the features of the various plans proposed. He referred to the California plan noting, under that state's procedure "confidentiality is maintained until the matter is referred to the Supreme Court for decision." As a possible drawback to the California system, he listed the potential for abuse in the investigative power conferred on an independent agency. He also discussed criticism of the plan followed in New York because its proceedings had no assurance of confidentiality. Laub, Issues Before the Judiciary Committee of the Pennsylvania Constitutional Convention, 39 Pa.B.A.Q. 390 (1968).

Materials submitted to the delegates discussed in even greater detail the pros and cons of confidentiality in Board proceedings. Removal, Suspension, and Discipline of Judges, reprinted in, The Pennsylvania Constitutional Convention, Reference Manual No. 5: The Judiciary. See also Woodside, Pennsylvania Constitutional Law, 433-38 (1985) (The author is a former Pennsylvania appellate judge and was a delegate to the convention).

In presenting the plan for a Judicial Inquiry and Review Board to the convention for a vote, William W. Scranton, Chairman of the Convention's Judiciary Committee and a former governor of Pennsylvania, urged adoption. He stated that if the proposal were approved it would be "a tremendous step forward for the people of Pennsylvania as well as for the judges of same. It protects the judges and at the same time makes changes where changes are necessary in the finest kind of way." Journal of the Constitutional Convention, February 29, 1968, page 1374. Review of the convention materials thus demonstrates that the question of confidentiality was actively considered in the preparation of the constitutional amendment.

As finally adopted by affirmative vote of the people of the state, the constitution provides for a Board composed of five judges, two lawyers, and two laymen. It receives complaints or reports and makes preliminary investigations. After further examination, the Board may order a hearing and direct the attendance and testimony of witnesses. "If after hearing, the Board finds good cause therefor, it shall recommend to the Supreme Court the suspension, removal, discipline, or compulsory retirement of the justice or judge." Pa. Const. Art. V, Sec. 18(g).

The state supreme court reviews the Board's record and may receive additional evidence. Following that, the court may order discipline as recommended by the Board, may impose a different measure of discipline, or may exonerate the accused judge. "All papers filed with and proceedings before the Board shall be confidential but upon being filed by the Board with the Supreme Court, the record shall lose its confidential character. The filing of papers with and the giving of testimony before the Board shall be privileged." Pa. Const. Art. V, Sec. 18(h).

Implementing legislation consistent with the constitutional language was enacted and codified at 42 Pa.Cons.Stat.Ann. Sec. 2101 et seq. (Purdon 1981). In accordance with a constitutional directive, the state supreme court drafted rules of procedure for the Board.

Certain features of this case discussed by the parties in their briefs play no part in today's decision. Initially, we note that although the Board's earlier practice was to file its record with the state court in some cases where the charges were dismissed, it has now been authoritatively established that the state constitution permits filing only where discipline has been recommended. The Coalition does not now contend that the Board's prior practice was consistent with the state constitution. The district court recognized the state's construction of its constitutional provision as binding on the federal courts, and we agree with that determination.

Additionally, it should be apparent that since there has been publication of substantial portions of the Larsen hearings, the merits of that proceeding are not particularly pertinent here. The case before us is not legally moot, but realistically what is at stake is the Board's procedure in future cases.

The obvious must also be stated. The Coalition's claims are based on an alleged right of access, not a right of publication. Although both have their roots in the First Amendment, these principles are doctrinally discrete, and precedents in one area may not be indiscriminately applied to the other. In general, the right of publication is the broader of the two, and in most instances, publication may not be constitutionally prohibited even though access to the particular information may properly be denied. New York Times Co. v. United States, 403 U.S. 713, 91 S. Ct. 2140, 29 L. Ed. 2d 822 (1971) (per curiam).

The issue before us is not whether the First Amendment prohibits the state from barring public observation of judicial disciplinary proceedings at all stages. Pennsylvania has provided for disclosure but has limited it to the situation in which discipline has been recommended and the record has been filed with the court. Consequently, we assume, but do not decide, that there is a constitutional right of access to disciplinary proceedings at some stage.

In argument before the court, counsel for the Coalition conceded that "the most difficult issue is the issue of where in the process the right of access attaches and we have struggled with this issue." Indicative of that difficulty is the Coalition's retreat from its original position that access was required to "all proceedings on charges which the Board has determined not to be frivolous or obviously unfounded." The plaintiffs' brief at 6. That would have permitted disclosure of informal action by the Board such as private reprimands and requests for resignations in lieu of formal hearings. The Coalition's present contention is that contemporaneous access should commence at the point where the Board issues formal charges against a judge.

Although we assume a right of access, it does not attach at the same time it might in certain other contexts and at the point the Coalition urges on us here. By analogy to the cases establishing a First Amendment right of access to criminal trials, see, e.g., Richmond Newspapers Inc. v. Virginia, 448 U.S. 555, 100 S. Ct. 2814, 65 L. Ed. 2d 973 (1980), and this court's decision to extend the rationale to civil trials, Publicker Industries, Inc. v. Cohen, 733 F.2d 1059 (3d Cir.1984), the Coalition maintains that where a constitutional right of access is found to exist, a "presumption of openness" is created. This presumption, the Coalition contends, places the burden on the state to justify restrictions on access by showing a "compelling governmental interest [which the restriction] is narrowly tailored to serve." Globe Newspapers Co. v. Superior Court, 457 U.S. 596, 607, 102 S. Ct. 2613, 2620, 73 L. Ed. 2d 248 (1982).

All rights of access are not co-extensive, however, and some may be granted at different stages than others. In assuming a right of access, we need not postulate a span as extensive as that in civil and criminal trials as such, but rather may be guided by the unique history and function of the Judicial Review Board.

Richmond Newspapers and the cases decided in its wake stressed the tradition of open trials in England and then later in colonial America. Since the Bill of Rights had been adopted "against the backdrop of the long history of trials being presumptively open," 448 U.S. at 575, 100 S.Ct. at 2826, the Court concluded that the First Amendment prohibits the "government from summarily closing courtroom doors which had long been open to the public at the time that Amendment was adopted." Id. at 576, 100 S.Ct. at 2827.

But the cases defining a right of access to trials are, at best, of limited usefulness in the context of the fundamentally different procedures of judicial disciplinary boards. These administrative proceedings, unlike conventional criminal and civil trials, do not have a long history of openness. Recognizing this fact, the Coalition points to time honored judicial removal in open impeachment hearings.

Had the state constitutional convention acted to replace traditional impeachment with a substitute vehicle like the Judicial Inquiry and Review Board, a closer question of public access to the successor proceedings would be presented. But it is clear that the Board's functions are intended to supplement rather than replace the historical methods of judicial discipline: impeachment and removal for conviction of a crime. Pa. Const. Art. V, Sec. 18(h), 18(1); see also Judicial Discipline, 84 Dick.L.Rev. 447, 449-52 (1980). It was largely the recognition that these traditional methods are cumbersome and ineffective, partly because of their openness, that spurred the constitutional convention to conceive the new judicial disciplinary procedure.

Against this background, the "presumption of openness" gleaned from the history of criminal trials surveyed in Richmond Newspapers lacks force. Rather, in judicial disciplinary proceedings, what tradition there is, favors public access only at a later stage in the process. A temporally based right is no stranger to the law. For example, tradition supports the secrecy of the grand jury, the entity in the criminal justice system to which the Board is most akin. Similarly, sidebar conferences between lawyers and judges at trial are contemporaneously confidential although they may later appear as part of the transcript. See United States v. Gurney, 558 F.2d 1202, 1210 (5th Cir.1977).

In choosing the point at which formal charges are filed as the stage when proceedings should become public, the Coalition uses an analogy with traditional criminal procedures beginning with the indictment. It would treat formal proceedings before the Board as the equivalent of the criminal pre-trial and trial proceedings. This analogy is faulty because the Board cannot impose, but only recommend, punishment, and in that sense its functions are similar to those of the grand jury. Only the state supreme court has the power to discipline just as in the criminal field only a court has the power to sentence.

It may be said that the Board's recommendation has the effect of an indictment, not a conviction. The traditional notion of protection for a non-indicted target applies equally well in the disciplinary setting. That the Board's rules grant the accused more extensive procedural rights than are allotted to the subject of a grand jury investigation does not undermine the analogy.

Forcing judicial review proceedings into an older criminal procedural mold would have a stifling effect on a state's ability to use creative methods in solving its problems. It is quite uncertain whether the state would have chosen a judicial disciplinary program or have been able to implement one in the absence of the confidentiality provision. The Coalition has failed to show that the right of access it urges is so compelling as to justify the restriction on the state's freedom of choice.

The Coalition goes beyond tradition when it argues that the "structural values" of the First Amendment are served by subjecting the proceedings of the Board to greater public access. In his concurring opinion in Richmond Newspapers, Justice Brennan wrote that the First Amendment embodied more than a commitment to free expression for its own sake but included a "structural" role in government. Falling within this concept is the notion that the public must be properly informed in order for a democracy to survive.

The Coalition contends that the Board, in carrying out its assignments, is performing a governmental function. The public has an interest in information about the conduct of its judiciary and consequently is entitled to assurance that the Board is properly discharging its duties. As we recently noted in United States v. Smith (Appeal of the Patriot News Co.), 776 F.2d 1104 (3d Cir.1985), structural values have been a consideration in the decisions granting a right of access to trial, and even some pre-trial proceedings. See United States v. Criden, 675 F.2d 550 (3d Cir.1982).

The "structural" argument alone cannot carry the day, however, because it has not developed independent of, and unrelated to, historical antecedents. The authority relied on by plaintiffs must be viewed in the context in which it was decided--the trial process which has a long common law tradition of openness. For a court to accept the structural considerations of the First Amendment without heed to the circumstances in which they are invoked would lead to an unjustifiably expansive interpretation. See In re Reporters Committee for Freedom of the Press, 773 F.2d 1325, 1332 (D.C.Cir.1985). Justice Brennan, an advocate of the structural theory of the First Amendment, has cautioned that "the stretch of this protection is theoretically endless." Richmond Newspapers, 448 U.S. at 588, 100 S.Ct. at 2833 (concurring opinion). Consequently, the right to know "must be invoked with discrimination and temperance." Id.

The structural arguments, of course, are legitimate constitutional underpinnings for the right of publication, but they do not mandate access. Chief Justice Warren must have envisioned such contentions when he wrote:

"There are few restrictions on action which could not be clothed by ingenious argument in the garb of decreased data flow. For example, the prohibition of unauthorized entry into the White House diminishes the citizen's opportunities to gather information he might find relevant to his opinion of the way the country is being run, but that does not make entry into the White House a First Amendment right. The right to speak and publish does not carry with it the unrestrained right to gather information."

Zemel v. Rusk, 381 U.S. 1, 16-17, 85 S. Ct. 1271, 1280-81, 14 L. Ed. 2d 179 (1965). The Supreme Court has applied this limitation even in cases where the public does have an interest in gaining information--for example, attempts to establish a special right of access to prisons have been rebuffed. See, e.g., Houchins v. KQED, Inc., 438 U.S. 1, 98 S. Ct. 2588, 57 L. Ed. 2d 553 (1978); Pell v. Procunier, 417 U.S. 817, 94 S. Ct. 2800, 41 L. Ed. 2d 495 (1974).2

Even in the litigation setting, the right of access is subject to limitations. In Seattle Times Co. v. Rhinehart, 467 U.S. 20, 104 S. Ct. 2199, 81 L. Ed. 2d 17 (1984), the Supreme Court stated that "pretrial depositions and interrogatories are not public components of a civil trial," and accordingly, "restraints placed on discovered, but not yet admitted information are not a restriction on a traditionally public source of information." 104 S. Ct. at 2207-08. It bears noting that the Seattle Times case upheld a trial court ban on dissemination of information, a form of prior restraint which is not often permitted.

The Coalition does not now contend that all of the Board's activity should be open to the public. If the access question is envisioned as a line stretching from the extreme of completely open proceedings to completely closed ones, we find that each party has taken a position somewhere in the center. The battle here is not at the extremes, but in the middle ground.

The parties have staked out their positions in that area, the Board arguing that access is proper only as mandated by the state constitution, and the Coalition advocating the point at which it believes that the First Amendment requires access. The Board's demarcation point is specific and fixed, whereas the Coalition's position is somewhat selective and must rely on rather vaguely defined concepts in decisional law.

As noted earlier, the delegates to the state constitutional convention confronted the question of when access should begin. Their resolution, which is presently under attack, was arrived at with full knowledge of the competing concerns.

The plaintiffs' contentions appear to be based primarily on the desirability of granting earlier access, an argument that influenced some nineteen states. But Pennsylvania and twenty-one other states have deliberately chosen to fix the point of access at a later stage.3

In the circumstances of this case, we find that the presumption of validity attaching to state legislative and constitutional provisions weighs heavy. See Clements v. Fashing, 457 U.S. 957, 963, 102 S. Ct. 2836, 2843, 73 L. Ed. 2d 508 (1982). This presumption does not relieve the courts of their obligation to make an independent inquiry when First Amendment rights are at stake, Landmark Communications, Inc. v. Virginia, 435 U.S. 829, 843, 98 S. Ct. 1535, 1543, 56 L. Ed. 2d 1 (1978), but it does require that the state's determination be upheld unless it is found to transgress a clear constitutional prohibition. In addressing a claim of access, the Chief Justice admonished the courts not to "confuse what is 'good,' 'desirable,' or 'expedient' with what is constitutionally commanded by the First Amendment. To do so is to trivialize constitutional adjudication." Houchins v. KQED, Inc., 438 U.S. 1, 13, 98 S. Ct. 2588, 2596, 57 L. Ed. 2d 553 (1978) (Opinion of Burger, C.J.).

The Coalition's burden is particularly heavy here because the concerns of access have been accommodated up to the point that the state has determined them to be outweighed by more compelling interests. In his oft-cited lecture, "Or of the Press," 26 Hast.L.J. 631, 636 (1975), Justice Stewart said,

"There is no constitutional right to have access to particular government information, or to require openness from the bureaucracy [citing Pell v. Procunier ]. The public's interest in knowing about its government is protected by the guarantee of a Free Press, but the protection is indirect. The Constitution itself is neither a Freedom of Information Act nor an Official Secrets Act.

"The Constitution, in other words, establishes the contest, not its resolution. Congress may provide a resolution, at least in some instances, through carefully drawn legislation. For the rest, we must rely, as so often in our system we must, on the tug and pull of the political forces in American Society."

In this case, we have a state constitutional provision, precisely addressing the point, arrived at as a result of the tug and pull of political forces. Federal courts should not overturn a state's evaluation of structural concerns in the absence of egregious circumstances. Here we are not presented with the fiat of a single official acting in a discretionary fashion, but with a constitutional provision enacted by a state in conformity with Article IV, Sec. 4 of the federal constitution guaranteeing each state a republican form of government. See Bauers v. Heisel, 361 F.2d 581, 588-89 (3d Cir.1966).

The notion that the effectiveness of judicial disciplinary boards depends to a large extent on confidentiality is not unique to Pennsylvania; the idea has been almost universally accepted. In Landmark Communications, 435 U.S. 829, 98 S. Ct. 1535, 56 L. Ed. 2d 1 (1978), the Supreme Court listed four advantages of confidentiality:

1. Encouraging the filing of complaints;

2. Protecting judges from unwarranted complaints;

3. Maintaining confidence in the judiciary by avoiding premature announcement of groundless complaints; and

4. Facilitating the work of a commission by giving it flexibility to accomplish its mission through voluntary retirement or resignation of offending judges.4

See also Mosk v. Superior Ct., 25 Cal. 3d 474, 159 Cal. Rptr. 494, 601 P.2d 1030 (1979).

The possibility that judges would be harassed and the judicial system disrupted in the event of open hearings before a judicial inquiry board has not been an illusory concern. California, the first to adopt a commission-form of judicial discipline, included a constitutional provision assuring confidentiality of proceedings until filing with the state supreme court. The California Commission provided the model for the Pennsylvania Board, including its confidentiality requirements. See Frankel, Removal of Judges: California Tackles an Old Problem, 49 A.B.A.J. 166 (1963). The intuitive wisdom of that concept was confirmed by the near disaster experienced by the California Supreme Court when confidentiality concerns were brushed aside and charges against its Chief Justice were heard in public. See Cameron, The California Supreme Court Hearings--a Tragedy That Should and Could Have Been Avoided, 8 Hast. Const.L.Q. 11 (1980); Mosk, Chilling Judicial Independence--The California Experience, 3 West.N.Eng.L.R. 1 (1980); Tribe, Trying California's Judges on Television: Open Government or Judicial Intimidation?, 65 A.B.A.J. 1175, 1178 (1979) ("Gallup Poll Justice at its Worst").

In his article, Professor Tribe describes confidentiality in disciplinary proceedings as "[p]rotection vitally needed to encourage collegiality, candor, and courage--both political and intellectual--protection needed not only for the benefit of judges but for the benefit of society as a whole." 65 ABA J. 1179. The effective operation of the judiciary is a matter of serious concern, one which may suffer from too much openness.

It is noteworthy too that the disciplinary procedures for the federal judiciary enacted by Congress provide for confidentiality. See 28 U.S.C. Sec. 372(c)(14). The Senate Committee explained its belief "that the establishment of a confidentiality provision will avoid possible premature injury to the reputation of a judge," and that specified measures ought to be taken in "protecting the judge from malicious publicity." See S.Rep. No. 362, 96th Cong.2d Sess. 16 reprinted in 1980 U.S.Code Cong. & Ad.News 4315, 4330. Approval of the rationale employed by the Coalition here would inescapably lead to a holding that the confidentiality provision of the federal statute is also unconstitutional. See also Rule 10J, Judicial Council of the Third Circuit. Because the federal statute is not at issue here, however, we need not pursue discussion of that ramification.

The district court observed that there is no special federal constitutional protection from public criticism given to judges. That is undoubtedly true in the publication context, but nothing in the state constitutional provision bars dissemination of information which the press has received. Indeed, the publication of the Larsen transcript in this case illustrates that point. See also Landmark.

Nevertheless, the state has demonstrated a substantial interest in preserving limited confidentiality. That interest rests not only on the reputation of the judiciary as an institution and judges who have been accused but not proved culpable, but also on the need for flexibility so that the Board may efficiently accomplish its purpose.

In practice, it has been demonstrated that one of the most effective methods of meeting the problem of the unfit judge is to remove him from the bench by voluntary retirement or resignation. Experience has shown that some judges would prefer to resign rather than undergo complete formal hearings. Pennsylvania has concluded that if the confidentiality provisions were not in effect, the accused judge might feel compelled to seek vindication by requiring a hearing.5 In the state's view the public would lose more than it would gain in this eventuality.

Nor is it a valid objection to say that the confidentiality provision gives more favored treatment to judges than other governmental officials. The reality is that only judges are subjected to such disciplinary procedures; rather than being favored they have been singled out for more rigid oversight than their counterparts in the legislative and executive branches.6 See generally Markey, The Delicate Dichotomies of Judicial Ethics, 101 F.R.D. 373.

In summary, even assuming a right of access, an issue we specifically do not decide, determination of the point at which it becomes overriding requires a balancing of the state's concerns against those of public disclosure. The state's interests are weighty, its resolution of a serious problem is not unreasonable, and the Coalition's claim, although presenting arguably desirable alternatives, is not supported by historical antecedents. In these circumstances, the Coalition has failed to demonstrate the unconstitutionality of the state procedure, and the district court's finding of a First Amendment violation is not warranted.

The state constitution and the enabling statute both provide that "papers filed with and proceedings before the Board shall be confidential." The Board's rules of procedure state that it is regarded as contempt of court "for a person subpoenaed to in any way breach the confidentiality of the investigation. All participants shall conduct themselves so as to maintain the confidentiality of the proceeding." The subpoena served on the individual plaintiffs contained the following statement: "By direction of the Constitution of Pennsylvania these proceedings are confidential and any disclosure outside the proceedings shall constitute contempt and be actionable."

When the reporters appeared before the Board they moved to quash the subpoenas but this request was denied. A motion for an expedited appeal to the state supreme court also failed. When plaintiff Biddle appeared before the Board, his counsel remarked

"MR. KLEIN: And it is the Board's position, I take it, that Mr. Biddle is precluded from disclosing in any way his own testimony or appearance before this Board.

[Board Member]

[Board Member]

MR. RACKOFF: Absolutely."

The district court acknowledged that the Board's mandate was a form of prior restraint but found the order "justified to the extent that the confidentiality rule is justified." 579 F. Supp. at 217. The court determined that there was a valid state interest in insisting on witness secrecy until the time the Board's record was filed with the state supreme court.

The claim of the individual plaintiffs in this phase of the case is based on the broader right of free speech, not simply access. Consequently, we must be less deferential to state interests. Any prior restraint on expression comes to the court with a presumption of unconstitutionality.7 Organization for a Better Austin v. Keefe, 402 U.S. 415, 418, 91 S. Ct. 1575, 1577, 29 L. Ed. 2d 1 (1971). Therefore, the district court's equation of the restrictions placed on witnesses with the limitations on access must not be read as applying the same standard in both instances. Analysis of the factors that distinguish the two situations must be observed.

In Landmark, the Supreme Court examined the grounds for confidentiality in the Virginia judicial disciplinary procedures. Assuming their validity, the court nevertheless held the regulation could not affect a newspapers' right to publish information it had obtained from an independent source. The Court carefully noted that the newspaper was a stranger to the Commission proceedings. No reporter, employee, or representative of the paper had been subpoenaed by, or appeared before, the Commission. 435 U.S. at 837 n. 10, 98 S. Ct. at 1541 n. 10. The court made clear, therefore, that it was confronted with a restraint on publication.

In Seattle Times, on the other hand, a ban was imposed on the defendant newspaper that sought to disseminate information it had secured through the court's own discovery process. The trial court issued a protective order enjoining publication of specified data, allowing its use only for purposes of trial. By its terms, the order did not apply to information gained by the defendant through means other than discovery.

The Supreme Court affirmed the trial court's order noting that the First Amendment does not prohibit all restraints on expression, "freedom of speech does not comprehend the right to speak on any subject at any time." 104 S. Ct. at 2207. A litigant has no First Amendment right to publish information made available only for purposes of trying his suit, and an important consideration was that the prohibition extended only to information gained through that means. Finally, the Court concluded that privacy considerations and proper functioning of a court's discovery proceedings to prevent abuse justified the restrictions.

In Rodgers v. United States Steel Corp., 536 F.2d 1001, 1006 (3d Cir.1976), we expressed a similar view in dictum, but held that a protective order prohibiting counsel from discussing material in an exhibit attached to a deposition was not permissible. Because the protective order applied to information that counsel had obtained from means other than through the court's process, we decided, to that extent, the restraint was impermissible. To the same effect, see State of New York v. United States Metals Refining Co., 771 F.2d 796 (3d Cir.1985).

The restriction imposed by the Board here differs from those considered in the preceding cases. Here, unlike Landmark, plaintiffs were participants in the Board's proceedings. The situation is also unlike Seattle Times in that plaintiffs did not seek to avail themselves of the Board's processes. The reporters were compelled to appear and testify on a matter that did not affect their personal interests. As counsel agree, the individual plaintiff's posture is much like that of a grand jury witness.

The scope of the Board's rule as well as that of the district court's order which refers to the "substance of any testimony" are not entirely clear. But it is important to recognize that the order was entered in conjunction with the directive requiring the Board to file its record with the state court in every instance in which formal proceedings were conducted. Hence, the district court's restraint on a witness would last only until that filing.

Under state law the Board will not file the record until and unless it recommends disciplinary action. Therefore, when the Board dismisses charges, its record will never be filed, and the restraint on the witness will be perpetual. The district court did not contemplate a restriction of such a duration.

The curb on disclosing the witness's testimony applies to information obtained from sources outside as well as inside the Board. In short, a person having any knowledge about the conduct of a judge, favorable or otherwise, might be forever barred from speaking, writing, or publishing it if he testified about that information before the Board. We find no state interest strong enough to justify such a sweeping measure.

Pennsylvania does not impose any such prohibition on witnesses who appear before a grand jury. They are permitted to disclose their own testimony although the traditional veil of secrecy applies to other participants in the process. 42 Pa.Cons.Stat.Ann. Sec. 4549 (1980); In Re November, 1975 Special Investigating Grand Jury, 299 Pa.Super. 539, 445 A.2d 1260 (1982). This procedure is consistent with the federal practice, see Fed.R.Crim.P. 6(e)(2), and its application in judicial disciplinary proceedings would not substantially impair the Board's function.

To the extent that the confidentiality requirement is intended to prevent improper defamatory publicity, other protection is available. The state constitution provides that testimony before the Board is privileged.8 A witness, therefore, may invoke that privilege as a defense to an action for defamation. If a witness chooses to speak outside the confines of a Board hearing, however, it would seem questionable whether the privilege would apply to that statement. See Hutchinson v. Proxmire, 443 U.S. 111, 99 S. Ct. 2675, 61 L. Ed. 2d 411 (1979). For that reason, witnesses may find it in their own best interests not to divulge their testimony.

We conclude that to the extent the Board's regulation and the district court's order prevent witnesses from disclosing their own testimony, those directives run afoul of the First Amendment as impermissibly broad prior restraints. They are not totally invalid however. The confidentiality requirement is reasonable and may be enforced insofar as it would prevent a person, whether a Board member, employee, or counsel, from disclosing proceedings taking place before the Board. The same limitation applies to witnesses with the exception of their own testimony. The state interest in this respect, as in the grand jury setting, is sufficiently strong to support such a ban.

It follows that although witnesses may, if they choose, disclose their own testimony, they may not reveal that of another witness whom they may hear testify. Nor are they free to disclose the comments of Board members or staff that are overheard during their appearance. Therefore, the Board's regulation must be read to permit witnesses, at their discretion, to disclose the substance of their testimony before the Board.

Accordingly, the order of the district court will be vacated, and the case will be remanded for entry of a decree in conformity with the views expressed in this opinion.

BECKER, Circuit Judge, concurring in part.

I join in the result reached by the majority and in Part III of its opinion. I disagree, however, that there is any need to assume a right of access as the majority does, and then to determine whether the right is overridden in this case. Rather, I believe that because there is no tradition or history of access to the documents under consideration, there is simply no right of access to them, and that the court should so hold.

For a party to establish a constitutional right of access, it must show two things: (a) a tradition of openness, and (b) important functional values that would be served by access. Press-Enterprise Co. v. Superior Court, 464 U.S. 501, 104 S. Ct. 819, 78 L. Ed. 2d 629 (1984); Globe Newspaper Co. v. Superior Court, 457 U.S. 596, 102 S. Ct. 2613, 73 L. Ed. 2d 248 (1982). The relevant historical inquiry here is whether there is a tradition of openness with respect to the Judicial Inquiry and Review Board's (the Board's) records when it votes after a formal hearing to dismiss the charges against the judge under investigation.

Examination of the historical record demonstrates that there is no tradition of openness with respect to such records.1 Of the fifty-three cases decided by the Board on the merits up to June 1, 1983 the Board dismissed the chargegnin twelve of them.2 One of the cases is the one before us, whose history is by now well-known. In six of the remaining eleven cases, the Board did not file its records in the Supreme Court. The records therefore did not become public. In the five remaining cases, the Board did file its records, but in three of them the Board had apparently initially decided not to make the record public and changed its mind only because one of the three district justices under joint investigation was killed trying to prevent a hold-up around the time of the investigation. The Board filed the record in the Supreme Court to clear that justice's name and memory. Thus, in only two of the twelve relevant cases did the Board decide to make public its records absent extraordinary conditions.

It is clear to me that this "tradition" is not sufficient to sustain the burden imposed by the Supreme Court. Any other conclusion would make a nullity of the tradition of openness requirement. Judge Adams obviously disagrees, stating that "the history of the Board up to the time of this case, though not a long one, demonstrates a tradition of openness in its proceedings...." Concurrence and dissent at 484. In reaching this conclusion, however, Judge Adams discounts the six cases in which the Board did not recommend sanctions and the records were not disclosed. Although, as noted above, these six cases constitute one-half the relevant history, Judge Adams refers to them as mere "simply exceptions" to an "overall pattern" of openness. Id. at 482. Aside from giving new meaning to the old adage that "the exception proves the rule," this position has little to recommend it. The pattern it purports to identify was honored more often in the breach than otherwise--it was no pattern at all.3 The six cases therefore cannot be trivialized as exceptions to a pattern that never existed, but must be given their due in an impartial examination of the whole relevant history. Such examination results in a conclusion opposite that of Judge Adams: there was no pattern or tradition of openness here--far from it.

Judge Adams also believes that the fact that impeachment proceedings have historically been open to the public is relevant to whether there is a tradition of openness to Board documents and records. Concurrence and dissent at 485. I disagree. As the majority explains, the history of impeachment proceedings would be relevant only if the Board had been intended to be, or had in fact become, a substitute for impeachment of judges. Majority Opinion at 473. However, the record is clear that Board was intended to supplement the impeachment process, not to supplant it, and that it has not become a de facto substitute for impeachment. The number of impeachments has not declined as a result of the availability of the Board.4 Cf. United States v. Smith (Appeal of the Patriot News Co.), 776 F.2d 1104, 1111 (3d Cir.1985) (holding that there is a right of access to bills of particulars, despite their brief history, because indictment have grown "skeletal," and bills of particulars play the role once played by more expansive indictments). Thus, the history of impeachment cannot shape our analysis of the appropriate practices of the Board.

Although the parties have discussed at some length the practices of other states' judicial review boards, and the majority and Judge Adams also refer to them, such practices cannot provide the historical evidence necessary to support the tradition of openness of the Pennsylvania Board. The very nature of the historical inquiry implies that states have some flexibility in deciding which of their institutions may be open and which closed to the public. States can shape their own histories and may therefore differ about whether their citizens have rights of access to otherwise identical institutions. Such differences are elemental to our system of federalism. Invoking the histories of other states' institutions abridges the states' flexibility; in this case, it denies Pennsylvania the opportunity to decide for itself whether its Board should be an open or a closed institution.

Our first obligation is to look at history for a tradition of openness. Having done so, I believe that there is no such tradition, and that our analysis should end there. I therefore do not reach the several issues pertaining to burdens of proof and appropriate standards of review in the majority opinion and disputed by Judge Adams.

ADAMS, Circuit Judge concurring in part and dissenting in part.

I join the majority's conclusion in Part III that the reporters may not be barred from disclosing their testimony before the Board. On the issue of public access, however, I respectfully dissent, because I differ with the majority's legal analysis and its result.

At the core of our differences on the access issue is a dispute over history. The majority believes that there is a "tradition of closed proceedings in the judicial disciplinary process ..." in Pennsylvania. This conclusion, however, is belied by the fact that throughout the state's history the central instrument of judicial discipline has been impeachment, from its inception a public proceeding. Nor, when the Judicial Inquiry and Review Board (the Board) was created in 1969 to strengthen the disciplinary process, did it deviate from this tradition of openness. Significantly, the Board granted the public access to its formal hearings in the vast majority of cases by filing a transcript with the Supreme Court, and did not adopt its current position on confidentiality until the present proceeding was underway.

The correct legal analysis here flows in large measure from the historical record. Under the standards set out by the Supreme Court in Globe Newspaper Co. v. Superior Court, 457 U.S. 596, 102 S. Ct. 2613, 73 L. Ed. 2d 248 (1982), and Press-Enterprise v. Superior Court, 464 U.S. 501, 104 S. Ct. 819, 78 L. Ed. 2d 629 (1984) a tradition of openness, coupled with the powerful functional values served by public access, gives rise to a presumption of access to Board proceedings. Pursuant to that authority, the presumption of openness may be overcome only by a showing that closure "is necessitated by a compelling governmental interest, and is narrowly tailored to serve that interest." Globe, 457 U.S. at 607, 102 S.Ct. at 2620. The Board's conduct here, it is clear, cannot meet this standard.

Building on its conclusion that there has been a tradition of secrecy, the majority determines that the Coalition, in seeking access, must justify its infringement of the state's choice of confidentiality even after a determination that a conflict is non-frivolous and requires a full adversary hearing; it proceeds to hold that the Coalition has not done so in this case. Because the majority's approach reflects a misreading of the historical record, and a sharp departure from controlling precedent, I must disagree.

To see this case in proper perspective, it is necessary to focus precisely on the provision of the Pennsylvania Constitution at issue. Article V, Sec. 18(h) governs the confidentiality of the state's Judicial Inquiry and Review Board; it states that "[a]ll papers filed with and proceedings before the board shall be confidential but upon being filed by the board in the Supreme Court, the record shall lose its confidential character."1

Nothing in this provision specifies under what circumstances the record can or must be filed with the Supreme Court. What is abundantly clear is that once it is filed, it is no longer secret. From the time of its founding through June 1, 1983, the Board had disposed of 52 cases on the merits following a hearing, not including the present case. In all but six, the record was filed with the Supreme Court, and thus made public. The proceedings filed included 41 cases where the Board recommended sanctions, including four where the court did not follow the recommendation, as well as five others where the Board had determined no punishment was warranted. Although as the concurrence stresses the record was not filed in six other cases where the Board did not call for sanctions, the overall pattern suggests that these were simply exceptions to its general practice of filing the record after a hearing. The Board's own characterization indicates as much: in pleadings before the district court, the Board declared that "In the past, the record has been filed with the Supreme Court after a hearing has been conducted, whether or not the Board has recommended the imposition of discipline."2 App. at 40.

Until the present disciplinary proceeding was in progress, the Board's filing practices had never been questioned. App. at 235. Indeed, at the time of its filings, the Board stated that it was acting pursuant to "powers provided for in Article V, Section 18, Subsection D of the Constitution of the Commonwealth." App. at 236. After the Board had heard the case against Justice Larsen, however, it abruptly altered its usual procedure, and, for the first time sought an opinion from special counsel on the propriety of filing a record in a case where the Board had not recommended discipline. It was on the basis of this opinion that the Board came to take the position that the state constitution prohibited it from filing a record where charges against the judge were dismissed, after 15 years of practice to the contrary.3II.

The past practice of the Board is directly relevant to one of the two criteria used to resolve a claim of access, whether there has been a history of openness. In Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 100 S. Ct. 2814, 65 L. Ed. 2d 973 (1980), the Supreme Court held that the press and general public have a constitutional right of access to criminal trials. The plurality opinion by Chief Justice Burger relied on a history demonstrating that a presumption of openness inheres in the nature of a criminal trial under an Anglo-American system of justice. Id. at 564-73. Concurring, Justice Brennan wrote that resolution of a claim of access under the First Amendment must be evaluated in light of two factors: "The weight of historical practice" and an assessment of the societal benefits, or "specific structural value of public access in the circumstances." Id. at 597.

This two-step analysis was reaffirmed by the Supreme Court two years later in Globe Newspaper v. Superior Court, 457 U.S. 596, 102 S. Ct. 2613, 73 L. Ed. 2d 248 (1982), which held that the First Amendment prohibited a mandatory closure rule denying the press and public access to trials involving testimony by the minor victim of a sex crime. Citing Richmond Newspapers, the Court noted that the criminal trial historically has been open to the press and the general public. Globe, 457 U.S. at 605, 102 S.Ct. at 2619. The majority then proceeded to evaluate the functional role of access to trials. It concluded that it enhances the factfinding process, fosters an appearance of fairness in our system of justice, and permits the public to participate in and serve as a check upon the judicial process. Id. at 606, 102 S.Ct. at 2619. See also Press Enterprise Co. v. Superior Court, 464 U.S. 501, 104 S. Ct. 819, 78 L. Ed. 2d 629 (1984) (applying two-part test in granting claim of access to voir dire proceedings).

This Court has further recognized a right of access beyond the criminal context. In Publicker Industries, Inc. v. Cohen, 733 F.2d 1059 (3d Cir.1984), we addressed the issue of public access to civil trials. Surveying such venerable authorities as Coke, Hale, and Blackstone, the Court emphasized that "historically both civil and criminal trials have been presumptively open." Id. at 1068. Turning to the second prong of the Globe analysis, we identified an important role for access in the functioning of the judiciary and the government as a whole. We found that access in the civil context served the same functions that the Supreme Court had enumerated in Globe. Id. at 1070. Perhaps most importantly, it was emphasized that access "permits the public to participate in and serve as a check upon the judicial process" and to ensure that the "constitutionally protected 'discussion of governmental affairs' is an informed one." Id. (quoting Globe, 457 U.S. at 604-05, 102 S. Ct. at 2618-19).

We have also confronted claims of access to judicial proceedings which, unlike the criminal and civil trial, lacked a counterpart at common law. In such cases, courts have placed greater reliance on the structural First Amendment values served by access as distinguished from the historical factor. Thus, in United States v. Criden, 675 F.2d 550 (3d Cir.1982), we identified a right of access to a pretrial suppression hearing, despite the absence of a long history of openness. Because "the relative importance of pretrial procedure to that of trial has grown immensely in the last two hundred years," we concluded that "we do not think that historical analysis is relevant in determining whether there is a first amendment right of access to pretrial criminal proceedings." Id. at 555.

More recently, in United States v. Smith (Appeal of the Patriot News Co.), 776 F.2d 1104 (3d Cir.1985), we held that there exists a presumptive right of access to a bill of particulars because of its close relationship to the indictment, which has historically been available to the public. Although "[b]ills of particulars, relative to indictments, have a brief history," we noted that in recent years indictments have become somewhat skeletal, and bills of particulars have thus grown to perform a large part of the role originally played by indictments. Id. at 1111. In addition, public access to such bills "serves the same societal interests served by access to the charging documents." Id. Thus, the bill of particulars was held entitled to the same presumption of openness as the indictment.

These cases frame the appropriate inquiry for a court reviewing a claimed right of access. The inquiry entails review of both the history of openness and the functional values served by access. In examining these elements, a court must be guided by the purposes of the First Amendment, and be sensitive to historical evolution. Where new structures develop to fulfill the role earlier played by other structures, we must seek guidance from the history of the earlier institutions, as in Smith. Moreover, where a particular judicial proceeding has become a critical governmental institution only recently, we place correspondingly greater weight on the structural values furthered by access, as in Criden.

In the present case, the Board itself has a relatively brief institutional history. Nationally, more than two-thirds of the present-day judicial-discipline organizations came into existence during the last 15 years. See I. Tesitor & D. Sinks, Judicial Conduct Organizations 19-27 (1980). Pennsylvania's Board began operation in 1969. While the state schemes vary as to the stage at which the board proceedings become public, the general practice of the Pennsylvania board has been to file a record with the Supreme Court, and thus allow public access after the Board has held a hearing on whether to recommend formal discipline. The Board's policy on confidentiality was adopted as a balance between the need to protect the "judiciary ... from frivolous complaints," and the recognition that "at the same time, the public must be made to know that its real complaints will be heard." McDevitt, Reviewing the Review Board, 42 Pa.B.A.Q. 33, 36 (1970-71); see also Jones, State of the Judiciary in the Commonwealth, 45 Pa.B.A.Q. 152, 155 (1974) (Board has function of "providing the public with the assurance that the conduct of judges at all levels is really being policed...."). Thus the history of the Board up to the time of this case, though not a long one, demonstrates a tradition of openness in its proceedings consistent with the performance of its disciplinary function.

More important to the inquiry is the history of functionally similar procedures designed to discipline judges. See Comment, The First Amendment Right of Access to Civil Trials After Globe, 51 U.Chi.L.Rev. 286, 291 (1984). The analogue to the Board discipline, as the majority notes, is impeachment and removal for conviction of a crime. An impeachment proceeding has always been a quintessentially "public business." C. Black, Impeachment: A Handbook 19 (1974). According to the majority, however, the analogy is inapplicable because "[t]he Board's functions are intended to supplement rather than replace" the impeachment process. Ante at 472.

Examination of the debate leading to adoption of the Article V, however, undercuts this assertion. Creation of the Board grew out of a recognition that "[i]mpeachment is rarely used because it has been shown to be almost ineffective." Laub, Issues Before the Judiciary Committee of the Pennsylvania Constitutional Convention, 39 Pa.B.A.Q. 390, 397 (1969). Thus a new method was needed. It is true that the Board was granted powers to punish more diverse types of judicial misconduct than could be remedied by impeachment, which was reserved for punishment of criminal acts of office and breaches of positive statutory duty. Id. at 397. Nevertheless, under the pre-1968 regime, impeachment was the primary means available for disciplining judicial misconduct; afterward, by contrast, action by the Board became the primary means. In functional terms, then, the proceedings before the Board replaced impeachment. Indeed, there have been no impeachment proceedings since the Board was created.

Even if we were to accept the majority's label of "supplement" rather than "replacement," moreover, we would still be compelled to look to the prior history of impeachment in analyzing the history of access to judicial disciplinary proceedings in Pennsylvania. To do otherwise would appear to ignore the teaching of Smith. There we noted that bills of particulars in recent years have come to perform part of the function initially served by the indictment, and so are now fairly seen "as supplements to the indictment," id. at 1111, subject to a comparable right of access. Similarly, Board discipline, under any characterization, has at the very least come to play a major portion of the role formerly served by impeachment, and thus must be similarly open.

The majority's assertion, then, that because of the lack of open proceedings in the judicial disciplinary process, the Coalition bears a burden of overcoming a presumption of confidentiality rests, I believe, on a misreading of the historical record. Together, the practice of the Board and the prior practice of its functional equivalent support the Coalition's argument that history comports with a presumption of openness.

Against this background, we turn to the second step of the Richmond-Globe analysis, which calls for consideration of the structural values served by access. These values overwhelmingly support a presumption of access.

The majority asserts that in some circumstances acceptance of the structural considerations of the First Amendment could lead to "an unjustifiably expansive interpretation" of that constitutional provision, ante at 473, and that, as Justice Brennan has argued, the right to know "must be invoked with discrimination and temperance." Richmond Newspapers, 448 U.S. at 588, 100 S.Ct. at 2833 (concurring opinion). As general propositions, these admonitions are unexceptionable. But they have little bearing on this case. Here, we consider the structural values of access in the context of a historically open proceeding, where the right of access sought attaches at the very point at which access generally has been granted. Moreover, this point occurs at a stage in the disciplinary proceedings where a review panel has found the complaint not to be frivolous, and where the subject of the complaint has had the benefit of a full adversary hearing. Thus, in neither theoretical nor practical terms does consideration of structural values in this case pose the danger of an "unrestrained right to gather information," Zemel v. Rusk, 381 U.S. 1, 16-17, 85 S. Ct. 1271, 1280-81, 14 L. Ed. 2d 179 (1965), a factor which apparently concerns the majority. Ante at 474.

One critical value furthered by access is its role in "heightening public respect for the judicial process." Globe, 457 U.S. at 606, 102 S. Ct. at 2619. The facts of this case dramatically illustrate the importance of this value. As the Pennsylvania Commonwealth Court stated in a related state court litigation, the underlying nature of the present matter included "allegations of rampant racism, vote-fixing, political maneuvering, personal eccentricities, and the indiscreet, if not unlawful, use of the influence of [a Justice's] office for the political and personal benefits of his friends." In re Subpoena Served by the Pennsylvania Crime Commission, 79 Pa.Commw. 375, 380, 470 A.2d 1048, 1051 (1983). The Commonwealth Court found that this matter had "shaken and continues to shake the public confidence in the administration of justice in this Commonwealth." Id.4

To deny public access to the disposition of such charges once they have reached a formal stage may well create an impression that the Board's "decisions [are] based on secret bias or partiality," Richmond Newspapers, 448 U.S. at 569, 100 S.Ct. at 2823 (plurality opinion of Burger, C.J.); see Criden, 675 F.2d at 556, or stem from "parochial protectiveness." In re Subpoena, 79 Pa.Commw. at 402, 470 A.2d at 1060. Completely closed Board proceedings may thus "breed suspicion of prejudice and arbitrariness, which in turn spawns disrespect for law." Richmond Newspapers, 448 U.S. at 595, 100 S.Ct. at 2837 (Brennan, J., concurring). The most stringent set of ethical standards for elected officials would be of limited worth if the public is not persuaded that the standards are being fairly enforced. Legitimacy rests in large part on public understanding. See Note, Judicial Removal--Establishment of Judicial Commission for Removal of Judges Precludes Legislative Investigation of Judicial Misconduct, 84 Harv.L.Rev. 1002, 1009 (1971).

A related but important benefit secured by access is a "community therapeutic value." Press-Enterprise, 464 U.S. at 508-09, 104 S.Ct. at 823; Richmond Newspapers, 448 U.S. at 570-71, 100 S.Ct. at 2823-24 (plurality). Allegations of misconduct by judges who are charged with meting out justice often provoke public outcries. A chance to verify that a public agency like the Board is effectively enforcing ethical rules provides an outlet for such reactions. As the Executive Director of the Board has stressed, "there is real value in giving citizens a sounding board for their grievances." McDevitt, supra, 42 Pa.B.A.Q. at 35.

Public access also demonstrably advances the First Amendment's "core purpose of assuring freedom of communication on matters relating to the functioning of government." Richmond Newspapers, 448 U.S. at 575, 100 S. Ct. at 2826 (plurality). Access here would educate the public both about the Board and the judiciary. See Comment, supra, 132 U.Pa.L.Rev. at 1183 (1984). A public informed by access would in turn be better equipped to serve as a check on the judicial system, an essential component in our form of government. See Globe, 457 U.S. at 606, 102 S. Ct. at 2619. As to the Board, the knowledge that the press and public may review its actions will act as a safeguard against bias or partiality, and enable the public to evaluate the Board's performance of its watchdog role. As to the judiciary, access to Board proceedings will enable citizens to evaluate judges, and make more effective decisions when the time comes for electing or re-electing them. This is especially appropriate in Pennsylvania which selects its judges by popular election.

Finally, opening Board hearings will likely make them more effective by enhancing their integrity. Factfinding before the Board may include production, and admission of documents, and testimony by witnesses. As in the criminal trial context, public scrutiny would discourage perjury or other misconduct. See Richmond Newspapers, 448 U.S. at 569, 100 S.Ct. at 2823 (Burger, C.J.) (plurality); Shaman and Begue, Silence Isn't Always Golden: Reassessing Confidentiality in the Judicial Disciplinary Process, 58 Temple L.Q. 755, 771 (1985). Access would serve an important purpose as "security for testimonial trustworthiness." Publicker, 733 F.2d at 1070.

Given the strength of these structural values, and the history of the Board's practices as well as the predecessor impeachment process, it seems clear that a presumption of access attaches to the Board's proceedings. To some extent the majority appears to concede this point. Although at one point it likens the Board to the Grand Jury, a secret body, ante at 473, it also "assume[s] a right of access" to Board proceedings at some stage, ante at 472, and states that "what tradition there is, favors public access only at a later stage in the process." Ante at 473. Under this latter view, the right of access apparently attaches at a point the Board now contends it does, when formal sanctions are recommended to the Pennsylvania Supreme Court. This contention, however, suffers from the same defect as the theory that there is a tradition of closure, since, as we have demonstrated, historically access was granted to formal hearings by the Board, regardless of their outcome. Similarly, in impeachment proceedings, access is granted to the trial at which the legislature determines whether to recommend conviction, regardless of whether conviction is recommended or not.

The only other support offered by the majority for its delayed point of access, under its theory of a "temporally based right," ante at 473, is an analogy: the Board's proceedings, the majority contends, are "most akin" to the grand jury, id., rather than the pre-trial and trial proceedings in Richmond Newspapers and its progeny. This is so, it is argued, "because the Board cannot impose, but only recommend, punishment...." Ante at 473.

In fact, however, the Board's real powers are considerably greater than this analogy suggests, and the protections provided the subject of a complaint are correspondingly more stringent than before the grand jury. Unlike a grand jury, the Board may conduct full adversary hearings, designed to reach a conclusion on the merits of the charges, rather than merely a determination of probable cause. Where the Board does recommend sanctions, the Supreme Court does not retry the case, but decides to affirm or reject them on the basis of the transcript of Board hearings. Where the Board dismisses a complaint--as occurs in the vast majority of cases--its decision is final and unreviewable. See First Amendment Coalition, 501 Pa. 129, 460 A.2d 722.5 As a result of its power, in hundreds of cases the Board itself has imposed what it calls "informal" sanctions, including verbal reprimands, letters of admonition, and solicited resignations. While the Board now argues that these are "consensual" sanctions, this label should not cloud the underlying reality: the Board's ability to obtain "consent" to such measures is in large part predicated on its unquestioned ability to effect formal punishment. Indeed, the Supreme Court has accepted the Board's recommendation 90% of the time, a practice apparently contemplated from the start.6 Given its autonomy and power, even representatives of the Board have acknowledged that it "sit[s] first as an investigatory body, then as a quasi-judicial body." McDevitt, supra, 42 Pa.B.A.Q. at 37 (1970).

Unlike the grand jury, then, but like the trial in Richmond Newspapers, judicial inquiry procedures in Pennsylvania are a "mechanism for judicial factfinding" and an "initial forum for legal decisionmaking," 448 U.S. at 596, 100 S.Ct. at 2838 (Brennan, J., concurring), and are thus entitled to the same presumption of openness under the First Amendment.

Concluding that a right of access exists, of course, does not complete the Court's task. Such a right is not absolute. It may be overcome if the state demonstrates that closure "is necessitated by a compelling governmental interest, and is narrowly tailored to serve that interest." Press-Enterprise, 464 U.S. at 510, 104 S.Ct. at 824 (1984) (quoting Globe Newspaper, 457 U.S. at 607, 102 S. Ct. at 2620 (1983)). See Smith, 776 F.2d at 1112. Here, however, the Board cannot meet this standard.

The Board offered three major interests that it contends are sufficient to overcome the right of access. The first is that confidentiality encourages the filing of complaints by providing protection against retaliation. This interest may be quickly disposed of, however, since under Pennsylvania's procedure, the judge under investigation, the person most likely to retaliate, has already been informed of the name of the initial complainant by the time a formal hearing begins. App. at 129-30, 262, 264. Although the subject is apparently not notified of the identities of the witnesses to appear, the disclosure of the complainant and the nature of the complaint will most often suggest those likely to be called. Thus, this interest is not a strong one once the proceedings pass the initial stage.

Two other rationales are suggested by the majority: the interest in the reputation of the judiciary and individual judges, and the need for flexibility in negotiating sanctions. The flexibility interest derives from the Board's ability to encourage judges to resign voluntarily by offering to spare them the publicity of a hearing. See Landmark Communications, Inc. v. Virginia, 435 U.S. 829, 835-36, 98 S. Ct. 1535, 1539-40, 56 L. Ed. 2d 1 (1978). Given the Board's limited resources, and the fact that its jurisdiction encompasses many and varied types of judicial misconduct, this interest is an important one, and unless it can maintain confidentiality while investigating complaints, the Board has no other effective means of preserving flexibility.

I also agree that there is a strong interest in protecting the reputation of judges from unfounded complaints. See Landmark, 435 U.S. at 835, 98 S.Ct. at 1539. But this interest exerts declining force as the Board concludes that the complaint is worthy of a full hearing. The first step by the Board in its investigative stage is to dismiss frivolous claims. It then informally investigates further, winnowing out other meritless claims, and allowing the judge to comment on charges that appear more substantial. If a complaint remains unresolved, the Board then conducts a formal investigation, after which it decides whether to lodge formal charges requiring an adversary hearing. It is during the investigative stage that, "the meritorious can be separated from the frivolous complaints," Landmark, 435 U.S. at 835, 98 S.Ct. at 1539. The Board can use confidentiality as a bargaining chip to resolve some complaints without filing formal charges leading to a hearing, and it can further determine which of the non-frivolous complaints are sufficiently substantiated to merit formal charges.

Once these steps have been completed, the concern over premature disclosure of unfounded complaints virtually vanishes. Moreover, to the extent that there is any remaining danger of unfair injury to reputation, it may be answered by a means other than confidentiality--that is, by disclosure of the full transcript of the proceedings, including the basis of the decision to sustain or dismiss the charges. This, of course, was the Board's practice until the time of this case. Under this procedure, every judge formally accused by the Board but eventually exonerated would be vindicated by the simultaneous disclosure of the transcript. Once the transcript is complete, then, the state can no longer justify its limitations on the right of access, and the press and public, as the district court held, are entitled to inspect the transcript.

In rejecting the Coalition's claim of access, the majority also engages in a balancing of access interests against confidentiality interests, but it never articulates the standard it applies to this weighing. Without citing authority, it seems to suggest that in evaluating the competing concerns, the Board's balancing deserves special deference from this Court. Thus it states that "[t]he coalition has failed to show that the right of access it urges is so compelling as to justify the restrictions on the state's freedom of choice," ante at 473, and that "[f]ederal courts should not overturn a state's evaluation of structural concerns in the absence of egregious circumstances," ante at 475. With all due respect, this approach turns the constitutional analysis upside down. The basic rule is that once a right of access is found to exist, then the burden falls upon the state to justify its infringement. See Globe, 457 U.S. at 606-07, 102 S.Ct. at 2619-20. Moreover, the Supreme Court has declared that a federal court is required to conduct an independent review of the asserted justifications, according to federal constitutional standards. See Landmark, 435 U.S. at 843, 98 S.Ct. at 1543.

To support its application of a "presumption of validity attaching to state legislative and constitutional provisions," ante at 475, the majority cites Clements v. Fashing, 457 U.S. 957, 963, 102 S. Ct. 2836, 2843, 73 L. Ed. 2d 508 (1982). That decision, however, provides little support for the majority's approach. Clements dealt with an equal protection challenge to state constitutional limitations on a public official's ability to become a candidate for another public office. The Supreme Court held that in the absence of special burdens on minority political parties or independent candidates, the state's classification should be analyzed under a lenient rational basis test, rather than the more demanding "strict scrutiny." Id. at 963-65, 102 S.Ct. at 2843-45. Thus the Court applied "traditional equal protection principles," id. at 965, 102 S.Ct. at 2844, developed in a long line of precedents. In this case, by contrast, the asserted right, whose existence is conceded by the majority, is a First Amendment right of access. The majority points to no case allowing for a rational basis review in a First Amendment access case, nor could it; once a right of access is found to exist, any infringement of that right must meet the "compelling interest" test set forth in Globe.

Ultimately, under the First Amendment, any attempt to justify the Board's position must establish that its current view of confidentiality is necessary to its operation. The majority hints that it believes this, see ante at 476-77, but it does not say so directly. The reason it does not should be clear enough: for its first 15 years, the Board operated without that degree of confidentiality, and there is no evidence that shows any problems as a result. Against such a record, a claim of necessity is anomalous indeed.

As judges, we instinctively react against public exposure, perhaps because we know the importance of confidentiality to certain aspects of our work, such as decisional conferences. In this context, secrecy is a goal worthy of protecting. But we have no general authority to "protect the court as a mystical entity or the judges as individuals or as annointed priests, set apart from the community and spared the criticism to which in a democracy other public citizens are exposed." Bridges v. California, 314 U.S. 252, 291-92, 62 S. Ct. 190, 207-08, 86 L. Ed. 192 (1941) (Frankfurter, J. dissenting). There is no legal support for the proposition that elected state officials who happen to be judges are entitled, in Judge Pollak's words, to "total exemption from the slings and arrows which others in the public arena must live with." 579 F. Supp. at 214. In the final analysis, what is most regrettable about today's decision is that in the absence of any other convincing rationale, it too will be seen as an example of the judiciary protecting its own interests.7

This case comes at a critical time in the history of the judiciary. Judges at both the state and federal level are receiving unprecedented scrutiny not only from the press and public, but also from law enforcement officials. Prosecution of judges for misconduct, unfortunately, has become increasingly common.8 As a result, the public needs assurance that it can continue to rely on the independence and impartiality that are the cornerstone of our judicial system; it needs to know that the non-criminal mechanisms created to monitor judicial conduct are functioning with vigor and effectiveness. In Pennsylvania, the Review Board's critics cry out that its decisions are based on partiality and self-protection. Nothing could give more credence to these charges than fencing out the public from the Board's proceedings; and nothing could rebut them more effectively than disclosure where complaints have been denominated by the Board as serious.

Because I believe that the First Amendment requires access to proceedings involving non-frivolous complaints regarding high elected officials, I would affirm the judgment of the district court, and order that the transcript of the hearing be made public.

Judges GIBBONS, SLOVITER and MANSMANN join in this dissent.